|Welcome Guest ( Log In | Register )|
Articles >> horse-racing >>
Pace in National Hunt Racing part 2
Pace in National Hunt Racing part 2
by David Renham
This is a second piece in a series of articles on pace in National Hunt racing. I focused in particular on novice chases, and for this second article I am sticking to chases, but turning my attention to handicaps. In order to help compile this article I am using some pace data from Proform www.proformprofessional.co.uk which is the best piece of software I have found for anything pace related.
To recap a stat mentioned in the first article, in 2011 the early leader has won 24% of ALL chases. Effectively this means front runners win twice as often as they should given a level playing field. Let me now focus on handicap chases, extending the statistical sample by looking at data from 2008 (to Nov 30th 2011). Firstly let us look at the results splitting them into three – by front runners (L), prominent racers (P) and hold up horses (H):
Now anyone looking at this table may think that prominent runners are the best value of the three groups as they have won the most races. However, we need to take into account the number of runners in each of the three categories as this gives us a better understanding of the pace bias. The following table gives us this extra information:
Now, as you can see, front runners account for a much smaller proportion of the total runners than the other two groups. The reason for this is fairly straight-forward – in general you only get one front runner per race (occasionally two if two horses are vying for the lead), but the remaining runners in a race will be split fairly evenly between prominent runners and hold up runners.
In order to explain the front running bias more clearly let us imagine a 9 runner handicap chase, which is roughly the average field size from this data sample. The split in terms of running style would be something like:
1 front runner; 4 prominent runners; 4 hold up runners
Therefore the winning percentages for each group in 9 runner races should be 11.1% for front runners, 44.4% for prominent runners and 44.4% for hold up horses. However, , using the results from the first table as our benchmark, the actual figures are 20% for front runners, 48.5% for prominent runners and 31.4% for hold up horses. Hence, front runners perform well above their perceived expectations, prominent runners slightly above their perceived expectations and hold up horses significantly below their perceived expectations.
So as with novice chases, we can see there is an advantage to front runners – the question is how can we take advantage of it? I noted in the last article that it is far from an exact science in terms of trying to pick the front runner before the race has started. The front running edge becomes somewhat redundant if you are unable to predict the front runner pre-race. The plain fact is that you will never be able to successfully predict the front runner in all handicap chases – what we need to do therefore is try to find ways to improve our chances of doing so. Trainer data is an option although I tend to use this for races where past horse data is limited (such as novice chases over the jumps, or 2yo maidens on the flat). For handicap chases we are dealing with more experienced horses that have run more often and tend to follow a specific style of running, be it front running, prominent or hold-up. This is shown quite neatly when looking simply at the running style of handicap chasers on their most recent run and seeing what effect that had on its next start:
- Horses that led LTO went onto lead again in their next race 33% of the time;
- Horses that raced prominently LTO went onto lead in their next race 12% of the time;
- Horses that were held up LTO went onto to lead in their next race 5% of the time.
Therefore in handicap chases, horses that led LTO are over 6 times more likely to lead next time out than horses that were held up LTO.
So we can improve our chances of predicting the leader in a handicap chase using the LTO running style. However, if we use the last three runs, we can improve our chances further of predicting the front runner:
- Horses that led in two of their last three starts and raced prominently in the other went on lead in their next race 36% of the time;
- Horses that led in all of their last three starts went on lead again in their next race 55% of the time.
Now, when trying to predict the leader, we want to also try and rule out as many runners as possible to increase our chances of doing so. We know that horses held up in their last race have a 5% chance of leading this time; the chances of leading lessens if horses have been held up for three consecutive runs – their chances of front running now drops to a dismal 2.7%.
So it is clear that a horses’ running style LTO has a bearing on their likely running style next time; this strengthens when you combine their running styles in their last three races.
One additional feature of the Proform software is that for each race, pace figures are calculated using the last three races. Scores are given as follows:
4 – for early leaders/front runners
2 – for prominent racers
0 – for hold up horses
Hence, a horse that has led in all of this last three starts gets 12 points; a horse that has race prominently in all 3 starts would get 6, etc, etc. Of course you could do this all by hand using past results on either www.sportinglife.com or www.racingpost.com and by checking the relevant races with their comments in running. However, this would take you about 5-10 minutes to rate a race, and if there were 6 handicap chases in the day then that could get close to taking up an hour of your time. Hence, using Proform can save you considerable time if wishing to use their horse pace figures.
We have already seen that past running styles do influence subsequent races, so one would expect the horses top rated by the Proform pace figures to lead more often than the lowest rated ones. I decided to check this theory out – I used the pace ratings rank data in the software to see what I found:
The figures correlate extremely well with the top rated runners clearly the most likely to lead than the second top rated, the second top rated more likely to lead than the third top rated, and so on. Of course a success rate of 33.4% still only means we a successful front running prediction roughly 1 race in every 3 – not only that we know that only around 20% of this leaders will win, so clearly this is not a license to print money. However, let us see how well these figures do on an average racing day – in terms of predicting the early leader. The day I have chosen is 1st December 2011. Firstly let us look at the pace figures for the 1.50 at Market Rasen which was a 3m5f handicap chase:
Copper's Gold 12 Canal Bank 4
Matmata De Tendron 6 Duke Of Ormond 4
Kercabellec 6 Our Columbus 4
This looks quite a clear cut race in terms of pace prediction – Copper’s Gold had led on all of his last three starts and was 6 points clear of his nearest rivals. Copper’s Gold therefore looked to have a very strong chance of being the early front runner – the figures proved spot on as Copper’s Gold led from the start. Indeed he was only headed after the final fence finishing a gallant second.
The next handicap chase was the 2.10 at Wincanton over 3m 1f. The pace figures were as follows:
Magot De Grugy 6 Master Cobbler 4
Just The Job 6 Upham Atom 2
Tarquinius 4 Nudge And Nurdle 2
Freeline Fury 4
These figures are extremely tight with two joint top rated horses, and no horse that has led more than once in their last three starts. Hence predicting the leader here is tricky and perhaps illustrates that we may need to look in more depth than simply the highest rated. As it turned out one of the top rated Just The Job did lead before being headed 3 fences from home (finished 3rd).
The next qualifying race was at 2.25 - the 3m 1f handicap chase at Market Rasen. The pace figures were thus:
Chesapeake 8 Lucky Landing 2
Flichity 6 No Principles 0
Kadouchski 6 Night In Milan 0
Larks Lad 6 Livvy Inn 0
The problem with this race was that it was a novices’ handicap chase meaning that all of the runners had limited chasing experience and hence past pace figures become much less relevant. Indeed, the running of the race proved this somewhat as Lucky Landing led despite having a very low pace figure of 2. I personally ignore novice handicap chases when analyzing the race from a pace perspective and suggest you do the same.
The final handicap chase of the day was the 3.50 Wincanton over 2m 5f. The Proform pace figures were as follows:
Gunship 10 Morenito 4
Ere Alfie 8 Betheholygobbs 2
Mylord Collonges 8 This Way 2
Randjo 6 Folie A Deux 2
Marias Rock 4 Local Present 0
Pacha D'Oudairies 4 Legends Lass 0
Silver Dollars 4
Gunship looked the most likely front runner but Ere Alfie and Mylord Collonges had the potential to lead also. As it turned out, Ere Alfie and Gunship shared the front running duties so again the figures were pretty much spot on.
As this typical day indicates, the raw Proform pace figures certainly give punters a decent edge when it comes to predicting the front runner in a handicap chase. Clearly there is more to it than that but it gives you an excellent starting point.
Earlier in this article I mentioned the fact that it is impossible to predict the front runner all of the time in handicap chases. The closer our success in front running prediction gets to 100% the more likely we are to make money betting these selections. We know that front runners in these races do have a distinct edge – the strike rates discussed earlier illustrated that. However, just for the record, if you had managed to predict pre-race all the front runners in handicap chases over the period of study, you would have secured a profit to SP of £11,972 to £10 level stakes! Indeed to Betfair SP this would have increased to over £18,000. Perhaps you can see why accurately predicting the front runner is worth the effort!
Next time I will expand further upon my National Hunt Pace research.
Copyright echo date("Y");?> PunterProfits.com