Back in September 2015 I looked at whether there was more value backing horses that were being backed, or horses that remained the same price, or horses that drift in price. The conclusion I came to was that if you used Betfair as many people do, drifters were better value. In that article however, when looking at horses that ‘drifted’ I focused on all horses that lengthened in price (from opening show to final SP). Hence I did not take into account how much the horse had drifted. I treated a horse opening up at 6/1 and ending up at 13/2 the same as one starting 6/1 and finishing up 10/1. In this article I am going into far more detail and looking at horses that change in price significantly – or at least what I perceive to be significantly! Also in this article I will be examining all races, in the original article it was just flat races. I have gone back to the start of 2012 so there is a mountain of data to crunch! For this article I have been using the excellent database from www.horseracebase.com.

 

Initially, I am going to look at horses that shorten significantly in price – what many people tend to call ‘steamers’. I have looked at the opening show odds (usually about 10-12 minutes before the race) and compared them with the final SP. All profits and losses however have been calculated to Betfair SP. Here is a price breakdown for our ‘steamers’:

 

Price change	Runners	Wins	Strike rate %	Profit/loss to BSP	ROI %
Opening 5/1 to 6/1  SP 100/30 or less	813	176	21.7	- £75.67	- 9.3
Opening 13/2 to 7/1 SP 4/1 or less	681	106	15.6	- £131.05	- 19.2
Opening 15/2 to 8/1 SP 5/1 or less	1402	208	14.8	- £127.40	- 9.1
Opening 17/2 to 9/1 SP 6/1 or less	843	117	13.9	- £41.09	- 4.9
Opening 10/1 SP 7/1 or less	1669	167	10.0	- £307.88	- 18.5
Opening 11/1 SP 8/1 or less	1277	139	10.9	- £12.64	- 1.0
Opening 12/1 SP 9/1 or less	3394	325	9.6	- £63.23	- 1.9
Opening 16/1 SP 11/1 or less	1599	128	8.0	- £85.63	- 5.4
Opening 20/1 SP 14/1 or less	1930	110	5.7	- £103.07	- 5.3
Opening 25/1 SP 16/1 or less	1558	69	4.4	- £134.01	- 8.6

 

Each price bracket has shown a loss and in general the horses that started nearer the head of market have lost more money in percentage terms than horses that started at bigger prices. Now it is time to look at those horses that drifted with similar price brackets in reverse:

 

Price change	Runners	Wins	Strike rate %	Profit/loss to BSP	ROI %
Opening 11/4 to 100/30 SP 5/1 or bigger	249	41	16.5	+ £27.11	+ 10.9
Opening 7/2 or 4/1 SP 13/2 or bigger	245	29	11.8	+ £21.69	+ 8.9
Opening 9/2 or 5/1 SP 15/2 or bigger	611	64	10.5	+ £91.38	+ 15.0
Opening 11/2 or 6/1 SP 17/2 or bigger	572	47	8.2	+ £28.68	+ 5.0
Opening 13/2 or 7/1 SP 10/1 or bigger	1185	92	7.8	+ £110.60	+ 9.3
Opening 15/2 or 8/1 SP 11/1 or bigger	2303	158	6.9	+ £217.05	+ 9.4
Opening 17/2 or 9/1 SP 12/1 or bigger	2166	135	6.2	+ £35.10	+ 1.6
Opening 19/2 to 11/1 SP 16/1 or bigger	1130	55	4.9	+ £84.03	+ 7.4
Opening 12/1 to 14/1 SP 20/1 or bigger	2470	87	3.5	+ £112.85	+ 4.6
Opening 16/1 SP 25/1 or bigger	1452	43	3.0	+ £60.34	+ 4.2

 

Remarkably all price brackets made a profit to BSP. Not surprisingly the strike rates are quite low, but making a profit is all about gaining value and clearly these drifters are value.

 

At this stage though I needed to change the parameters slightly, or at least how I looked at market drifters. If I wanted to use the exact data from each price bracket and test each one against a number of ideas, that was going to take me far too long – or at least with the software I was using I could not work out a shortcut. Hence I decided to use a percentage price change option that the software offered. I could not quite match up my original data exactly, but using a price drift of 35% or more (from opening show to SP), I got pretty close. The initial overall figures for all these drifters produced a smaller overall return than my specific price brackets did, but it made the next testing phase much easier. If I found some promising angles then I could go back and test them on the original price brackets.

 

Using the 35% or more drift figures gave me over 9000 qualifiers and a small 3% profit. However with a strike rate less than 7% a great deal of patience would be required. I decided to look at the going to see if that improved returns – my hypothesis was the softer the ground the better, and so it proved. The figures for soft or heavy going with horses that drifted in price 35% or more produced the following results:

 

Runners	Wins	Strike rate %	Profit/loss to BSP	ROI %
1864	135	7.2	+ £417.53	+ 22.4

 

We have to appreciate that we are not going to magically improve the strike rate – but despite winning roughly just 1 race in every 14, a profit of 22p in the £ has been achieved.

 

Moving onto class of race now:

 

Class	Runners	Wins	Strike rate %	Profit/loss to BSP	ROI %
Class 1-3	1356	80	5.9	- £268.36	- 19.8
Class 4	2587	161	6.2	- £28.93	- 1.1
Class 5-7	5320	385	7.2	+ £585.59	+ 11.0

 

The lowest class of races have produced some decent overall profits – 11p in the £ from over 5000 qualifiers. Possibly this is due to the races being generally less competitive, or at least contested by more inconsistent animals.

 

What area to test next? I suspected that the less punters knew about the horse the better from the perspective of drifting out in price. If an unexposed horse was drifting few punters would bat an eyelid and certainly not many would back them. They would not know enough about the horse and would assume that it is drifting for a reason. Hence the price is likely to drift even further and potentially give us even greater value. This certainly is the case when we look at horses that have never run before and have drifted 35% or more from their opening show odds. These runners produced 97 wins from 1243 runners (SR 7.8%) for a huge profit to BSP of £807.33 (ROI + 65%). It should be noted that each of the 6 years of study produced a profit for these debutants so it was not just one crazy year when a few big priced winners happened to win.

 

Horses that had run just once previously also showed a profit, although nowhere near as big as the one for debutants – these runners produced 60 winners from 809 runners (SR 7.4%) for a profit of 82.65 (ROI + 10.2%).

 

For the system fans out there I suppose I should give the figures for the system that combines the three factors I have highlighted so far namely going, class and career runs ….. so looking at horses that have drifted 35% or more from opening show that raced on soft or heavy ground in class 5 or lower and were making their debut – they have produced 23 winners from 183 runners (SR 12.6%) for a profit of £269.12 (ROI + 147.1%). Yes this is a serious case of back-fitting a system to suit ones data, but hopefully it is interesting nonetheless.

 

Having dug a bit deeper I thought it was time to go back and look at the price changes I looked at earlier in this article. I felt as the class data had more data, it made sense to look at that. Hence I am using the price brackets I initially used on just races class 5 or lower. I hoped this would improve the figures, albeit from fewer selections. Here is what I found:

 

Price change	Runners	Wins	Strike rate %	Profit/loss to BSP	ROI %
Opening 11/4 to 100/30 SP 5/1 or bigger	164	20	12.2	- £30.09	- 18.3
Opening 7/2 or 4/1 SP 13/2 or bigger	149	20	13.4	+ £39.38	+ 26.4
Opening 9/2 or 5/1 SP 15/2 or bigger	374	42	11.2	+ £89.67	+ 24.0
Opening 11/2 or 6/1 SP 17/2 or bigger	330	27	8.2	+ £25.29	+ 7.7
Opening 13/2 or 7/1 SP 10/1 or bigger	687	59	7.8	+ £156.16	+ 22.7
Opening 15/2 or 8/1 SP 11/1 or bigger	1277	97	7.6	+ £277.93	+ 21.8
Opening 17/2 or 9/1 SP 12/1 or bigger	1128	72	6.4	+ £53.61	+ 4.8
Opening 19/2 to 11/1 SP 16/1 or bigger	598	26	4.3	- £24.46	- 4.1
Opening 12/1 to 14/1 SP 20/1 or bigger	1366	52	3.5	+ £209.85	+ 15.4
Opening 16/1 SP 25/1 or bigger	775	22	2.8	+ £6.96	+ 0.9

 

As one can see we have gone from a full house in terms of profit for each price bracket to two losing ones. Having said that, when we compare the overall figures for both sets, those including the class factor have improved matters. The original totals saw 12383 qualifiers of which 751 won (SR 6.1%) for a profit of £788.83 (ROI + 6.4%); adding the class 5 or lower factor gives us 6848 qualifiers of which 437 won (SR 6.4%) for a profit of £804.30 (ROI + 11.7%).

 

As I have mentioned before, it is difficult to replicate in real life research data profit which takes into account Starting Prices or opening shows. However, if you are able to bet when racing is on, it is possible to apply the theory. You will ideally need two screens or a split screen to monitor Bookmaker odds while having your Betfair screen available too. In addition a live feed from Racing UK or At the Races will give you the opportunity to place any bets just before the off. Of course you won’t be 100% sure of the eventual SP but you will invariably predict the correct SP or worst case just one ‘tick’ away.

 

I may look further into this idea or compare forecast price to SP in a future article. I hope you have found this article enlightening and given you some food for thought.

 

David Renham