SECTIONAL TIMING

“An Introduction by Timeform”.

 Simon Rowlands has worked at, or for, Timeform since 1986,
with the exception of a short period earlier this century, and
is now Head of Research & Development.

He has developed many numerical and statistical ways of

analysing form over the years, often utilising algorithms and
automation. He launched Timeform’s Irish service in 2007,
the free-form US/Canadian service in 2010 and was involved
in the successful launch of TimeformUS in 2013. Simon is also
a recognised industry-wide authority on sectional timing
and has blogged extensively on the subject, including the
popular ‘sectional debrief’ series on timeform.com

Simon can be followed on Twitter @RowleyfileTF

© TIMEFORM LIMITED 2015

COPYRIGHT AND LIABILITY
Copyright in all Timeform Publications is strictly reserved by the Publishers and no material therein
may be reproduced stored in a retrieval system or transmitted in any form or by any means electronic
mechanical photocopying recording or otherwise without written permission of Timeform Limited.
Sectional Timing: An Introduction by Timeform is published by Timeform Limited, Halifax,
West Yorkshire HX1 1XF (Tel: 01422 330330 Fax: 01422 398017;
e-mail: [email protected]). It is supplied to the purchaser for his personal use and on
the understanding that its contents are not disclosed. Except  where the purchaser is dealing as a
consumer (as defined in the Unfair Contract Terms Act 1977 Section 12) all conditions warranties or
terms relating to fitness for purpose mer­chant­ability or condition of the goods and whether implied
by Statute Common Law or otherwise are excluded and no responsibility is accepted by the Publishers
for any loss whatsoever caused by any acts errors or omissions whether negligent or otherwise of the
Publishers their Servants Agents or otherwise.
 AN INTRODUCTION BY TIMEFORM

CONTENTS
AN INTRODUCTION BY TIMEFORM 4

WHAT IS SECTIONAL TIMING? 7

SECTIONAL TIMING IN ACTION 11

HOW RACES TEND TO BE RUN 17

HOW TO RUN FAST: ESTABLISHING SECTIONAL PARS 22

SECTIONAL TIMES IN ACTION: LINGFIELD 2014 AND 2015 27

SECTIONAL TIMING IN ACTION, ELSEWHERE 31

TAKING YOUR OWN SECTIONALS BY HAND 38

UPGRADING SECTIONAL PERFORMANCES 43

ILLUSTRATING SECTIONAL UPGRADES 46

FAQS ABOUT SECTIONALS 50

IN-PLAY SECTIONALS 54

THE FUTURE OF SECTIONALS 57

APPENDIX 58

3
 SECTIONAL TIMING

CHAPTER 1
AN INTRODUCTION BY TIMEFORM
According to some authorities, both the title and the claims of a book
need to be arresting and compelling, regardless of the content. The
object of a book is to sell that book, not necessarily to deliver upon
what is promised by the title or the claims made in its pages.
That is not the case with this booklet, which in any case is free. From its title –
“Sectional Timing: An Introduction by Timeform” – to the claims made within it, the
intention is to state no more than can be justified.
In general terms, it is an “introduction” to the subject of sectional timing, rather than
the definitive last word on the matter: sectional timing is in its infancy, in Britain and
Ireland at least, and any discussion of the subject needs to acknowledge that.
More specifically, no claims are made that sectional timing will provide “a passport
to riches”, or that this booklet will make that happen. Sectional timing is another way
of looking at the age-old conundrum of racing analysis. Analysis of sectional times
is a powerful tool, if used properly, but it needs to be used in conjunction with more
established forms of analysis rather than viewed as a silver bullet.
Nonetheless, it is possible to make a claim of sectional timing which is realistic for
anyone who reads this booklet, in whole or in part, namely:
An appreciation of sectional timing will increase your understanding of, and
enjoyment of, horseracing and may help you to strike more good bets.
The pages which follow will attempt to take the reader on the road to that
understanding, from a description of what sectional times are – via how to place
sectionals into context, how to take sectional times of your own and upgrade
performances on the back of sectionals – to using sectionals to tackle in-play betting.
Illustrations will be given along the way from real-life sectional performances.
These have been taken from those provided electronically by the likes of TurfTrax
(sometimes through funding provided by the British Horseracing Authority and
the Department of Culture, Media and Sport) and from a now-extensive archive of
sectional times sourced manually by Timeform itself.

4
 AN INTRODUCTION BY TIMEFORM

It is beyond the scope of this booklet to deal at length with sectional timing (also
known as fractional timing) outside Britain and Ireland, but readers are advised
that much has been written and spoken on the subject elsewhere. The existence
of comprehensive sectional timing enables the pioneering analysts at TimeformUS
to assess racing in North America in a more multi-dimensional way than their
competitors.
Media coverage of sectionals in Britain and Ireland is sadly limited, but the
outstanding work in this area of James Willoughby, particularly on dedicated racing
channel RacingUK, must be flagged up. James has been a tireless and persuasive
advocate of sectional timing for a long time.
A basic premise of sectional analysis is that there is an efficient way for a horse to run
in a given circumstance, and a significant departure from that efficiency will result in a
significant cost to that horse’s overall time.
Efficiency maximises average speed and minimises overall time: inefficiency
decreases the former and increases the latter. This is basic physics, not to mention
common sense.
The results of races occur from the interaction between this basic truth and wider
considerations of ability and aptitude, played out for each and every horse within that
race.
Far from being revolutionary, this has been understood – at least implicitly – in
horseracing for a very long time. Any race reader worth his or her salt will understand
that a horse must be disadvantaged by being set too much to do, by being asked to
go too fast, by being expected to show stamina when sprinting is its forte, and so on.
Sectional timing formalises many of the sensible tenets of race reading, while
banishing uncertainty regarding measurements. One consequence of this increased
confidence is that sectional analysis based on numbers, rather than merely on
impressions, can move things on several steps further.
That fast-finishing second can be rated the winner – visual race reader and sectional
timer may agree on that – but by how much? The creation of sectional datasets, and
the considered analysis of those datasets, allows for the possibility of quantifying such
things at the very least.

5
 SECTIONAL TIMING

If you accept that energy distribution and the pace of a race influences results – as
the laws of physics dictate – it then becomes a question of how you measure these
phenomena: precisely, through sectionals (electronically or by using a video and
stopwatch), or by using your eyes.
There is no doubt that experienced race readers are capable of doing a good job of
interpreting a race without having sectionals to hand. But there is also no doubt that
it is impossible to gauge visually the difference between travelling at, say, 35 mph and
36 mph with any consistency or precision.
And, yet, that difference – the difference between 12.86 sec/furlong and 12.50 sec/
furlong – is more than enough to influence, potentially, the outcome of a race. The
difference is more than two lengths per furlong, or more than 16 lengths per mile.
Recording sectionals precisely may not always be possible, but it seems obviously
advantageous when it is. Not only does doing so mean that visual impressions
can be confirmed, or sometimes refuted, with confidence; it also means that you
have numbers which may be manipulated to tell you very much more than general
impressions do.
Timeform was founded on analysis of overall race times. Sectional analysis attempts
to put the skills first deployed in visual observations into a formalised timing
framework, much like was done with overall times originally. Along the way, some
surprising and highly illuminating insights have become apparent.
Sectional-timing analysis can be laborious, depending on the method of delivery and
the means for putting the information into context. The time spent on it could be
spent profitably on other forms of analysis. But there should be no doubt that analysis
of sectional times itself has merit.
To the uninformed, most racing seems the same: it is “much of a muchness”. No
two races are ever quite the same from a sectional-timing point of view, however.
Every event has its sub-plot of pace angles and stamina-related nuances. Otherwise
mundane racing can suddenly become gripping when viewed through the prism of
sectional timing. What’s more, good bets and clever angles can be found that occur to
few others.
If nothing else, that seems reason enough to attempt to find out more about the
subject.

6
 AN INTRODUCTION BY TIMEFORM

CHAPTER 2
WHAT IS SECTIONAL TIMING?
The importance of timing of horseraces to allow for comparisons
between those races and the horses within them has long been
established, even in Britain and Ireland where “the clock” has seldom
been considered as important as in, say, the US. A look at the record of
The Derby at Epsom shows that official times (initially to the nearest
second) for the race as a whole existed in the middle of the nineteenth
century.
Relatively little was done with these times to begin with, but one pioneer, Phil Bull,
analysed them mathematically (and took many of his own) between the Wars, an
enterprise so successful that it eventually gave rise to Timeform in 1948.
Bull was primarily concerned with overall times: the time from start to finish of a
race. Sectional times are overall times broken into sections, either electronically or
manually.
Much as a bare result – Horse A beat Horse B by so much – does not tell you of the
many incidents that gave rise to that result, overall times, while important, do not tell
you how those overall times came about.
Times need to be put into context, always. A time of, say, 100.00s may be fast in one
context and slow in another. In addition to the obvious considerations of course,
distance and speed of surface, the abilities of the horses themselves need to be taken
into consideration. That 100.00s might be fast for a poor horse and slow for a good
one under precisely the same circumstances.
Overall times may also be heavily influenced by pace. That 100.00s might have been
achieved by the runners going flat out early and hanging on grimly by the end; or
it might have been achieved by a slow early pace followed by a sprint (or by any
variation in between).
At both extremes, and for most of the possibilities in between, the overall time will
have been compromised by the manner in which it was achieved.

7
 SECTIONAL TIMING

Sectional times record the times taken by horses between established points.
Providing the distance between those points is known, or can be estimated to
acceptable accuracy, the sectional speed of the horse in question can also be
deduced.
As sectional times deal with distance and time, they are, in an important respect,
materially no different to overall times, which deal with time over the distance of an
entire race: an overall time of a race is essentially a sectional which covers the whole
of a race.
While the analysis of overall race times has a fairly lengthy history in Britain, the
analysis of sectional times does not, in no small part due to the fact that they have
often been absent. While such times were recorded as far back as the nineteenth
century in the US, they have appeared only recently, and only patchily, in Britain.
Wolverhampton Racecourse and Newmarket Racecourse produced sectional times
late in the twentieth century; the technology company TurfTrax produced them
extensively for all all-weather racing in Britain from 2005 to 2008; more recently,
TurfTrax have provided sectional times for most of the British Champions Series and
for races at Lingfield which form part of the All-Weather Championships.
TurfTrax’s technology involves a transmitter in the horse’s saddlecloth which relays
a signal several times a second to portable receivers positioned around the course,
so that the horse’s positions, and the time taken between those positions, can be
established and converted into sectional times.
Other technologies that have been investigated, or may yet be investigated, include:
break-beam technology (in use in America), which records the times of the leaders
only; permanent wireless systems, such as those used successfully elsewhere in the
world by Trakus and trialled in 2014 in Britain by RaceTech; and Global Positioning
Systems, which would obviate the need for much on-course infrastructure.
All such technologies cost money – plenty of money in most cases – and British
racing has not yet found a way of funding sectional timing widely but in an affordable
manner. TurfTrax’s comprehensive coverage of British all-weather racing from 2005
to 2008 was subscription-based but was shelved when the company had to be
restructured (but, fortunately, only after a very useful sectional dataset had been
established).

8
 AN INTRODUCTION BY TIMEFORM

In the absence of high-technology solutions, and even in their presence, there is

much to be said for the far more straightforward and affordable approach of timing
races from video coverage. There can be some issues regarding camerawork and
continuity, but “manual” sectionals are thereafter free from the sorts of glitches from
which higher-technological solutions tend to suffer.
Conventional television pictures are fit for purpose in most cases, but online replays
can also be captured and analysed using editing software to make timing easier.
Video validation is, according to some accounts, used extensively in Hong Kong
and Australia, despite those major racing jurisdictions having sectional-timing
technologies of various degrees of advancement. In Britain, the manual sectional
cause has recently been advanced by At The Races providing a live on-screen
stopwatch, an initiative which other broadcasters would do well to follow.
Against a backdrop of non-existent or at best sporadic sectional timing provision,
Timeform has been taking manual sectional times since the 1990s, selectively to
begin with, but more widely of late. It assists in the understanding of races, and of
individual performances within those races, and feeds into Timeform’s analysis on a
daily basis.
The TurfTrax data from 2005 to 2008 gave rise to the establishment of sectional
pars at the courses which were covered by it, and to a solid understanding of how
races tend to be run and of what constitutes efficiency. Timeform now has its own
comprehensive database of manually acquired sectional times – and, in many cases,
sectional pars – which enables further analysis of the subject independently of
whether sectional times are being provided electronically and officially.
In the absence of electronic sectionals, there is much to be said for filling the gaps –
“the yawning chasms” might be a better term – with manual times derived from video
analysis. As a result, one of the following chapters will deal specifically with how the
reader can compile his or her own sectional times.
This applies to turf racing as much as all-weather racing, and to jumps as well as Flat:
the laws of physics do not suddenly disappear because a horse runs on grass rather
than synthetics, or over obstacles rather than on the level.
That said, as with overall time analysis, there are factors specific to different
circumstances which need to be taken into consideration. Rail movements are more

9
 SECTIONAL TIMING

frequent on turf Flat than on all-weather Flat; overall distances, sectional distances
and even overall times are open to dispute in jump racing.
These factors complicate time analysis – whether it be analysis of overall times or of
sectional times – but have not rendered, and should not render, that analysis without
worth. Sectional timing can be challenging, even before you get to the analytical part,
but the upside is that it provides a very real opportunity for those prepared to invest
the time, effort and acquired knowledge to get ahead of the crowd.
Over and above the increased enjoyment that is likely to result from knowing a
subject better than do others, getting ahead of the crowd is what punters everywhere
aspire to, after all.

10
 AN INTRODUCTION BY TIMEFORM

CHAPTER 3
SECTIONAL TIMING IN ACTION
May 4th 2002, The Rowley Mile Course at Newmarket. The Two
Thousand Guineas, the first colts’ classic of the British season, had
attracted 22 runners, but one of them stood head and shoulders
above his rivals as the short-priced favourite.
Hawk Wing had won three of his four starts as a two-year-old and was spoken of in
reverential terms by those closest to him. Despite having not run for eight months,
Hawk Wing was sent off at 6/4 under Jamie Spencer. To some, it seemed to be a case
of “the bigger the field, the bigger the certainty”.
What happened next sparked a debate that raged long after: an early controversial
chapter in what became a controversial career for the horse.
With such a large field on a straight mile course, it was hardly surprising that the
runners split into three groups, with Hawk Wing in the stand-side one. By halfway,
Hawk Wing was some way back in his bunch and several lengths adrift of Redback,
who led on the centre-to-far side. Meanwhile, Hawk Wing’s talented stable-
companion Rock of Gibraltar was tracking the pace towards the far side.
Rock of Gibraltar moved through to lead in the penultimate furlong, with Hawk Wing
finally in full cry but still a few lengths down. The latter gained throughout the final
furlong but was still a neck down at the line.
Due to the initiative shown by Newmarket Racecourse in funding sectional timing, it
was possible to dissect what had happened from a time point of view. The sectionals
for Rock of Gibraltar and Hawk Wing, post-processed from the originals to allow for
some small discrepancies due to the technology used, are shown in isolation below.

Two Thousand Guineas, Newmarket, 04 May 2002 - Individual Sectionals

Horse After 3f 3f to 4f 4f to 5f 5f to 6f 6f to 7f 7f to line Overall
ROCK OF GIBRALTAR 38.69s 11.43s 11.20s 11.37s 11.35s 12.46s 96.50s
HAWK WING 39.06s 11.58s 11.25s 11.29s 11.31s 12.07s 96.56s

11
 SECTIONAL TIMING

Hawk Wing looked unlucky, but what did those sectionals say? They identified that
he made up 0.39s on Rock of Gibraltar in the final furlong, and 0.08s and 0.04s in the
two furlongs that preceded it. But they also identified that, despite visual impressions
to the contrary, both colts ran the final furlong more slowly than the furlongs that
immediately preceded it.
Anyone who knows Newmarket Racecourse will point to the topography of the track.
The final furlong on the Rowley Mile course rises in the region of 13 feet, according to
Google Earth, which is a pretty significant incline.
All other things being equal, we would expect horses to be slowing at the end of a
race at Newmarket. We would also expect horses to start to tire at the end of a race.
It is clear that those individual sectional times need context. They need the context of
the course at which they occurred, the distance over which they were achieved, and,
not least, the horses to which they apply. We should expect a horse of Hawk Wing’s
ability to run faster in a given furlong than an average horse, but how much faster is
another matter entirely.
The first step in making sense of such information is to convert the raw times into
something which can be compared across different circumstances, intuitively and
mathematically. For this, it is necessary to tackle one of the few equations in this
booklet, an equation which is central to much of what follows:

A horse’s finishing speed, expressed as a % of its average race speed, is derived from
(100*T*d)/(D*t).

Where “T” is its overall time (in seconds), “d” is the distance of the sectional to the
finish (in furlongs), “D” is the overall race distance, and “t” is the sectional time (“*”,
“/” and the use of brackets follows the Microsoft Excel convention of multiplication,
division and completion of calculations within those brackets before reinserting the
result into the equation).

First off, it is necessary to express those earlier individual times as times for the last
furlong, two furlongs, or whatever, from which finishing speed %s can be easily
calculated according to the above equation:

12
 AN INTRODUCTION BY TIMEFORM

Two Thousand Guineas, Newmarket, 04 May 2002 - closing sectionals

Horse Last 5f Last 4f Last 3f Last 2f Last 1f Overall
ROCK OF GIBRALTAR 57.81s 46.38s 35.18s 23.81s 12.46s 96.50s
HAWK WING 57.50s 45.92s 34.67s 23.38s 12.07s 96.56s

As an example, Rock of Gibraltar’s last-two-furlongs finishing speed % in the 2002 Two

Thousand Guineas was (100*96.50*2)/(8*23.81) = 101.3%
Hawk Wing’s last-two-furlongs finishing speed % was (100*96.56*2)/(8*23.38) =
103.3%
Those finishing speed %s show that both horses were running at average speeds in
the last quarter of a mile that were slightly faster than their average speeds for the
mile race overall.
However, if their final furlongs are considered, the figures become (100*96.50*1)/
(8*12.46) = 96.8% for Rock of Gibraltar and (100*96.56*1)/(8*12.07) = 100.0% for Hawk
Wing.
Rock of Gibraltar was slowing in the final furlong, compared to his average speed for
the race itself, while Hawk Wing was running at exactly the same average speed as he
was for the race overall.
The important thing about finishing speed %s is that they are ratios. They place a
horse’s sectional time and speed into the context of that same horse’s overall time
and speed. As such, they are independent of many of the factors which make raw
sectional times of limited use in isolation.
A finishing speed % also implies something significant about the remainder of the
information that has been used to calculate it. If a horse is finishing faster than its
average race speed, it follows that it was running slower than its average race speed
earlier in the race. Conversely, if a horse’s finishing speed % is low, its speed compared
to its average race speed must have been higher earlier.
An appreciation of this reality – that a finishing speed % not only explicitly describes
one phenomenon but implicitly describes another – is crucial for an understanding of
finishing speed %s.

13
 SECTIONAL TIMING

Also crucial for an understanding of sectionals more widely is the stark fact that
the best way for a horse to get from A to B in the shortest possible time is to run
efficiently.

“Running efficiently” means running evenly, or something close to that. Physics and
biomechanics tell us that there is a price to be paid for running faster or slower at one
stage of a race than another.
A horse will slow down more as a result of having gone too quickly early than it
gained from those earlier exertions; conversely, it will lose more time by going too
slowly early than it gains by then going quickly at the end. Even pace maximises
energy use and minimises overall time.
Reality can be subtly different, but only because there are complicating factors, not
because the theory is wrong.
Races are timed from a standing start, in Britain and Ireland at least, so an opening
sectional will include acceleration and cannot be as fast as it might be otherwise;
bends and inclines/declines will affect energy distribution; horses will usually begin to
slow just before the end of a race, else a jockey will be guilty of having energy in the
bank which has gone unused. All of these factors will be considered in more detail in
the next chapter.
There are also a multitude of reasons why a horse may win a race by running in an
inefficient manner. It may be advantageous to get out in front – to take the shortest
route or to avoid kickback, for instance – or it may be acceptable to run inefficiently if
others in the race are running inefficiently too.
So-called “pace biases” are often (even when not completely illusory) in reality
“track biases” connected to a physical advantage in assuming a position at a small
expense in efficiency. Pace varies on a by-race basis, and the correct way to look at the
possibility of a bias connected to pace is on a by-race basis.
The important thing to bear in mind is that inefficiency will have a negative effect
on a horse’s overall time – this is an unambiguous physical law – even if it does not
necessarily have a negative effect on that horse’s chance of winning a race in the
complicated and highly tactical environment of a horserace.

14
 AN INTRODUCTION BY TIMEFORM

There are a multitude of possible applications of sectional times, including simple

descriptive ones which tell us that, say, Horse A ran faster at a certain stage than
Horse B, or that Horse C passed all of its rivals in the last two furlongs of a race.
But perhaps the most powerful application is as an extension of conventional time
analysis. On one important level, the deconstruction of overall times into sectionals
simply gives us a far more powerful, informed and multi-dimensional way of tackling
an age-old racing conundrum: namely, what a horse’s timed performance tells us
about its ability.
The Two Thousand Guineas of 2002 served as an illustration of how an individual
horse’s sectionals can be converted into a finishing speed % which is at once intuitive
– a figure over 100% indicates that the horse finished quickly compared to its average
race speed, one under 100% indicates the opposite – and have wider applications.
Sectional times for individual horses are not always (indeed, not usually in Britain
and Ireland) available, but we can also learn a lot about the general run of a race by
considering headline figures for that race, and that can often be done easily.
If all we knew was the time lapse between the leader passing a given point and the
leader (that is, the winner) at the finishing line, we could compare this with the overall
time for the race to calculate a “race” sectional. This would tell us – in general terms,
not specific ones about individual horses – whether the race was run at a quicker or
slower relative pace than usual, given the course, distance, conditions and abilities of
the horses concerned.
In the 2002 Two Thousand Guineas, the leader two furlongs out was the
aforementioned Redback, who eventually finished third. He reached that juncture
after 72.47s. It follows that the “race” sectional for the last two furlongs is the overall
time for the race itself minus this leader’s time, or 24.03s.
By using the same finishing speed calculation as previously, the finishing speed figure
for the race itself can be seen to be (100*96.5*2)/(8*24.03) = 100.4%.
Overall, the race finished slightly faster than its average speed, at least as judged
by the time of the last two furlongs of it. Incidentally, the race sectional for the final
furlong is the same as for Rock of Gibraltar individually – 96.8% – for that horse was in
the lead at both the furlong marker and the finishing line.

15
 SECTIONAL TIMING

Both types of information – race sectionals and individual horse sectionals – can
increase understanding of what happened in a race, though individual sectionals are
more powerful, if more difficult to obtain manually.
In both cases, evidence of inefficiency would be evidence that the horse, or horses,
could have run faster. In both cases, a slow early pace and a fast finish would indicate
that speed was tested more than stamina, while a fast early pace and a slow finish
would indicate the opposite.
A slow-early/fast-late race profile may well indicate that horses prominently placed
when the pace quickened were advantaged, while a fast-early/slow-late profile may
well indicate the opposite (though, as will be shown, this is in reality overly simple).
This is familiar territory for the experienced visual race reader, but the crucial
difference is that there should be little or no opportunity for error when such
impressions are based on hard figures. Even better, those hard figures mean that the
degree to which these impressions are true can be quantified and incorporated into a
wider knowledge of the subject matter.
Where Hawk Wing, Rock of Gibraltar and the 2002 Two Thousand Guineas are
concerned, we need to know more about how races tend to be run, and in particular
about how to run a mile at Newmarket’s Rowley Course efficiently, before we can
state with confidence whether one horse was superior to another once the sectionals
have been taken into account.
Such matters will be the subject of the next few chapters. In the meantime,
however, the finishing speed % metric is perhaps the most important single piece
of information in this booklet. The simple (100*T*d)/(D*t) calculation provides the
foundation of a better understanding of sectional timing, of what has taken place,
and of what may yet take place, both intuitively and at a deeper level.

16
 AN INTRODUCTION BY TIMEFORM

CHAPTER 4
HOW RACES TEND TO BE RUN
At one basic level, sectional times can tell us about how races tend to
unfold. Horses accelerate from a standing start at the beginning of a
race, so common sense dictates that the opening furlong will not be
as fast as subsequent ones (providing that standing start is included
in the overall race time, which is the case in Britain and Ireland but not
always elsewhere).
Common sense (and physics) also dictates that running round a bend will slow a
horse down compared to running the same distance unimpeded in a straight line.
Inclines also slow down a horse, but, as physics shows, by less than for the equivalent
decline (a recent study which claimed to prove the opposite was questionable to say
the least).
It is also to be expected that the majority of horses will begin to tire by the very end
of a race. If a horse does not, then it has unexpended energy, and the jockey is likely
to be guilty of failing to get the most out of it.
We should expect a slow (but accelerating), then faster, then slowing tempo to most
races, though with those speeds affected at any juncture by bends and undulations.
Sectionals show that is exactly what we get, as the following tables for races at short
of a mile on British AW tracks illustrate:

Winners' median by-furlong times at AW courses, 2005 to 2008 (all races) FIVE FURLONGS
Course Count Opening 1f 4f-3f out 3f-2f out 2f-1f out 1f out-line Overall
KEMPTON 68 13.73s 11.35s 11.75s 11.66s 11.98s 60.47s
LINGFIELD 85 14.13s 11.08s 10.85s 11.24s 11.64s 58.94s
SOUTHWELL 38 13.41s 10.83s 11.20s 11.72s 12.89s 60.05s
WOLVERHAMPTON 135 14.74s 11.30s 11.91s 11.95s 12.55s 62.45s

17
 SECTIONAL TIMING

Winners' median by-furlong times at AW courses, 2005 to 2008 (all races) SIX FURLONGS
Course Count Opening 1f 5f-4f out 4f-3f out 3f-2f out 2f-1f out 1f out-line Overall
KEMPTON 141 14.07s 11.60s 11.77s 11.62s 11.60s 12.38s 73.04s
LINGFIELD 193 14.10s 11.66s 11.54s 11.19s 11.53s 11.83s 71.85s
SOUTHWELL 51 15.12s 11.95s 12.22s 12.13s 12.28s 13.50s 77.20s
WOLVERHAMPTON 187 14.92s 11.25s 11.92s 12.30s 12.23s 12.70s 75.32s

Winners' median by-furlong times at AW courses, 2005 to 2008 (all races) SEVEN FURLONGS
Course Count Opening 1f 6f-5f out 5f-4f out 4f-3f out 3f-2f out 2f-1f out 1f out-line Overall
KEMPTON 134 14.38s 11.49s 12.05s 12.25s 11.92s 11.77s 12.45s 86.31s
LINGFIELD 245 14.23s 11.58s 12.16s 11.98s 11.42s 11.66s 11.88s 84.91s
SOUTHWELL 47 15.33s 11.92s 12.34s 12.69s 12.39s 12.55s 13.56s 90.78s
WOLVERHAMPTON 224 16.31s 11.79s 11.84s 12.29s 12.52s 12.35s 12.75s 89.85s

Knowledge of the courses in question is useful at this point.

Kempton is largely flat; 5f and 10f races are run around the inner bend, with a run-in
of just under 2f; races over other distances use the outer bend, with a run-in of just
under 3f.
Lingfield undulates more, with its highest point being around the 4f pole and the
biggest decline being from 3f out to 2f out, though the run from the home turn of
around 350 yards is level.
Southwell is mostly flat; it has a 5f straight but other races are on a round course with
a run-in of just under 3f; Southwell is the only course with a testing fibresand surface,
and times there are slower than elsewhere.
Wolverhampton’s straight is the shortest, around 365 yards in length, with a circuit
being a fraction under a mile in length; the back straight is slightly lower than the
home straight, but there are no significant undulations otherwise.
Also very much to the point, with those by-furlong times in mind, is that
Wolverhampton’s distances are 20 yards over at 5f, 4 yards under at 6f, and 32 yards
over at 7f: the excess is included in the times of the opening furlongs. Wolverhampton

18
 AN INTRODUCTION BY TIMEFORM

switched from a polytrack to a tapeta surface in the summer of 2014 but none of the
starts were moved, and the topography of the course remained the same.
Lingfield was remeasured in 2014, after Timeform’s sectional and video analysis
identified anomalies there, and was found to be 6 yards over at 5f, 1 yard over at 6f, 7f
and 8f (and a whopping 51 yards short at 16f ). The new distances at the course reflect
this, but starts used in the past have not been moved.
Those undulations and bends affect the by-furlong times, but it can be seen that in
all instances the opening furlong was the slowest, and that in all but one instance (7f,
Lingfield) the final furlong was the second-slowest.
Perhaps less to be expected, the fastest furlong, or joint-fastest furlong, was the
second one at all distances at Kempton, Southwell and Wolverhampton; at Lingfield,
it is that downhill third-last furlong in each case. It appears that runners, who are
accelerating for much of the opening furlong of a race, “overshoot” somewhat in the
second furlong before settling into a pattern quite close to “even splits”, once bends
and undulations are allowed for.
The profile for longer-distance races tends to be slightly different, a function of the
fact that initial acceleration accounts for a smaller proportion of the race overall and
also because a smaller proportion of races at longer distances tend to be truly-run
(something of which anyone looking to devise their own standard times should be
aware).
Following on from the previous chapter, we are interested in seeing how these
times appear when converted to finishing speed %s: that is, the speed of the finish
compared to the average speed of the race overall.

Winners' median fin spd at AW courses, 2005 to 2008 FIVE FURLONGS

Course Last 5f spd Last 4f spd Last 3f spd Last 2f spd Last 1f spd
KEMPTON 100.0% 103.5% 102.5% 102.3% 101.0%
LINGFIELD 100.0% 104.6% 104.2% 102.4% 100.7%
SOUTHWELL 100.0% 103.0% 100.6% 97.6% 93.2%
WOLVERHAMPTON 100.0% 102.9% 101.1% 100.2% 97.8%

19
 SECTIONAL TIMING

Winners' median fin spd at AW courses, 2005 to 2008 SIX FURLONGS

Course Last 6f spd Last 5f spd Last 4f spd Last 3f spd Last 2f spd Last 1f spd
KEMPTON 100.0% 103.2% 102.8% 102.6% 101.5% 98.3%
LINGFIELD 100.0% 103.7% 103.9% 104.0% 102.5% 101.2%
SOUTHWELL 100.0% 103.6% 102.7% 101.8% 99.8% 95.3%
WOLVERHAMPTON 100.0% 104.3% 102.5% 101.5% 101.0% 99.2%

Winners' median fin spd at AW courses, 2005 to 2008 SEVEN FURLONGS

Course Last 7f spd Last 6f spd Last 5f spd Last 4f spd Last 3f spd Last 2f spd Last 1f spd
KEMPTON 100.0% 102.8% 102.0% 101.9% 102.4% 101.8% 99.0%
LINGFIELD 100.0% 103.0% 102.6% 103.4% 104.1% 103.1% 102.1%
SOUTHWELL 100.0% 103.1% 102.1% 101.3% 101.1% 99.3% 95.6%
WOLVERHAMPTON 100.0% 102.5% 101.8% 100.7% 100.2% 100.1% 98.6%

These confirm that finishing speed %s, compared to average race speeds, are high
from shortly after the start of a race (as the run from that point to the line does not
include the initial period of acceleration) but decrease near the end. The latter effect
is particularly pronounced on the fibresand at Southwell, where winners are usually
running quite a bit slower at the end of a race than on average for the race overall.
This may be down to a number of factors, including the greater likelihood of kickback,
but it does seem as if the unique nature of the fibresand surface lends itself to
attritional finishes, in which the winner is slowing but might have settled the issue
some time before.
Interestingly, this is similar to the profile that tends to be observed on dirt in the US:
late finishing speeds there tend to be well below 100% even after run-up (a small
section at the start of the race ignored for timing purposes) is allowed for.
There is one major caveat to all this, however, for anyone looking to establish the
optimum way to run a race from winners’ times. Some winners run efficiently, and
some do not. It is easier to run slow than fast, and one consequence is that the
majority of races are run at a pace slower than would be required for the horses to run
a good time (though Southwell may be an exception).

20
 AN INTRODUCTION BY TIMEFORM

In order to establish how to run efficiently, from which sectional pars follow, it is
necessary to look at performances in which horses appear to have run about as fast
from the start to the finish of a race as their innate ability allows them to. This means
considering a subset of these winners (and of certain beaten horses), not all of them.
That is covered in the next chapter.

21
 SECTIONAL TIMING

CHAPTER 5
HOW TO RUN FAST: ESTABLISHING SECTIONAL PARS
The outcome of a race depends on many things, some of which
can be measured with sectional times. Running efficiently is not a
prerequisite of winning a race, and neither will running inefficiently
lead to defeat in every instance. That depends on other factors, not
least the other horses in the race. Those other horses may be running
their races even less suitably.
What we can be sure of, however, is that a horse running its race in a significantly non-
efficient way will record an overall time slower than it otherwise would (disregarding
personalised factors, such as an aversion to being in front). That is simple logic: by
definition, inefficiency leads to slower times.
This begs the question “what is an efficient way of running a given race?”

The best way to establish this is through evidence of how horses run fast times, which
they should only be able to do by running efficiently. “Fast times” can be defined as
times which closely reflect the abilities of the horses concerned.
Timeform’s preferred methodology is to identify all performances at a given course
and distance in which a horse has run a time no more than 5 lb below its apparent
ability. Only handicaps – in which differences in abilities are theoretically equalised –
are considered, as horses are quite capable of running misleadingly “good” times in
“bad” sectional ways by simply outclassing their opponents in other types of races.
With a big enough dataset – such as that provided by the 2005 to 2008 TurfTrax
all-weather figures – it is possible to consider only winners. With less information,
estimates may need to be based on close-up beaten horses also. Estimates are
something which overall time analysts use all the time.
An alternative to this evidence-based approach exists and was alluded to previously.
It is possible to ascertain the general effect of bends, inclines and declines on
sectional times to a reasonable level of accuracy and to estimate sectional pars (and
overall times) armed with this knowledge.

22
 AN INTRODUCTION BY TIMEFORM

This “topographical” approach was used by Timeform at York (a level course with
sweeping bends, so one of the easier ones to tackle) when sectional times appeared
there as part of the British Champions Series, despite there being little historical
evidence to fall back upon. It meant that performances like Frankel’s exceptional
Juddmonte International Stakes win at the course could be put into a meaningful
context.
Nonetheless, evidence is preferred, where it exists. In the case of all-weather tracks in
Britain, more than enough evidence exists.
Again, it is necessary to convert raw sectional times into %s, in order that they can be
compared one with another. Any approach dealing with the raw times themselves
must be subject to those raw times being heavily influenced by differences in abilities
and in circumstances.
As an example, the average and median finishing speed %s for six furlongs (which
we now know to be six furlongs plus one yard) at Lingfield for fast-time winning
performances from 2005 to 2008 were as follows:

Par finishing speed %s, Lingfield 6f, 2005 to 2008

Last 6f Last 5f Last 4f Last 3f Last 2f Last 1f
AVERAGE 100.0% 103.4% 103.4% 103.1% 101.4% 99.9%
MEDIAN 100.0% 103.3% 103.1% 102.8% 101.1% 100.2%

If nothing else, this does establish that there is little difference between measuring
this by averages and by medians. Medians (the mid-values of ranges of data) tend to
be more appropriate when dealing with uneven distributions of data, and times tend
to be unevenly distributed. But the process of transforming times to %s which cluster
around 100 renders this distinction small if not negligible.
If we knew nothing else about Lingfield, this would be a perfectly good starting point
for sectional pars, or “how to run fast/efficiently”, at six furlongs there.
But we do know more about Lingfield than just the six-furlong times in isolation. We
know the five-furlong times, and the seven-furlong times, for instance. Indeed, we
know the apparently efficient times for races at all distances at the course. Because
the closing stages of races at one distance are the same closing stages as races at
another distance, they are not independent of one another.

23
 SECTIONAL TIMING

To model sectional pars properly, you need to consider what the information from all
distances at a course tells you about the effects of shared bends, inclines, declines and
the like. Although minor variations will exist, it would be illogical to have, say, horses
speeding up markedly three furlongs out in a mile race but slowing down markedly at
the same juncture for a mile-and-a-quarter race, as measured by sectional pars.
This modelling process is quite complicated, and involves converting those %s back
into times, mapped to standardised times for the course and distance. The reader is
asked to take this process on trust. One consequence of not treating every distance
in total isolation is that the sample size being used effectively increases and the
resulting figures are more robust.
It is beyond the scope of this publication to provide sectional pars for every course
and distance at which Timeform has calculated them, but Lingfield itself provides a
good and manageable example. The closing sectional pars there are as follows:

Sectional pars at Lingfield in terms of finishing speed %s

Distance Last 5f Last 4f Last 3f Last 2f Last 1f
5F 6yd 100.0% 105.0% 103.6% 101.5% 99.3%
6F 1yd 103.3% 103.0% 102.5% 100.9% 100.0%
7F 1yd 102.0% 102.2% 102.7% 101.1% 100.2%
8F 1yd 101.7% 102.5% 102.9% 101.3% 100.2%
10F 102.4% 103.1% 103.9% 103.4% 102.5%
12F 101.8% 103.0% 104.3% 103.6% 102.6%
13F 102.5% 103.1% 104.4% 103.7% 102.7%
15F 169yd 102.0% 103.1% 104.6% 103.1% 101.9%

READ ROWLEYFILE TODAY!

Catch up with the latest from Simon Rowlands online.
Look out for his unique approach to trends analysis in the Big-Race Previews.

Exclusively at
timeform.com Previews Jury Daily Rowleyfile TV Focus The Last Word

24
 AN INTRODUCTION BY TIMEFORM

Sectional pars at Lingfield in terms of by-furlong times

Distance 5f to 4f out 4f to 3f out 3f to 2f out 2f to 1f out 1f out to line Overall
5F 6yd 14.35s 10.75s 10.88s 11.34s 11.84s 59.16s
6F 1yd 11.47s 11.46s 11.31s 11.77s 11.98s 71.88s
7F 1yd 11.99s 12.04s 11.44s 11.89s 12.11s 84.92s
8F 1yd 12.44s 12.10s 11.53s 11.97s 12.24s 98.08s
10F 12.64s 12.50s 12.01s 12.08s 12.29s 125.99s
12F 13.05s 12.78s 12.01s 12.13s 12.37s 152.32s
13F 12.69s 12.79s 12.02s 12.14s 12.38s 165.29s
15F 169yd 13.26s 13.13s 12.04s 12.44s 12.73s 204.60s

The finishing speed % pars will be used in the next section to look at some recent
action at the course and to illustrate how this measure plays out in reality.
First, though, a few words about which sectionals you should use. Where multiple
sectionals exist, you should consider them all, though this may be too lengthy
and complicated a process for some. All sectionals contain information which may
override in importance the sectional information found elsewhere.
Nonetheless, the reality is that many sectionals will be one-off affairs, taken by hand,
and that even where multiple sectionals exist it is impractical in many cases to display
them all at one go. It is, therefore, important to understand which sectionals are likely
to be more meaningful than others.
Manual sectionals are likely to be heavily constrained by camerawork. You need a
juncture of the race which can consistently be gauged through the pictures, and at
which it is possible to estimate the margins back for all or most of the runners.
Over and above this, study of the 2005 to 2008 record shows that the “ideal” sectional
– the one that captures the most information about a race in the smallest total
distance – is one taken around 25% to 35% from the finish. This is the point in races
around which things tend to “happen”. Moves are made, sometimes prematurely,
sometimes too late; and the before-sectional/after-sectional comparison often tells
you plenty about the degree to which horses’ overall times have been compromised
by the run of the race.

25
 SECTIONAL TIMING

As a rule of thumb, a last-two-furlongs sectional tends to be appropriate for races at

up to and including a mile, a last-three-furlongs sectional tends to be more fitting for
races beyond a mile but short of two miles, and a last-four-furlongs sectional may be
best for races at two miles and more.
That last-named guidance applies to all jumps racing, but the reality is that sectional
distances over jumps will be dictated by the location of obstacles, which are the
obvious and easy juncture at which to take times. It is also possible that jumps races
develop in a manner which requires one-off sectionals to be taken at an earlier stage:
without sectionals in the first place, we are guessing.
The constraints of camerawork mean that manual sectional times are usually taken for
the last two furlongs at Lingfield and Wolverhampton and for the last three furlongs
for Kempton and Southwell, regardless of the race’s distance.
Chelmsford City, which recently reopened having been known as Great Leighs in
the 2000s, and Dundalk in Ireland are usually last three furlongs. The sectionals used
for turf Flat tracks vary between two furlongs and four furlongs (and are sometimes
judged by paths and crossings rather than by furlong markers) but are most often
three furlongs.
Fortunately, but not illogically, cameras tend to be positioned where things “happen”
or are about to “happen”.
For simplicity’s sake, much of what follows will consider a horse’s, or a race’s, one-off
sectional. But it bears repeating that all sectionals “count”, not least because races do
not always happily unfold for the convenience of those tackling sectional times!

26
 AN INTRODUCTION BY TIMEFORM

CHAPTER 6
SECTIONAL TIMES IN ACTION: LINGFIELD 2014 AND 2015
Lingfield Park has long been one of the best racecourses for sectional
timing. It was the first all-weather track to open in Britain, in 1989,
and camerawork there has usually been conducive to the taking of
accurate sectionals. More recently, it has been one of the venues for
the All-Weather Championships and was the location of the first All-
Weather Finals Day, on Friday 18 April, 2014.
Some of the All-Weather Championships and all of the All-Weather Finals Day were
covered with electronic sectional timing, courtesy of a TurfTrax project funded by a
grant scheme administered by the British Horseracing Authority and the Department
for Culture, Media and Sport. Those sectional times were available freely to the public
at the time and exist, along with many other historical electronic times, on TurfTrax’s
website.
All-Weather Finals Day in 2014, which played to a sell-out crowd and was widely
considered to be a great success, provides an opportunity to see sectionals in action
and to apply some of the principles mentioned so far. Firstly, we can look at the
headline times for the races, known as the “race sectional” or “field time”, which is
the time for the leader at a sectional to the leader (the winner) at the line. Last two
furlongs is used in each case for ease of comparison.

TIMEFORM.COM
Join today and we will give you a tenner! RACE CARD

FIND TIMEFORM
ONLINE CHELTENHAM
HORSES TO FOLLOW
Register a new account with Timeform and we will give you £10 to spend on downloads. timeform.com
Get the full premium
with full commentaries,
a day, great features

Downloads
service for Britain
Flags and ratings
from Timeform writers

Features
and Ireland including
for £5 each, Race
and much more.
Timeform Race Cards,
Passes from less
Register free!
than £2.50
EXTRA
Shop

You could try Race Passes, grab two Premium Race Cards or the latest tipping features. Listen Live
timeform.com/radio On the Move
timeform.betfair.com

Follow us on Twitter
@Timeform1948 Find us on Facebook
facebook.com/timeform194
8

Get the best advice 2014/15

wherever you are.
and analysis,
JUMPS SEASON CHELTENHAM FESTIVA

Enter voucher code ROW10 when you sign-up

It’s all about winners!
L ISSUE
Flags | Ratings
| Insight | Analysis

HTFX-J1415-Cover1
.indd 1

02/10/2014 11:46

Free credit must be used within 28days. Offer applies online at timeform.com and no
alternative is available. Voucher code may be used once per customer or household. Flags | Ratings | Insight | Analysis
See timeform.com for full conditions and terms of use.

27
 SECTIONAL TIMING

AW Finals Day, Lingfield, 18 April 2014

Race Winner Dist Overall Last 2f Fin spd 2f par Diff Pace
1 VIEWPOINT 12f 148.38s 23.99s 103.1% 103.6% -0.5% true
2 LIVING THE LIFE 7f 82.84s 23.02s 102.8% 101.1% 1.7% true
3 ERTIJAAL 7f 83.19s 24.19s 98.3% 101.1% -2.8% strong
4 LITIGANT 15.77f 197.30s 24.08s 103.9% 103.1% 0.8% true
5 ALBEN STAR 6f 69.73s 23.44s 99.2% 100.9% -1.7% true
6 CAPTAIN CAT 8f 94.51s 24.11s 98.0% 101.3% -3.3% strong
7 GRANDEUR 10f 125.59s 22.40s 112.1% 103.4% 8.7% slow
Last 2f time is time lapsed for leader at that point to leader (i.e. winner) at line
Fin spd % is 100*(Overall time*sectional distance)/(sectional time*overall distance)

Those “race” sectionals for the last two furlongs, converted into finishing speed
%s and compared with the pars for the course and distance (given previously),
immediately highlight whether the races were truly-run, slowly-run, or something
else.
The contests won by Ertijaal, Alben Star and Captain Cat all resulted in “race” finishing
speeds below 100%, and comfortably below the par finishing speed %s for the
respective race distances. It follows that, if the finishing speed of a race is slow
compared to the speed of the race overall, the speed earlier must have been high:
those three races were truly-run or strongly-run.
By contrast, the contest won by Grandeur featured a particularly fast finish, not just
in terms of that 22.40s time, but in terms of that time compared to the overall time. A
finishing speed of 112.1% is well above the sectional par for the course and distance
of 103.4% and indicates a slow early/fast late race.
Some general conclusions can be made on the back of this, such as that the truly-run
and strongly-run races tested stamina more than speed (within the context of the
distances they took place over) and that the slowly-run race tested speed more than
stamina.
It might also have been the case that a position away from the pace will have been an
advantage in the former group of races and one near the pace would have been an

28
 AN INTRODUCTION BY TIMEFORM

advantage in the latter. That is a reasonable rule of thumb, but, as will be seen, things
are not always that simple.
Such general conclusions are useful: quite a lot has been learned from not a lot of
data. Visual race readers may come to similar conclusions but can confirm their views
and, crucially, quantify their views by such methods.
However, we do have access to a lot more data than is being used here. Not only do
we have detailed furlong-by-furlong sectionals, we have those sectionals for each
individual horse.
The “race” sectionals give us an overall context but individual horses run in individual
ways, and that is what counts in terms of energy expenditure where that individual
horse is concerned.

It is worth looking at one of those races in more detail: the Captain Cat race – the
Ladbrokes All-Weather Mile Championship – which looks to have been the most
strongly-run race of all. These are the figures for each individual horse:

Ladbrokes All-Weather Mile Championship, Lingfield, 18 April 2014

Posn Horse Overall Posn 2f out Last 2f Fin Spd Diff
1 CAPTAIN CAT 94.51s 10 23.34s 101.2% -0.1%
2 HIGHLAND KNIGHT 94.71s 5 23.93s 98.9% -2.4%
3 ALFRED HUTCHINSON 94.89s 6 23.98s 98.9% -2.4%
4 GREY MIRAGE 95.02s 7 24.06s 98.7% -2.6%
5 GEORGE GURU 95.13s 4 24.45s 97.3% -4.0%
6 NOBLE CITIZEN 95.14s 12 23.76s 100.1% -1.2%
7 SIRIUS PROSPECT 95.14s 7 24.17s 98.4% -2.9%
8 CHOOKIE ROYALE 95.35s 2 24.89s 95.8% -5.5%
9 ANACONDA 95.37s 1 24.97s 95.5% -5.8%
10 FRONTIER FIGHTER 95.60s 2 25.14s 95.1% -6.2%
11 MOONDAY SUN 95.66s 10 24.49s 97.6% -3.7%
12 SILVERHEELS 97.49s 9 26.42s 92.2% -9.1%

29
 SECTIONAL TIMING

It can be seen that Captain Cat and Noble Citizen ran the last two furlongs at a slightly
faster speed than their average race speed but that every horse in the race finished
more slowly than the established sectional par for a mile at Lingfield of 101.3%.
Some finished more slowly than others, and, interestingly, the winner was nearest
to par of all. The horses which occupied the first three positions at the two-furlong
marker – Anaconda, Chookie Royale and Frontier Fighter – lost several positions from
that point and recorded notably slow finishing speed %s of below 96.
The difference between a horse’s actual finishing speed % and the optimum finishing
speed % defines the degree to which its overall time was affected by its energy
distribution along the way.
A more detailed explanation of this crucial consideration in sectional-timing analysis
comes in a later chapter.
For now, it is sufficient to know that Captain Cat came closer to par in winning than
Anaconda and others did in losing, at least as judged by final-two-furlong times.
That is not to say that Anaconda was a better horse than Captain Cat – who beat
him by some margin and may have had more to give – just that the former’s (slower)
overall time was more compromised by the run of the race than the latter’s. The
“race” sectionals show that the pace was strong, and the horse which set that pace,
Anaconda, suffered as a result.
Incidentally, Captain Cat’s own times for the last five furlongs, four furlongs, three
furlongs, two furlongs and one furlong equate to finishing speeds of 100.3%, 101.1%,
101.8%, 101.2% and 99.4% respectively. He raced quite close to par throughout, but
especially so for those last two furlongs.
Captain Cat was probably the best horse – as he showed at times before and after –
but it helped that he got a well-judged ride from James Doyle. Sectionals leave no
room for doubt about that.

30
 AN INTRODUCTION BY TIMEFORM

CHAPTER 7
SECTIONAL TIMING IN ACTION, ELSEWHERE
Lingfield Park is currently at the forefront of sectional timing in
Britain, though there is disappointingly little competition for this
accolade. Nonetheless, other courses have provided some sectional
times over the years, while it has been possible to take sectional times
by hand with acceptable accuracy at many other tracks, as well as at
Lingfield when electronic ones are not provided.
The provision of electronic sectional times for many years at Newmarket means that
the aforementioned 2002 Two Thousand Guineas there – the race in which Rock of
Gibraltar held off Hawk Wing – can be put into a fuller context, as can other races.

Two Thousand Guineas, Newmarket, 04 May 2002 - Finishing Speed %s

Horse Last 5f Last 4f Last 3f Last 2f Last 1f
ROCK OF GIBRALTAR 104.3% 104.0% 102.9% 101.3% 96.8%
HAWK WING 105.0% 105.1% 104.4% 103.3% 100.0%
Par Finishing Speed 101.8% 101.0% 99.7% 98.3% 93.4%

It can be seen that both Rock of Gibraltar and Hawk Wing were finishing faster than
would have been the case in a truly-run race, though Rock of Gibraltar was closer to
those pars than his rival. It seems the damage was done early on, a steady opening
meaning that Hawk Wing, in particular, was out of his ground.
While Hawk Wing ran the last furlong at precisely the same average speed as for the
race overall, this needs to be compared to what would be efficient for that course and
distance, mindful of the fact that the Rowley Mile course rises from the three-furlong
marker, much of that incline coming in the final furlong.
We have other Two Thousand Guineas, from around the same time and more recently,
with which to compare those times and finishing speed %s, but sadly nothing from
the greatest Guineas performance of the modern era, Frankel’s in 2011, for which no
electronic sectionals were provided.

31
 SECTIONAL TIMING

Selected Winning Two Thousand Guineas Performances

Year Horse Going Time/Form Last 3f Last 2f Last 1f Overall
36.07s 24.57s 13.03s
2001 GOLAN Good 122T, 122F 97.48s
(101.3%) (99.2%) (93.5%)
36.96s 24.83s 12.90s
2004 HAAFHD Good 129T, 129F 96.64s
(98.1%) (97.3%) (93.6%)
37.21s 25.20s 13.09s
2012 CAMELOT Soft 108T, 123F 102.46s
(103.3%) (101.6%) (97.8%)
36.55s 24.99s 13.21s
2013 DAWN APPROACH G/F 109T, 130F 95.84s
(98.3%) (95.9%) (90.7%)
Par (99.7%) Par (98.3%) Par (93.4%)

Hopefully the reader will see that Camelot’s finishing speeds (but not the absolute
times, which were slowed by softer going) are fast compared to the pars, while Dawn
Approach’s are slow; Golan’s and Haafhd’s differ from par, but by less.
The inference to be drawn is that Camelot’s fast finish in relative terms indicates a
steady pace earlier, and Dawn Approach’s slow finish indicates a fast pace earlier.
You would expect both Camelot and Dawn Approach to be capable of better times
had they raced more efficiently, whereas Golan and Haafhd should have recorded
times more closely reflecting their true abilities.
The Timeform form-based and overall-time-based assessments back this up: Golan’s
and Haafhd’s overall-time-based figures are good compared to their form assessment,
while Camelot’s and Dawn Approach’s are not, but for differing reasons (one having
gone steadily early, the other having gone strongly).
Over and above considerations of merit, this arguably tells us something about the
aptitudes of the horses involved also. Camelot had running left in him at the end of a
muddling race, but Dawn Approach was emptying, if not as quickly as his rivals.
Interestingly, both colts contested The Derby over half a mile further at short odds on
their next starts, with Camelot winning and Dawn Approach finishing last. There is a
danger of hindsight bias here, but were some of the portents there for all to see in the
Guineas sectionals?!

32
 AN INTRODUCTION BY TIMEFORM

The Derby itself provides an illustration of how much an extreme topography can
influence par finishing speed %s. Epsom is renowned for its mostly downhill finish, as
well as its bends and tricky cambers, but at least as significant a feature is the opening
five furlongs of the Derby course, which rises around 125 feet according to Google
Earth, making it the toughest incline in British racing, Flat or jumps.
Hand times from the end of the path just after the start of the home straight can be
taken and Google Earth shows that this sectional distance is just less than three and
a half furlongs. The par finishing speed % for mile-and-a-half races at Epsom is about
111.0%. The fastest Derby-winning sectional Timeform has recorded is the 38.0s
of Galileo on firm going in 2001, or 115.1% finishing speed. Those figures are a lot
different to those from the flatter courses considered previously.
By contrast, the stiffest last-three furlongs on the Flat in Britain and Ireland – again
according to Google Earth – are to be found at Beverley, which rises over 60 feet, and
where par sectionals for longer-distanced races are in the region of 97%.
Most sectional pars for the last two or three furlongs on British Flat tracks tend
towards 100% or slightly over. The same tends to apply for closing sectionals for
British jumps tracks, though sectional distances vary considerably according to
the location of obstacles. This is what we should expect given that even pacing
approximates to efficiency (and that most courses do not resemble the roller-coaster
ride of The Derby at Epsom).
It is worth seeing how some other notable sectional performances from recent years
have shaped up.
Ascot Racecourse has been active in sectional timing due to the regular existence
of TurfTrax figures there as part of the British Champions Series. The unique nature
of big events at the track (including large crowds, large infrastructure and wireless
interference) has posed some unique problems, however: the times returned have
not always seemed valid.
Nonetheless, a combination of electronic and manual times at the course since it was
relaid in 2005 has made it possible to establish solid sectional pars there.
It was most disappointing that the course executive chose to drop electronic
sectionals for Royal Ascot 2014 at very short notice and make no provision for
alternatives. Fortunately, the camerawork (including uninterrupted coverage from the

33
 SECTIONAL TIMING

inside of the course) on this flagship occasion made the taking of accurate manual
sectionals relatively easy for the majority of races.
As a result, it was possible to capture one of the great sectional efforts of the modern
era: a performance which was visually stunning but which would otherwise have left
no permanent sectional record.

Major races on Day One of Royal Ascot 2014

Winner Distance Overall Race sect (Fin Spd) Win sect (Fin Spd)
TORONADO 8f (str) 97.73s 25.2s (97.0%) 24.95s (97.9%)
THE WOW SIGNAL 6f 72.99s 24.2s (100.5%) 24.15s (100.7%)
SOLE POWER 5f 58.85s 23.5s (100.2%) 22.7s (103.7%)
KINGMAN 8f (rnd) 99.06s 22.6s (109.6%) 22.35s (110.8%)

Bated Breath had recorded a last-two furlongs time of 22.87s (TurfTrax) when second
in the 2012 King’s Stand Stakes, but that was comfortably bettered by Sole Power
in victory in the same race on this day. As can be seen, the “race” finishing speed of
100.2% was not extraordinary (the par for five furlongs at Ascot is 100.5%), but Sole
Power himself produced a prodigious burst of speed to come from behind and post
103.7%.
If that was prodigious, then it is difficult to find a word to describe the finishing
effort of Kingman, who ran about 0.35s quicker still (than a Group 1-winning sprinter
showing a rare turn of foot) for the final two furlongs of a three-furlong-longer race
when winning the St James’s Palace Stakes 40 minutes later.
All such times need to be put into context, of course. The context here was a quick
surface (quicker than the “good” going returned officially), a following wind and, in
Kingman’s case, a race run at a steady pace: the “race” sectional for the St James’s
Palace was 109.6% compared to a course-and-distance par of 99.4%.
That all said, Kingman’s time for the last two furlongs (and his 34.2s for the last three
furlongs) was still extraordinary. He was running faster at the end of a mile race
than top-class sprinters had managed at shorter distances just before. It is likely
that no other horse has run the final quarter-mile at the new Ascot track faster: with
widespread sectionals we would know for sure.

34
 AN INTRODUCTION BY TIMEFORM

Kingman’s overall time suffered for his running slow early/fast late, of course. While
his Timeform performance rating – an assessment of the result largely independent of
time – was 124, his timefigure for the race overall was a modest 105. The gap between
those two figures is actually not as great as might be expected in the circumstances.
How much faster Kingman might have run if paced evenly is something that will be
dealt with later on. For the time being, it is tempting to bask in the memory of such
scintillating speed: 22.35s for two furlongs is over 40 mph, and this took place on a
track with a slightly uphill finish!
Such speeds do not occur in jumps racing, but sectional analysis is still relevant in
that sphere, with one or two caveats. After all, racehorses are subject to the laws of
physics, and to the effects of running efficiently or inefficiently, whether they jump
obstacles or not.
The main problem in jumps racing is with accurate measurement. Google Earth
means it is possible to measure sectional distances with accuracy (though still with
some difficulty where transportable obstacles are concerned) and has highlighted
some significant errors in the official distances of races themselves.
Race distances over jumps in Britain are approximated to the nearest half a furlong –
don’t ask why, that’s just the way it is – but some of them have been manifestly wrong
even so. A Timeform survey of the Wetherby 25f chase course late in 2014 identified
it as being around one and a half furlongs short of the distance advertised (the BHA
eventually conceded that the distance was wrong and adjusted accordingly).
In addition, overall race times suffer from a lack of agreement as to when a race starts
in a meaningful sense. Timeform has long adopted the approach of timing races
from when the leader breaks the plane of the start, rather than from when the tape
goes up (which can add many meaningless seconds to the time). Some bodies have
belatedly followed suit, but others seemingly have not.
This all makes the establishment of average race speeds (not to mention timefigures
for the overall race) fraught with difficulties. Until the authorities take measurement
of time and distance more seriously, sectionals will be compromised without a good
deal of independent manual validation.

35
 SECTIONAL TIMING

In other respects, however, sectionals over jumps have a lot of worth. The times
themselves can be taken easily when a leader jumps an obstacle, and, even without a
wider context, raw sectionals on the same day can make for meaningful comparisons.
The more attritional nature of much of jumps racing means there can be some wild
swings in finishing speed %s, pointing to the need to interpret results accordingly.
An example of jumps’ finishing speeds can be given with the Cheltenham Gold Cup of
recent years.
Cheltenham Gold Cup, 26.5f (sectional from 3 out = 3.84f)
Year Winner Overall Race Sect (Fin Spd) Win Sect (Fin Spd)
2008 DENMAN 407.8s 63.2s (93.5%) 63.2s (93.5%)
2009 KAUTO STAR 404.9s 58.3s (100.6%) 58.2s (100.8%)
2010 IMPERIAL COMMANDER 403.8s 59.3s (98.7%) 59.1s (99.0%)
2011 LONG RUN 389.5s 56.8s (99.4%) 56.5s (99.9%)
2012 SYNCHRONISED 396.2s 57.0s (100.7%) 56.1s (102.3%)
2013 BOBS WORTH 424.9s 65.30s (94.3%) 64.36s (95.7%)
2014 LORD WINDERMERE 404.6s 58.35s (100.5%) 57.25s (102.4%)

The par finishing speed for chases over 26.5f on the New Course at Cheltenham
appears to be around 99.0%. It can be seen that the 2008 and 2013 Gold Cups were a
long way below that, indicating a slow finish and the high likelihood of a strong pace
prior to that.
In addition, Bobs Worth’s Gold Cup was run on softer going than the other races
listed. The combined effect of soft ground and a strong pace suited a horse with
abundant stamina, which Bobs Worth demonstrated by rallying to lead approaching
the last as others wilted.
He was nowhere near as effective on good ground and with a much less searching
pace 12 months later, a race in which he finished fifth at short odds to Lord
Windermere.
In the three months from November 2014 to January 2015 inclusive, jumps finishes
ranged from 80.8% (Bob Ford in a controversial long-distance chase at Ffos Las on all-

36
 AN INTRODUCTION BY TIMEFORM

but unraceable ground) to 125.9% (Sweet Boy Vic in a farcically slowly-run hurdle at
Fontwell).
Such wild extremes are obvious to those using eyes as well as to those using
stopwatches, but many of the scenarios in between are less cut and dried, and
putting hard figures on them should help us to understand precisely what happened
and why.
It must be clear from some of the preceding text that an ability to take serviceable
sectionals by hand will continue to be a useful skill to possess unless British racing
suddenly (and unexpectedly) funds widespread and accurate electronic sectionals.

37
 SECTIONAL TIMING

CHAPTER 8
TAKING YOUR OWN SECTIONALS BY HAND
Compared to having electronic sectionals delivered to your inbox, or
available to you through the internet, recording sectionals by hand
may seem both laborious and of questionable worth.
However, the vast majority of races in Britain are not covered by electronic sectionals,
while those that are arguably need to be validated independently, as they are in some
other jurisdictions. What’s more, with some basic guidelines and practice, manual
sectionals can be both acceptably accurate and achievable without too much effort.
Timeform augments what sectional coverage exists with manual coverage at all other
meetings in Britain and, more recently, in Ireland.
At a basic level, it is easy to record a “race” sectional for the leader at a pre-defined
sectional and the leader (that is, the winner: this will often enough not be the same
horse) at the line. Simply set a stopwatch running as the nose of the leader passes the
sectional and stop it when the nose of the winner crosses the line.
On-screen clocks provided by some broadcasters make this even easier to do, and
there is also the option of capturing a video and using film-editing technology to the
same ends. Avoid taking times from the start of a race to the sectional, as response
delay means the starting time is likely to be out by as much as 0.5s.
A race sectional can be converted into a race finishing speed by the 100*(T*d)/(D*t)
equation detailed earlier in this guide. It is possible to learn much about a race in
general from this information.
But individual horses run races in individual ways, and it is these individual ways – not
the race itself – which define whether the horse in question has run efficiently or not.
It is entirely possible to run efficiently in an otherwise inefficient race, and vice versa.
In order to use sectionals more fully, we need times for each and every horse in a race.
Contrary to what might be imagined, this does not mean repeating the race-sectional
process, at length, for every horse. That race sectional, plus the race result, gives you
quite a lot of information: all that is then needed to calculate individual sectionals
is the margin back at the sectional, which is something that is usually best gauged
visually.

38
 AN INTRODUCTION BY TIMEFORM

In Britain, horses’ overall times can be engineered from the race result, for the margins
returned are conversions of the time margins between them (though the Racecourse
Judge responsible reserves the right to tweak these). These conversions are fixed
according to circumstance, even though the reality is that horses may well be
travelling at speeds markedly different to those implied by the methodology.

Official Lengths-Per-Second Scale in use in Britain, February 2015

Race Type Going Range LPS
Flat Turf Good or quicker 6
Flat Turf Good (Good to Soft Places) to Good to Soft (Soft Places) 5.5
Flat Turf Soft (Good to Soft Places) or slower 5

Flat AW Kempton, Lingfield, Wolverhampton, Chelmsford City 6

Flat AW Southwell 5

Jumps Turf Good or quicker 5

Jumps Turf Good (Good to Soft Places) to Good to Soft (Soft Places) 4.5
Jumps Turf Soft (Good to Soft Places) or slower 4

Jumps AW Kempton, Lingfield 5

Jumps AW Southwell 4

Going changes during the day may alter the scale to be applied
When going varies, the scale for the going on the straight is applied

So it is that we know that a horse given as having been beaten three lengths on all-
weather at Kempton has been beaten 0.50s (give or take some rounding); one given
as having been beaten four lengths on soft going on Flat turf has been beaten 0.80s
(ditto); and so on.
Sectionals themselves show that this is a very crude conversion. We saw much earlier
that the median final-furlong times for winners of five-furlong races at Lingfield from
2005 to 2008 was 11.64s; in the same period, the equivalent figure for extended two-
mile races at Wolverhampton was 13.17s. A time margin of a second was treated as six

39
 SECTIONAL TIMING

lengths in both cases even though horses at the shorter distance were covering 13%
more ground in that second than those at the longer distance.
Sectionals show that Lingfield finishes are quicker than those at Kempton and
Wolverhampton, and that Chelmsford City and Southwell are not as much out of line
as the conversion implies.
Why the racing authorities cannot simply ensure that the overall time of each horse is
published – as it is in Ireland and in many other jurisdictions – is a mystery, given they
have this information and that it has been requested repeatedly over many years by
members of the public.
Be that as it may, the published methodology means conversions can be made
according to that methodology so that race times for individual horses result.
Conversions at the juncture of the sectional can – and should – take into account
the speed at which the horses appear to be travelling, with the race speed from that
sectional to the line a reasonable proxy for this.
Timeform converts its own visual margins at sectional junctures taking 12.50s/
furlong to be equivalent to 6 lengths per second, based on there being 75 lengths in a
furlong, though the “true” figure may be slightly more or slightly less. This means that
a horse running at 10.71s/furlong is covering 7 LPS ((12.50/10.71)*6) and one running
at 14.58s/furlong is covering 5 LPS ((12.50/14.58)*6).

DOWNLOAD THE
NEW APP TODAY!
Find winners on the move

40
 AN INTRODUCTION BY TIMEFORM

The upshot is a scenario like the following:

Manual sectional calculations for Race 2, Wolverhampton, 09 January 2015

Posn Horse Cum Marg As time 1 Indiv Time Marg Sect As time 2 Sect time Fin spd
1 RICH AGAIN 0 lengths 0.00s 73.36s 5.2 lengths 0.81s 22.63s 108.4%
2 INVINCIBLE RIDGE 1.5 lengths 0.25s 73.61s 2.7 lengths 0.42s 23.26s 105.8%
3 ACE MASTER 2 lengths 0.33s 73.69s 1 length 0.16s 23.61s 104.4%
4 STRATEGIC FORCE 2.75 lengths 0.46s 73.82s 3.9 lengths 0.61s 23.29s 106.0%
5 LIGHT FROM MARS 4 lengths 0.67s 74.03s 0 lengths 0.00s 24.11s 102.7%
6 CLUBLAND 4.5 lengths 0.75s 74.11s 1.9 lengths 0.30s 23.89s 103.8%
7 DREAM SCENARIO 5.5 lengths 0.92s 74.28s 3.7 lengths 0.58s 23.78s 104.5%
8 RASAMAN 7.5 lengths 1.25s 74.61s 5.6 lengths 0.87s 23.82s 104.8%

Winner’s time: 73.36s Race sectional: 23.44s Race distance: 5.98f Sectional distance: 2.00f
Sectional Lengths Per Second: (((12.50*2)/23.44)*6) = 6.40

The individual sectional times are arrived at by subtracting “As time 2” (the time
margin behind the leader at the sectional) from “As time 1” (the time margin behind
the winner at the finish) and adding the result to the race sectional.
In the case of the winner, who came from behind, this individual sectional is quicker
than the race sectional. In the case of Light From Mars, who went from leading at the
sectional to finishing fifth, this individual sectional is slower than the race sectional.
The resulting finishing speed %s are fast for the course and distance (at which the par
is 101.1%), showing that the winner did well in sectional terms to come from several
lengths back despite a generally fast finish.
The calculations are straightforward for anyone with more than a passing knowledge
of Microsoft Excel and of basic arithmetic, though the populating of such a
spreadsheet with results in the first place is more of a challenge.
The same processes can be followed for jumps races, though the nature of racing in
that sphere means that precision is not quite so paramount. Margins back to non-
leading horses at a sectional are probably best assumed to be at 0.2s per length for
the first several lengths and then timed for longer margins.

41
 SECTIONAL TIMING

There will be, inevitably, occasions when horses are out of picture at important stages,
and other occasions when information is incomplete or questionable.
We work with what we have in imperfect circumstances. That applies to racing more
widely, and not just sectionals. And that applies to how we analyse the sectionals that
we manage to obtain, also.

The late and renowned British mathematician and statistician George Box put it very
neatly: “All models are wrong but some are useful.
The more advanced sectional analysis that is to follow is a “model” of the real world
arrived at through interpretation of data. The true test of such a model is utility, not
unerring accuracy: specifically, does our “model” allow us to understand the subject
matter better and to make better decisions as a result?

42
 AN INTRODUCTION BY TIMEFORM

CHAPTER 9
UPGRADING SECTIONAL PERFORMANCES
It was established earlier how to create sectional pars – expressed as
“efficient” finishing speed %s – against which performances could be
compared. It follows that differences from these finishing speed %s
represent inefficiencies to varying degrees, and that the presence of
inefficiency implies a performance which could have been improved
upon, in time terms.
“How to run fast” was a relatively simple thing to establish: “what effect does running
inefficiently have on a horse’s performance?” is a good deal more complex, but must
be answered in order to upgrade efforts on the back of sectionals rather than simply
observe that a horse “went too fast” or “went too slow”, as has been the case to this
point.
The theory is that the negative effect on a horse’s overall time is a function of the
difference between the Optimum finishing speed % for the circumstances (“O”) and
the horse’s Actual finishing speed % (“A”).
This difference needs to be squared: the underlying laws are related to velocity, and
the effect of going too fast or too slow should be multiplicative rather than additive
as actual finishing speed % moves further away from the optimum.
A difference of 1% from optimum should have little effect on overall time (1
multiplied by 1 = 1), but a difference of 5% from optimum should have a much bigger
effect (5 multiplied by 5 = 25).
In addition, what precisely is being measured to come up with Optimum and
Actual finishing speed %s should have some bearing: in particular, the length of the
sectional distance compared to the length of the race itself should affect the degree
to which inefficiencies are upgraded.
All of this was incorporated into a speculative first equation – a heuristic in technical
jargon – involving O (Optimum finishing speed %), A (Actual finishing speed %),
d (sectional distance) and D (overall race distance), which was investigated by the

43
 SECTIONAL TIMING

academic and acclaimed author Bob Wilkins and found to be fit for purpose, subject
to a couple of small revisions.
Wilkins’ mini-paper can be found in the Appendix to this booklet. Incidentally, his
2010 book “Bioenergetics and Racehorse Ratings” deals with efficient pacing, among
other things, and is highly recommended reading.
The paper has not been peer-reviewed, and anyone with the inclination to critique
it, or the modified equation which results from it, is very welcome to do so. That is
part of the scientific process whereby we edge towards a closer approximation of the
truth.
The revised equation in use at Timeform is 1.5*(O-A)*(O-A)*((d/D)^1.4), where “O” is
the Optimum finishing speed % for the circumstances, “A” is the Actual finishing speed
% achieved by the individual horse, “d” is the sectional distance, “D” is the overall race
distance (and ^1.4 means d/D is raised to a power of 1.4, according to Microsoft Excel
conventions).
The original multiplicative constant of 1.25 was increased to 1.5 following further
research. Further tweaks as more data, and more experience of dealing with that data,
becomes available are entirely possible.
As Wilkins confirms, the effect of running inefficiently on overall time (relative to
the time that could be expected) should describe a parabola. That is borne out by
applying the Timeform equation to the winners of all-weather races in Britain in 2014:
the further away from sectional par a performance was, the more the horse’s overall
time was affected, and in an approximately parabolic fashion, too.
One important observation of Wilkins’ work is that “the correction required for slower
finishes is higher than for faster finishes”. There is a limit to how fast a horse can run,
and running efficiently comes quite close to that: it is more closely bounded on one
side of the parabola than the other. This is also taken into account when applying the
equation.
The figure that results is an upgrade to the horse’s overall time in terms of pounds. If a
horse has a timefigure (based on its overall rather than sectional time) of, say, 100 and
a sectional upgrade from the above equation of, say, 6, its sectional rating would be
106.

44
 AN INTRODUCTION BY TIMEFORM

It is beyond the scope of this booklet to explain how timefigures (comparative

assessments of overall times) are arrived at, but Timeform produces them for all Flat
racing in Britain and much of Ireland now. Such timefigures have considerable worth,
but overall times are always subject to the pace at which a race was run. Timefigures
plus sectional upgrades have greater worth still.
This was confirmed by a survey of all performances on all-weather in Britain in 2014
in which a Timeform sectional rating and a rateable next-time performance rating
existed. Across these 12,739 instances, sectional ratings were closer to predicting
next-time performance than were timefigures on 9,586 occasions (75.2%) and were
an average of 2.6 lb closer overall.
In addition to prompting sectional ratings, the upgrade method may be used to
re-rate a race according to the margins-beaten result that would have happened
(in theory) had each of the runners in that race performed efficiently. Non-specific
general visual impressions about a race are hereby replaced with specific and
formalised quantitative assessments of it.
Theory needs to be tested against reality, to see if it holds up. The evidence, from
Wilkins and from Timeform’s own research, is that the upgrade equation works well
for normal values of d/D – especially in the 25% to 35% range identified much earlier
as the best one-off sectional – and for all but extreme finishing speeds.
George Box (mentioned in the previous chapter) might approve of such utility.

45
 SECTIONAL TIMING

CHAPTER 10
ILLUSTRATING SECTIONAL UPGRADES
The sectional-upgrade process described in the previous chapter can
be applied retrospectively to some of the historical races that have
already been considered, such as the 2002 Two Thousand Guineas:
Two Thousand Guineas, Newmarket, 04 May 2002 - finishing speeds (and upgrades)
Horse Overall Last 5f Last 4f Last 3f Last 2f Last 1f
ROCK OF GIBRALTAR 100.0% (+0) 104.3% (+5) 104.0% (+5) 102.9% (+4) 101.3% (+2) 96.8% (+1)
HAWK WING 100.0% (+0) 105.0% (+7) 105.1% (+10) 104.4% (+9) 103.3% (+5) 100.0% (+4)
PAR FINISHING SPEEDS 100.0% (+0) 101.8% (+0) 101.0% (+0) 99.7% (+0) 98.3% (+0) 93.4% (+0)

Upgrades result from difference between actual finishing speed % and par finishing speed %,
according to the equation 1.5*(O-A)*(O-A)*((d/D)^1.4) explained previously

By definition, an overall time is not upgraded if the only “sectional” we have is for
the entirety of the race itself (finishing speed = 100.0%); also, by definition, finishing
speed %s which are bang on par are not upgraded as they define the efficient way of
achieving a given overall time.
But the differences between these par finishing speed %s and the actual finishing
speed %s result in various upgrades, according to which sectional is used. The correct
thing to do is to take the maximum value shown for a horse (summing the values
would result in multiple counting of the same phenomenon), which, in this instance,
is +5 for Rock of Gibraltar and +10 for Hawk Wing.
The conclusions would have been similar in general terms, but not so favourable for
either horse, had only last-three-furlong or last-two-furlong times been known. The
existence of electronic sectionals for each furlong back in 2002 was a clear positive.
Rock of Gibraltar’s base timefigure of 112 gets upgraded to 117, while Hawk Wing’s
base timefigure of 112 to gets upgraded to 122. The implication is that the latter was
about 5 lb superior to the former on the day, as judged by the sectional-upgrading
process, which is about two lengths for a mile run in a time of 96.50s.

46
 AN INTRODUCTION BY TIMEFORM

Recalculating the result on the back of sectionals is likely to give a better guide to the
respective merits of the protagonists than the bare facts of the result itself, while the
sectional ratings which have been prompted are likely to be a better guide than any
assessment based on overall times alone.
The same process was applied to the performances of one of 2014’s stars, Kingman,
including in the St James’s Palace Stakes referred to previously, though in that case
only last-two furlong times, taken by hand, exist.
St James's Palace Stakes, Royal Ascot, 17 June 2014
Horse Overall Last 2f Fin Spd Upgrade
KINGMAN 99.06s 22.35s 110.8% 24
NIGHT OF THUNDER 99.44s 23.0s 108.1% 14
OUTSTRIP 99.60s 22.95s 108.5% 15
WAR COMMAND 99.69s 22.65s 110.0% 21
YUFTEN 99.85s 23.35s 106.9% 10
TOORMORE 100.06s 23.45s 106.7% 9
PRINCE OF ALL 100.07s 23.25s 107.6% 12

The slow early pace meant that every horse in the field ran inefficiently and came
home a good deal faster than the course and distance par of 99.4%. But some did
better than others, and Kingman did best of all.
His 24 lb upgrade from a timefigure of 105, gave him a high-class sectional rating of
129. A reworking of the race result, based on this sectional information, meant an
even higher form-based assessment was justifiable.
Kingman’s sole defeat came in a more truly-run Two Thousand Guineas. In each of
his other races in his classic season, he showed a terrific turn of foot to win tactical
contests. The bare form of those contests meant he could not be rated as highly as
he seemingly deserved by those constrained by the final result rather than by a fuller
picture of how that result was arrived at.
Incidentally, the same process of sectional upgrading pointed to Sole Power being
about 4 lb better than the result in the King’s Stand Stakes on the same card. Sole

47
 SECTIONAL TIMING

Power’s last-to-first turn of foot was visually stunning, but sectionals did show that
many of his rivals had done plenty of running early on.
Potential illustrations of the sectional upgrading process are legion: it is hoped the
reader gets the picture. It is, however, worth considering an arguably counter-intuitive
example at a much lesser level before moving on.

Ladbrokes Handicap, Southwell, 08 January 2015

Horse Overall Last 3f Fin Spd Upgrade
WARFARE 103.08s 39.4s 98.0% 1
PUTIN 103.83s 40.35s 96.5% 5
DHA CHARA 103.98s 40.45s 97.6% 5
TWO SHADES OF GREY 104.02s 39.95s 97.3% 2
SOOQAAN 104.42s 40.25s 96.4% 3
I'M SUPER TOO 105.32s 40.95s 96.8% 5

It is not just by running slow early and fast late that a horse earns an upgrade: running
fast early and slow late leads to the same. Indeed, while the former happens more
often, theory and evidence shows that the latter is proportionally more harmful to a
horse’s overall time.
However, it is possible to run fast early and slow late and still win: that is what
happened in the above example. Each of the first six horses shown finished slower
than the course-and-distance par of 99.9%, but the winner was closer to that par than
the horses it beat.
The upshot is that the magnitude of the winner’s success seems to have been
exaggerated by the way the race was run, despite that winner’s having “gone (slightly)
too fast”. This happens more than occasionally at Southwell, easily the most stamina-
sapping of the British all-weather tracks, but it can happen anywhere.
Cape Blanco gave an even more vivid demonstration of this sort of phenomenon
when winning the Irish Champion Stakes at Leopardstown in 2010. The top-class colt
set off like a scalded cat and was soon several lengths clear. Pretty soon, all of his rivals
were busting a gut in giving chase, including his odds-on stablemate Rip Van Winkle.

48
 AN INTRODUCTION BY TIMEFORM

Far from being overhauled, Cape Blanco remained in front by a comfortable margin,
as, one by one, his rivals tried to bridge the gap but failed. And, yet, Cape Blanco did
weaken – sectionals (including a 90.4% finishing speed for the last two furlongs)
clearly showed that – it was just that his rivals weakened also. By comparison, the
legendary Sea The Stars had recorded a finishing speed of 100.1% in the same race 12
months earlier.
There are many different ways in which to win a race, and they include ensuring your
rivals run inefficiently through running inefficiently yourself. It bears repeating that
the object of the exercise is not to run the fastest possible time, but to beat one’s
rivals.

It is best to view each individual horse as an entity of its own. The overall time each
horse achieves will result from its ability and aptitude, and from how its energy is
rationed. But the outcome of a race depends on the interplay of all of those factors for
each and every horse.

49
 SECTIONAL TIMING

CHAPTER 11
FAQS ABOUT SECTIONALS
Sectional timing is concerned with variations in pace, so, in this
booklet about sectional timing, we will steady the pace slightly after
some fairly frenetic chapters. We have straightened for home, the
winning post is in sight, and we will gather ourselves for the closing
effort.
A chapter covering “Frequently Asked Questions” – or a thinly-veiled means for
addressing some misconceptions on the subject of sectional timing – could have
gone just about anywhere, but finds itself here. Much of what follows will reprise
some of what went before, but some of what follows is new.
“Isn’t comparing sectionals on different tracks pointless?”
In much the same way that comparing overall times on different tracks is not
pointless, once you have standard times, sectionals – which deal with the same issues
of time and distance – can be adjusted to take into account different circumstances.
A “par” sectional will be specific to a course and distance, but the conversion of
sectional figures to finishing speed %s allows much readier comparisons between
different circumstances, different continents and even different generations.
“Isn’t comparing sectionals on different going, such as when going is
deteriorating, pointless?”
As with the first question, the procedures outlined previously allow for ready
comparisons when the racing surface varies. The important thing to remember is
that a before/after sectional compares one part of a race with the remainder of that
race, or one part of a horse’s performance within that race with that same horse’s
performance in the remainder of that race. By converting sectional times into
finishing speed %s, we are dealing with ratios, and ratios are minimally affected by
circumstances, for those circumstances are heavily implied in the overall and sectional
times themselves. So, comparisons between races which took place on different
goings are perfectly valid by this method.

50
 AN INTRODUCTION BY TIMEFORM

“Aren’t sectionals compromised by inaccuracies in race distances?”

They are, but much less so than would be the case for any analysis of overall race
time. Other than in hurdles races, closing sectionals nearly always remain constant
in length (whatever that length happens to be). Failing everything else, the raw time
taken by a horse in that sectional will tell you something compared to other raw
sectional times on that day and on other days.
“Doesn’t the tactical nature of racing in Britain and Ireland render sectionals
of little worth compared to, say, the US, where races tend to be more strongly-
run?”
Quite the opposite. Indeed, the asking of such a question (and it happens, frequently)
betrays a fundamental misunderstanding of the subject. Sectionals are at their
strongest when pace varies and overall race time tells you less than it might about
the respective abilities of the horses. Best of all, perhaps, is a mixture of well-run races
(from which sectional pars, or “How To Run Fast”, may be established) and tactical
races. That neatly describes the state of affairs in Britain and Ireland.
“Don’t the individual requirements of individual horses render a “one-size-fits-
all” approach worthless?”
A more sensible question, for individual horses must have individual ways of running
efficiently, and individual strengths and weaknesses in terms of stamina and speed.
However, all the evidence is that these differences are subtle ones, and that horses
simply do not exist who record their best times by running in ways markedly different
to general pars of efficiency. In the absence of specific and extremely detailed
knowledge about individual horses, we regularly apply general concepts (like the
effect of weight and of maturity) to them. Furthermore, the exponential nature of
sectional upgrading means that such small differences will often amount to little.
“What is the point of sectionals when there is no appetite for them from the
public?”
The premise of such a question needs to be challenged. Timeform’s own articles
about sectional timing – including the regular Sectional Debriefs on timeform.com
– have sparked great interest. Plenty of people seem intrigued by sectional timing
but need it to be put into context, by broadcasters and by writers, which has seldom

51
 SECTIONAL TIMING

happened to date. Whether this justifies spending large sums of money on installing
sectional timing is another matter, and, besides, much cheaper alternatives (including
sectionals taken from video analysis) exist. Ultimately, though, those who are not
interested in sectionals can ignore them. Universal approval was not required for
British racing to “do the right thing” and introduce overall race timing, starting stalls or
photo finishes, for instance.
“Don’t one-off sectionals miss out on a lot of important information?”
They may do. Most races soon settle into a certain tempo and then feature a distinct
increase in effort approaching the finish. One-off sectionals – which imply what
happened before the sectional and not just after it – often capture these races
perfectly well. However, some races do not conveniently fit this profile and would be
better dealt with by being broken into yet smaller parts. The Derby at Epsom (uphill,
round bends and downhill in roughly one-third portions) is a specific example.
The 2004 King George VI and Queen Elizabeth Stakes won by Doyen – in which the
runners went slowly, then too fast, and, as a result, finished quite slowly – is another.
Both those races are at a mile and a half, and such extreme variations seldom happen
at shorter. Even so, splitting a race into just two parts (before and after) significantly
increases the dimensionality of time analysis compared to considering just an overall
time.
“What is the point of sectional timing without considering a multitude of other
factors also?”
Anyone imagining that sectional timing in isolation will provide a passport to riches
is likely to be disappointed. Racing analysis is a complex and multi-factorial discipline,
in which one area influences another. However, in terms of assessing the ability and
aptitude of the protagonists, sectional-timing analysis is extremely powerful when
linked up with more accepted forms of time and form analysis. But other factors –
such as a horse’s fitness, its draw and the strength of its opposition – will have a major
say in how a horse performs on a given day: that will always be the case. Despite this,
Timeform’s lists of Sectional Debrief “to follow” and “to oppose” horses have shown
a good profit since their inception in April 2013, even though the circumstances of
a horse’s next race were unknown at the time of listing. Identifying horses that have
more ability than is generally supposed simply has to have worth.

52
 AN INTRODUCTION BY TIMEFORM

“There may be some worth in sectionals on all-weather, but what point is there
in them on Flat turf and over jumps?”
The main benefit of having sectionals on all-weather in Britain is one of scale: a
small number of all-weather courses cover a much larger amount of the total racing.
It is not the consistency of the courses (which vary much more than all-weather
critics acknowledge) or of the surfaces (which can vary in speed by large amounts).
Running efficiently, or non-efficiently, will have a more or less predictable effect on a
horse’s performance, and that applies whether the performance has taken place on a
synthetic surface, on Flat turf, over obstacles, or on a beach. All-weather racing has an
advantage in that rail movements are almost unheard of, and camerawork, through
which sectionals can be taken or validated, is usually the same from one meeting to
another. But there is no theoretical reason why sectional timing could not be applied
far and wide: indeed, turf Flat and jumps may even offer richer pickings than all-
weather, which is being extensively analysed already by a growing body of sectional-
timing enthusiasts.
“There may be some use in sectionals for data analysis after the event, but on-
screen, or “live”, sectionals serve no purpose other than to distract, yes?”
This is worth a chapter of its own: the next chapter.

53
 SECTIONAL TIMING

CHAPTER 12
IN-PLAY SECTIONALS
The final FAQ of the preceding chapter comes up quite often,
especially from those who seem to be antagonistic towards the idea of
sectional timing. The shorter version – “What possible point can there
be in displaying sectional times live?” – may be answered with “they
can enhance the understanding and increase enjoyment of an event
as it unfolds and provide new and exciting opportunities to bet on the
event as a result”.
But, in order to “enhance”, “increase” and “provide”, the meaning of those live
sectionals really needs to be spelt out. There has been a surprising reluctance to do
this among broadcasters, despite Timeform, for one, providing proof of worth on
occasions.
In order to understand whether a displayed sectional for a leader indicates a “fast”,
“slow” or “about right” pace, you need to apply the known sectional pars to a
projected time for the race. Crucially, this projected time is not a prediction of the
time in which the race will be run, but the time the race should be run in if the pace
were true. We are interested in how close to true a pace is, after all.
As has been emphasised a number of times already, it is easier to run slow than to run
fast, and more races are run inefficiently than efficiently (Timeform reckons as little as
20% of races at 14 furlongs plus are truly-run, or close to truly-run). It follows that the
majority of overall race times are not as fast as they could be.
To project an overall race time for a truly-run race, you need an estimate of the speed
of the surface and the ability likely to be shown by the winner (within the context of
its age and the weight it carries).
This is not as difficult as might be imagined: speed of surface can be judged from
race times of the opening contest or contests, especially as sectionals themselves will
identify the degree to which those early race times reflect the horses’ abilities; and the
likely ability shown by the unknown winner is straightforward for anyone who uses
standardisation as part of their form assessment, as Timeform long has done.

54
 AN INTRODUCTION BY TIMEFORM

It is beyond the scope of an introductory booklet on sectional timing to cover such

auxiliary material in detail, but an explanation of race standardisation by the author
can be found by using an internet search engine.
Having projected an overall time for a truly-run race (in which that overall time
reflects the ability of a standard winner of that race), you reapply the sectional pars
established previously to determine the sectional times which should be posted
along the way to result in that projected overall time.
The following table gives projected sectional times for Lingfield – one of the
few courses to display sectional times live – for races in which the overall time
equals Timeform standard time. But it must be emphasised that such a time will
vary (sometimes considerably) according to the precise speed of the surface and
the abilities of the horses involved. The par sectional times should be increased
or decreased in a pro-rata fashion as the projected overall time is increased or
decreased.

In-play sectional pars at Lingfield

Distance after 2f* after 4f after 6f after 8f after 10f after 12f Overall
5F 6yd 25.10s 47.32s 59.16s
6F 1yd 25.36s 48.13s 71.88s
7F 1yd 25.45s 49.48s 72.81s 84.92s
8F 1yd 25.96s 50.24s 73.87s 98.08s
10F 27.75s 52.31s 77.11s 101.62s 125.99s
12F 26.77s 52.18s 77.43s 103.03s 127.82s 152.32s
13F 26.95s 52.37s 77.63s 103.28s 128.75s 152.91s 165.29s
15F 169yd 22.42s 49.22s 74.60s 101.94s 128.06s 154.26s 204.60s

* opening sectional is not 2f in all instances

From this, it is not especially difficult to figure the range of times covering more-
or-less truly-run races, so that overly-fast and overly-slow times account for what
remains. Thereafter, it would be simple to colour-code these scenarios on-screen,
so that, say, an opening four furlongs of under 49.0s would be highlighted in red to

55
 SECTIONAL TIMING

indicate too fast, one of over 50.5s would be highlighted in green to indicate too slow,
and any time in between would show in amber to indicate the pace was “about right”.
That would be the pace for the leader, of course. A horse several lengths back in the
former scenario might be running close to par.
Over and above an increased understanding and appreciation of the sport,
knowledge of whether the pace is “too fast”, “too slow” or “about right” should make
in-play betting more informed, for everyone and not just an elite.
That is not to say that betting in-play would suddenly become easy – not a bit of it! –
but it might well become a more level playing field.
There are often still a bewildering number of options for converting the knowledge
of race pace into a bet, and sometimes precious little time in which to do it (bet-
placement software can help here). While a late sectional is preferred for post-race
analysis, an earlier sectional – or at least one which gives the punter a reasonable
amount of time to execute a bet – is likely to be better for in-play betting purposes.
Live-displayed sectional times do not come cheaply, thus far at least. But, where they
exist, they do provide opportunities for the clued-up punter to tackle in-play markets
with increased confidence. The future of live sectionals, accompanied by benchmarks
specific to the circumstances, may or may not be bright: but, with a little effort, it
could easily enough be red, amber and green.

56
 AN INTRODUCTION BY TIMEFORM

CHAPTER 13
THE FUTURE OF SECTIONALS
Predicting the future is difficult enough at the best of times, but predicting the future
of sectional timing has proved to be all but impossible to this point. Despite some
good initiatives, British and Irish racing is no further forward now than it was at the
turn of the century. There have been countless false dawns along the way.
Funding has been the main problem. Many of those in the corridors of power in
British horseracing have expressed enthusiasm for sectional timing, but someone has
to pick up the bill, and enthusiasm for doing that has been in much shorter supply.
The result is that, while Britain can justly claim to be a world leader in some aspects
of horseracing, it is a distant also-ran behind even minor racing jurisdictions like
Mauritius and Barbados where sectional timing is concerned.
Technology has also been an issue, with some of the higher-tech initiatives so far
having been dogged by problems. Technology has a habit of improving, however,
and whether it is GPS-based technology, wireless-based technology, break-beam
technology, drones, or some hitherto-unimagined technology, the future could well
involve better and cheaper solutions.
That aspect is in the hands of others, but a personal future of sectionals is something
you, dear reader, can be in charge of. It is not necessary to have access to high-tech
solutions to enter the intriguing world of sectional analysis, just access to video
replays and a stopwatch, or access to articles and discussions about the subject.
It is hoped that the preceding text will arm the reader with some of the tools required
for a fuller understanding of sectionals, from the basic approach of race sectionals to
the more complex task of upgrading performances.
Sectional timing does not appeal to everyone, clearly, and that is unlikely to change.
But the vast majority of those who have taken the trouble to understand it properly
find that it opens new and intriguing avenues of inquiry and provides insights
unavailable to those who ignore it.
As much as the potential for profit that arises from sectional analysis, it is this
potential for increased interest and enjoyment that should mean that sectional timing
– in some form or other – is here to stay.

57
 SECTIONAL TIMING

APPENDIX (By Bob Wilkins)

INTRODUCTION
Conventional speed figures are calculated in pounds, based on the “final time” (overall
time of the race), and make no allowance for the way the race was run.
However Simon Rowlands has proposed that if the pattern of running is not optimal,
the horse could have done better, and that a correction can be applied be adding a
few pounds to allow for non-optimal running.
The correction is derived from sectional timing over the last few furlongs of the race,
and two cases arise:
• If the horse finishes relatively slowly, then it probably ran too fast in the earlier stages
of the race, and could have recorded a better final time if it had run at a more even
pace.
• If the horse finishes relatively fast, then it probably ran too slowly in the earlier
stages of the race, and could have recorded a better final time if it had run at a more
even pace.
The proposed correction is

 d 
Dw = C   (O - A) 2
 D  [1]
where
Dw = correction in lbs
d = distance of the final section of the race
D = total race distance
O = optimal race speed
A = actual finishing speed (over the distance d)
C = a constant, approximately 1.25 if O and A are given as a percentage of the
overall speed

58
 AN INTRODUCTION BY TIMEFORM

Intuitively, the squaring of the

(O - A) term seems reasonable, as it w
gives a positive correction whether
the finishing speed is either too
slow or too fast (but this would
also be true if the magnitude of
the difference, i.e. ABS(O - A) were
used).
Equation [1] gives a parabolic
model of the dependency
between Dw and (O - A), which
is shown here. The graph is 0 (O-A)*(O-A)
symmetrical about the y-axis.
When d is small compared with D (i.e. the final finishing section is small), the
correction needs to be reduced correspondingly, which explains the appearance of
the term (d/D) in the correction formula.
The purpose of this work was to look at the fundamental basis of this idea, and to see
whether equation [1], or something like it, can be derived from first principles.

SIMPLE ANALYTICAL MODEL

Since the definition of what is an optimal pattern has not been given, the benchmark
used here is even pace, which is widely regarded as the best way to utilise energy
when running a race.

Kinematics (neglecting effect of start)

v0
For a 1-stage race run at a constant speed
v0 , the race time is given by T = D/v0).
Assume that constant speed v0 represents speed
the optimal case. The graph on the right
shows instantaneous speed versus distance
travelled.
0 D

59
 SECTIONAL TIMING

For a 2-stage race, the speed is different for the final distance d. See the graph below:

p
v0
r

0 D

Let p = per-unit increase is speed over 2nd stage, distance d (if increase is 5%, p = 0.05)
r = corresponding decrease in speed over the 1st stage (D - d)
q = d/D
This then gives:

stage 1 stage 2
speed (1 - r) v0 (1 + p) v0
distance travelled D-d d
time taken (D - d) / [(1 - r) v0 ] d / [(1 + p) v0 ]

If the race is completed in the same overall time as would be obtained with a constant
speed v0 , then the total time taken must equal D/v0 , so

(D - d) d D
+ =
(1 - r) v0 (1 + p) v0 v0

60
 AN INTRODUCTION BY TIMEFORM

Solving gives r
pq
r=
1+p-q [2]
For an increase (or decrease) in speed of p per unit over the 2nd stage, equation [2]
gives the corresponding decrease (or increase) required over the 1st stage which
would result in the same overall time (and hence the same speed figure).
Equation [2] is only useful for relatively small values of p and q, say p < 0.1 and q < 0.4.
If these parameters are too large a physically impossible value of r results e.g. if the
2nd stage is long and its speed is very low it is impossible to achieve an overall time
of D/v0 however fast the 1st stage is run. This is not a problem, however, because the
values of p and q being investigated are usually well within these limits.

Energetics
A 2-stage race which results in the same overall time as a 1-stage race results in the
same speed figure, but we will see that more energy and power is used for the 2-stage
race, which means that a faster time could have been achieved if the race had been
run optimally.
Assume that the power expenditure is can be represented as

P = k vx [3]
In Bioenergetics and Racehorse Ratings (BRR) it is shown that the power expenditure
is the sum of three components (a) the cost of locomotion, for which X = 1 (b) the
power to overcome air resistance, for which X = 3, and (c) the power needed to
supply kinetic energy, for which X = 2 (although this third component is also time-
dependent). As a first approximation the dependency can be represented by (3) with
1 < X < 3.
For the 1-stage race P0 = k v0 x is the power used in the optimally-run race.

61
 SECTIONAL TIMING

For the 2-stage race:

stage 1 stage 2
speed (1 - r) v0 (1 + p) v0
power k (1 - r) x v0 x k (1 + p) x v0 x
time taken (D - d) / [(1 - r) v0 ] d / [(1 + p) v0 ]
energy (= power x time) k (D - d) [(1 - r) v0 ] x-1 k d [(1 + p) v0 ] x-1

Adding the energies for the two stages and dividing by the total time T (= D/v0) gives
the following result for the average power:

[
Pav = P0 (1 - q)(1 - r) x -1 + q(1 + p) x -1 ] [4]
The red triangles on the graph below show the correction required DP = Pav - P0 , as a
fraction of P0 , plotted as a function of p, with X chosen to be 1.5. The correction can be
converted to a difference in pounds using the method described in BRR.
Positive values of p
0.003
represent faster than DP
optimal finishes, while
negative values are for
slower finishes. q = 0.375
0.002
For the graph q was
chosen to be 0.375,
which corresponds to
the case where stage 2 is
the final 3 furlongs of a 0.001
mile race (q = d/D = 3/8 =
0.375).
The blue circles plotted
on the graph are points -0.15 -0.1 -0.05 0 0.05 0.1 0.15
on the parabola DP = k p
p2, with k = 0.19. slower faster

62
 AN INTRODUCTION BY TIMEFORM

It can be seen that the data for positive p fits well to the parabola. The data for
negative p can also be fitted to a parabola, but with a higher value of k, suggesting
that the weight correction for slower than optimum races should be higher than for
faster than optimum racing.
The asymmetry of the data (red triangles) about the y-axis results from the asymmetry
in the speeddistance diagram shown on page 2 of these notes.
The effect of varying q (= d/D) can be seen in the next graph.

0.003

p = - 0.1

0.002

p = 0.1

0.001

0
0 0.1 0.2 0.3 0.4
q (=d/D)

The required correction increases with q and as seen before, the correction
required for slower finishes (negative p) is higher than for faster finishes. To a first
approximation the increase in DP with q could be represented by a straight line, but
an increase proportional to q1.4 fits the above curves better.
All these results were obtained with the exponent X chosen to be 1.5, but similar
results are obtained if higher values of X are used.

63
 SECTIONAL TIMING

CONCLUSIONS
The correction required is roughly proportional to d/D and a parabolic fit can be used
to represent the dependence on (O-A), so the proposed correction of equation [1] is
reasonable, i.e.

 d 
Dw = C   (O - A) 2
 D 
However, this work suggests that
1. The value of C needs to be higher for slower-finishing races
2. The dependency on (d/D) is not linear but more like (d/D)1.4
However, to date I have been to unable to estimate the value of C and for the present
it will have to be determined by fitting to actual race data.

Bob Wilkins
3 April 2013

64
Sectional Timing: An Introduction by Timeform
DOWNLOAD THE
NEW APP TODAY

Includes free cards, results and features

plus premium tips and Race Passes

Find winners on the move