Welcome to a new weekly feature, Clock Watcher, where we'll shine a light on a few horses that might be interesting to follow from a speed and/or sectional perspective. It is my hope that this column will also serve to introduce, embed and reinforce various concepts which may be unfamiliar at this stage.
Generally speaking, a run considered of sufficient merit to appear here will have two components: the horse will have recorded a time which is at least reasonably quick for the conditions; and the horse will have recorded a noteworthy upgrade on that performance.
The first component is fairly self-explanatory even if defining what is "at least reasonably quick" is highly subjective. Geegeez.co.uk doesn't currently produce its own speed ratings (and there is no plan for that to change at this stage), so for our purposes we will use Racing Post's Topspeed figures, which are published under license on this site.
The second component requires a bit more introduction. What is an 'upgrade', how does a horse achieve one, and how is this quantified?
What is an upgrade?
Track and field athletes run at their most efficient level - enabling them to produce their fastest times - when they travel at a constant speed. For instance, when Kenenisa Bekele broke the 5000m world record in 2004, a record which still stands today, his 1000m split times - or sectional times - were as follows:
12:37.35 (2:33.24, 2:32.23, 2:31.87, 2:30.59, 2:29.42)
Let's tabulate that:
A few months later, in the Olympic 5000m final, they covered the first 4 kilometres in 648.62 seconds, almost 41 seconds slower than the world record pace. Bekele, overwhelming favourite for gold, was readily out-sprinted and had to settle for silver, the winner recording a final time of 794.39 seconds, 37 seconds slower than the world record.
In a race where they crawled (relatively) and then sprinted, Bekele was unable to produce his best form. He could not run inefficiently to the same effect as his vanquisher, the Olympic 1500m champion Hicham El Guerrouj, whose superior kick facilitated his victory.
We know what is 'efficient' based on the body of similar historical races, and we call this par.
In simple terms, any deviation from efficiency - or par - whether fast early then fading, or slow early with a rapid finish, earns an upgrade. Thus, in this case, both the winner and second - as well, indeed, as the third through to sixth placed finishers - would have received upgrade figures.
An upgrade, then, is a recognition of the degree to which a horse raced inefficiently.
It should be noted that racing inefficiently will not necessarily prevent a horse, or an athlete, from winning. Indeed, El Guerrouj 'got the run of the race' back in 2004, that slow time suiting his questionable stamina but stronger kick. The primary objective is, after all, not to break records but to win the race.
What about par?
There is more detail in the User Guide around what may be new terms to some readers, and I'd encourage you to check that document (page 63 onwards in the current version, click here).
However, for the purposes of expediency, a quick line on par here. Par is the threshold against which all subsequent races over a course and distance are measured. From the User Guide,
Par is an assessment of the optimal energy distribution – based on relative time – between the sections of a race. It is not an average of all sectional times. Rather, it follows a fairly complex formula which uses an ‘nth percentile’ race as par. Further information can be found in Simon Rowlands’ excellent Sectional Timing Introduction report, available at this link. Indeed, that document is highly recommended for anyone keen to get a head start with the applications of sectional information.
So, in simple terms, par is a baseline, a means by which we may better understand the context of a performance.
Let's look at some examples.
An obvious one...
We'll start with a sore thumb, a horse on everyone's radar regardless of whether via visuals, sectionals or form. The very well related Waldkonig made his debut in a run-of-the-mill Wolverhampton novice stakes for two-year-olds on 7th December. A Kingman half-brother to Arc winner Waldgeist, he was sent off 6/4 favourite over the extended mile trip.
In the end, he won by nine widening lengths; the data offer some interesting footnotes to that emphatic victory.
There is a lot going on in this image, so let's take it step by step. First up, note that I have selected 'Call Points' (a five section breakdown) top left and I have clicked the 'Show Chart' button, which then changes colour and displays 'Hide Chart', the action that would happen upon a further click. So those are my selected parameters. (I also have the data view selected from my My Geegeez page, check the User Guide for more on that).
Beneath the blue buttons is a line of five coloured rectangles. These are the Call Points sectionals for the race. That is, they relate to the race leader at five points during the race, specifically the six-furlong, four-furlong, three-furlong, two-furlong and finishing posts.
The colour of the rectangles indicates the relative speed of each section, on a cold/slow/blue to hot/fast/red scale. Thus, this race was even (green) early, slow (blue) in the middle, and fast (orange) late. The OMC (Opening, Midrace, Closing) view below captures this more succinctly and is a better place for newbies to start, due to there being fewer data points.
Getting back to the main image, and the main part of it, we see a chart. This chart is highly configurable but the image shows the default, which is the sectional percentage data, by furlong, for the winner - and with the black par line also displayed. Any/all runners can be added or removed to/from the chart by clicking their name underneath or using the 'toggle' button top left.
Next to the toggle button is a statement of how many races comprise the par calculations and, therefore, the degree of confidence in par. In this case - indeed in the vast majority of all-weather race cases - confidence is high. At this stage, confidence is more limited elsewhere while the body of data grows as more races are run over various courses and distances.
The chart reflects what the coloured rectangles are saying: that the leader went even-ish (slightly above par) early, slowed up notably in the middle of the race, before finishing very strongly - well above the black par line.
Beneath the chart is the full result table, which has a familiar look to it. I have clicked on the winner's finishing position (i.e. on the text that says '1st') to reveal his sectional data - coloured rectangles for Call Points (including split times, aggregates time, and sectional time as a percentage of overall time, i.e. sectional percentage), running lines (the horse's position in the field and distance behind the leader, or in front if the race leader) - and in-running comment.
The rightmost column in the result table is 'UP', and it contains the upgrade figure. In this example, Waldkonig was calculated as having an upgrade of 29 by our algorithm. Again, in terms of quantifying ability, this tells us little more than that, like his father, Waldkonig is able to quicken impressively off a steady pace.
Waldkonig was given a Topspeed rating of 47 for his time performance in the race. That is far from a standout rating and would not highlight the horse's effort as noteworthy, though of course the nine length winning margin would be missed by nobody. By applying the upgrade figure to a representation of the time performance we get closer to an understanding of the merit of the effort: clearly it takes more ability to quicken off a fast pace than a slow one, with the degree to which a horse quickens also worthy of note.
We've been playing with combining various numbers to produce some sort of 'composite' time/performance rating, though I must declare at this stage that I'm not 100% certain that adding upgrades to Topspeed is a sensible thing to do.
We are currently trying to establish whether it improves the predictive ability of the raw speed figure: they are calculated on different scales so it is probably not entirely sensible to simply add the two together.
Nevertheless, there is some indication in the work done to date that this somewhat contrived 'combo' number has merit. In the case of Waldkonig, his 47 gets an extra 29 for a 76 overall. That is a better reflection of his performance, though probably not of his ability given this was a debut on a track that was likely not ideal. In any case, what it tells us unequivocally is that, in a race where the pace scenario looks muddling, Waldkonig is capable of a searing turn of foot.
A (slightly) less obvious one...
At a slightly less 'could be anything' level, trainer David Brown rewarded his and connections' patience when National Anthem, off the track for 417 days since running poorly at the same venue, blasted home in a six-furlong novice event at Southwell. Brown is the horse's third trainer in three career starts spanning 821 days and a wind operation!
Sent off at 15/2, fifth favourite of six but not completely unfancied, his performance was very different in sectional terms to that of Waldkonig, as the image below illustrates:
Here we see from the running line that National Anthem jumped very alertly and maintained that advantage, albeit that it was diminishing in the final furlong. He was better than four lengths in front after a furlong and fully nine lengths clear with an eighth to go. Little wonder that he tired close home. Also little wonder that he's entered over five furlongs at the same track on Monday where it will be very interesting to see how he goes in a handicap off a mark of 75, if taking up his engagement.
Far more speculatively...
Meanwhile, down in the basement, a horse called Disruptor might pop up at a price some time soon. He ran on 30th December at Lingfield, finishing third, and as can be seen from the below he ran an almost polar opposite race to par - based on the five Call Point sections:
This lad has had a few goes - twelve, including one since, to be precise - and has showed much improved form when leading or racing prominently recently. Prior to his run on Monday, where his inexperienced (14 rides) jockey shot up in the air as the stalls opened and then got sandwiched between two no-hopers for most of a furlong, he'd run his three best races from the front.
If/when he can get a slightly softer advantage - note the undesirable red zone section from five to four in his running data above - he has a chance to see his race out more effectively, albeit very likely in low grade company and with a more experienced pilot on top.
That said, looking more closely at the draw (DR) column below, it is worth noting that he has been consistently fortunate with his stall position in recent starts.
[NB note also that, in 'Show Sectionals' mode, races without sectionals have blanks. Hovering over the running lines segment displays details of the performance, including comments, position, distance beaten and jockey].
That's all for this inaugural edition of Clock Watcher. I hope it has provided food for thought and that, over time, it will support your understanding of the new data we are beginning to provide and how you might best take advantage of it for yourself.
Until next time...
p.s. as of Wednesday 8th January, sectional data is now live for Gold subscribers on geegeez.co.uk. You will need to enable it from the Race Card Options section of your My Geegeez page. On that page, you will also find a link to the most recent version of the User Guide, in which there is a comprehensive outline of sectional timing and how it is published on this site.
The current coverage comprises Total Performance Data tracks, as it is from them that we license our data. We hope to be able to integrate both Ascot and RTV (UK) tracks in due course. To be clear, we have no in house sectional aggregation function. Rather, we license 3rd party data as a publisher and aim to add value in the visualisation of that data. I very much hope by mid-year we have a far more comprehensive provision in terms of track coverage.