Over the past year I have been working on a method to use in sprint handicap races that gives me an accurate numerical representation of a horses’ chance. Essentially I have been trying to create an accurate ratings system.
I am very excited about my latest research as I feel it has the potential to make significant profits over the months and years to come.
The research to date has indicated that there are currently two optimum betting strategies based on my ratings yielding over 30% return in investment with strike rates of 29.5% and 31.5% – backing at £25 per bet both methods would have produced a combined profit of £4540.
The article below describes my methodology and approach.
As stated above over the past year I have painstakingly been developing a ratings system. The long term plan is to produce a set of ratings that will outstrip the best ratings systems on the market. Time will tell whether I can achieve this but that is the ultimate goal.
My opinion is that any approach that attempts to rate horses must be specialized and focused only on certain race types. The qualities a horse needs to win over a mile and half for example are different to those required to succeed over shorter distances. Similarly what works for handicap races is unlikely to work for listed races or claimers etc.
For my research to date I have targeted only 5 and 6f all age handicaps. I have already started research on longer distance races with a view to creating additional ratings methods for other targeted situations. However for this piece I want to concentrate your attention on completed research and that is 5 and 6f all age handicaps.
What Are Ratings?
Ratings are simply a way of giving runners a numerical figure that represents their chance of winning. The higher the rating the more likely the horse should win – well that is how it should work if the ratings are any good!! Many people use ratings to help them narrow down a race to a shortlist of candidates; others back the top rated; others use them in conjunction with their form study.
I do not want to let the cat out of the bag as to the precise nature of my ratings approach. However, to give you an idea of how it works I can tell you that each horse in a qualifying race type is assessed on ten key elements of form. Hence a 102 rated horse for example will have passed two form tests; a 105 rated horse on the other hand will have passed 5 tests, etc. As there are 10 key elements used the maximum possible score a horse can achieve with my method is 110.
Are Ratings Better Than Systems?
Both ratings and systems share the advantage that strategy can be based on many years of historical racing data. Betting is based on cold hard statistical facts and figures. A vast improvement over the random hunch backing employed by the average betting shop mug punter. A lot of the possibilities for human frailty are also removed as the strategy is researched and defined in advance. Following it then becomes mechanical each day.
I have had great personal success with systems however there are certain theoretical arguments as to why my ratings research may yield even better returns.
One of the common arguments against a traditional system type approach is that a qualifier will be a qualifier irrespective of the opposition faced on the day. Your system bet would be a system bet for example with or without the addition of Kauto Star as opposition within the runner line up. A ratings approach aims to overcome this by assessing and scoring each horse in the race.
A further common argument against systems is that they can suffer from over betting or odds available for selections may drop over time. This effect is not always a result of money from specific system followers. It can be a result of bookmakers adjusting the prices when certain elements of form are displayed. Take a system based on past trainer performance for example. If that trainer has been profitable in the past rest assured the bookmakers know that too and are likely to account for that fact in future odds offered.
A ratings approach provides extra protection against this due to an effect termed “Redundancy”.
A bridge engineer explained to me redundancy in bridge design terms. A motor way bridge for example should still stand if a truck were to hit and destroy one of the central supporting columns. The design has redundancy if it can withstand the removal of key supporting elements.
Systems can maintain strike rate but start to lose due to odds shortening as selections tend to focus on specialised areas of form.
My ratings approach having ten separate form elements is more likely to withstand odds reduction on one or more individual element.
Summary So Far
Right, at this stage I have explained some of the reasons that inspired me to investigate and research a ratings approach. What I have produced is a specialized method of scoring horses in 5 and 6f all age handicaps.
You are likely asking yourself key questions such as:
1 - Are these top rated horses profitable?
2 - What happens only if we focus only on horses who achieve a score of X
or more?
3 - What happens if we focus only on horses who out score / are clear rated
by a margin of X?
4 - At sprint distances draw effect can be considerable. Is there benefit
in accounting for that?
These are four important questions I asked myself as the answers are crucial in determining the optimum betting strategy derived from the raw ratings.
Past Performance Analysis
My race research has involved considerable data analysis and tedious number crunching. During the 51 months from January 2004 to March 2008 the all age sprint handicap ratings produced the following results for clear top rated horses:
Clear Top Rated runners 1593
Winners 337
Strike Rate 21.2%
Profit (to 1pt level stakes) +84.5pts
ROI +5.3%
You may say – a 5% return coupled with a strike rate of around 1 win in 5 – what is so special about that? Well let us have a look at the performance of the Racing Post Ratings during this period with their top rated horses in all age sprint handicaps:
Strike Rate 12.2%
ROI -18.7%
Now let’s look at the performance in all age sprint handicaps for the speed ratings in the Racing Post – known as Topspeed:
Strike Rate 11.4%
ROI -20.9%
Looking at those figures, I hope you can appreciate that the initial results are encouraging. They are even more encouraging perhaps when we also look at the percentages for horses finishing in the first three.
My Ratings 47%
Racing Post Ratings 34%
Topspeed 32%
So the basic raw figures look very good. In addition to the overall strike rate and profit level, when we look at the ratings’ scores we can see that there is a definite pattern:
101 – SR 18.7%
102 – SR 17.3%
103 – SR 17.9%
104 – SR 22.9%
105 – SR 31%
106 – SR 31.9%
107 or more – SR 33.3%
As these strike rate percentages show, the higher rated horses have a much better win ratio. You would expect this of course, but it is good that the figures show this so clearly. Indeed, focusing on 105 rated horses or bigger we get the following results:
Strike rate 31.5%
Profit +82.6pts
ROI +30.9%
For the record horses rated 104 also have proved profitable but the ROI is 6.3%. Hence, this seems to suggest that horses rated 105 or more are definitely worth following.
The next area of research to show you is the performance of the top rated in terms of how many points clear they are of the second top rated. The results are shown below:
Clear by 1 point – SR 19.1%
Clear by 2 points – SR 18.9%
Clear by 3 points – SR 26.5%
Clear by 4 points or more – SR 30.9%
Again the trend is the bigger the gap, the better the strike rate. This once again gives more confidence to the ‘feel’ and make up of the ratings system. Hence it makes sense to investigate the performance of the higher rated horses coupled with a gap of 3 or more points at the top of the ratings. The figures look good. For those rated 104 or more (+3 or more point gap):
Strike rate 29.5%
Profit +99pts
ROI +30.5%
For rated 105 (+3 or more point gap):
Strike rate 32.3%
Profit +66.2pts
ROI +34.5%
All the stats look very promising.
Moving on with the research it seems that the all weather has provided some good results in terms of strike rate especially. The overall aw results are as follows:
Clear Top Rated runners 604
Winners 145
Strike Rate 24.0%
Profit (to 1pt level stakes) +48.4pts
ROI +8.0%
Part of the reason for a higher strike rate as compared with the overall figure is almost certainly down to number of runners. On the all weather there tend to be less runners than on turf so it would come as no surprise for the strike rate to be higher. The good news is though that the clear top rated runners have made a profit on both turf and the all weather.
One factor I deliberately did not factor into the ratings (the only one I may add) is draw bias. Draw bias is one variable that is simple just that … VARIABLE. It is difficult to quantify these days even at the most draw biased tracks. My hope was that the results would show that courses with no traditional draw bias would outperform courses that are draw biased, or courses that show periodic draw bias.
Essentially courses will a level playing field should see better results for a ratings method.
Looking at the all weather first the theory seems to hold true. Essentially,
Kempton has a significant draw bias over 5 and 6f, while Wolverhampton also
has a bias, albeit not as strong as a year or two back. Lingfield has no significant
draw bias, and Southwell also has no strong bias these days.
Let us look at the results for the clear top rated horses at the courses with
draw bias – Wolverhampton and Kempton:
Clear Top Rated runners 304
Winners 64
Strike Rate 21.1%
Profit (to 1pt level stakes) -10.9pts
ROI -3.6%
Now let’s look at the non draw biased Lingfield and Southwell:
Clear Top Rated runners 300
Winners 81
Strike Rate 27.0%
Profit (to 1pt level stakes) +59.3pts
ROI +19.8%
Hence, as stated earlier the theory seems to hold true. The non draw biased
courses have shown much improved results – both strike rate and returns.
In order to test this theory further though, it has been necessary to look
at more data for turf racing.
I decided to look at the strongest group of draw biased courses, after which
I would test the strongest group of non draw biased courses. The draw bias
courses chosen were: Beverley, Chester, Thirsk, Catterick, Warwick, Ripon,
Newmarket, Newcastle, Carlisle.
Beverley and Chester have been consistent draw biased courses; Thirsk fairly consistent with high draws usually best, but certainly middle draws at a disadvantage; Catterick and Warwick see low draws tending to be best; Ripon either rail tends to have an advantage; Newmarket and Newcastle are two courses where draw bias occurs strongly from time to time, but difficult to predict; Carlisle sees middle draws at a big advantage in recent years.
Clear Top Rated runners 236
Winners 42
Strike Rate 17.8%
Profit (to 1pt level stakes) -18pts
ROI -7.6%
A loss made which is what I was hoping for, but I had expected a poorer performance. Having said that, Catterick provided a profit of 22.8pts and Beverley 14.8pts, so take that those out of the equation and the picture looks even rosier. I am still confident in my judgment on this and hoped that the non draw bias stats would bear this out.
The courses I have chosen for the non draw biased ones were Chepstow, Leicester, Redcar, Yarmouth and York. These to me are the five fairest courses for sprinters – none have a rail bias and indeed 4 of the 5 tend to race down the centre of the track as either rail offers no advantage. The results for the top rated at these courses were as follows:
Clear Top Rated runners 133
Winners 33
Strike Rate 24.8%
Profit (to 1pt level stakes) +45.9pts
ROI +34.5%
A very positive result, although the data is limited from just the 5 courses. It should be stated that these courses had not been back checked previously and indeed Redcar actually showed a loss of 17%.
At this juncture therefore there is plenty of data pointing to the fact that these ratings will work best on courses without draw bias. This has positive repercussions for the ratings I am devising for longer distances where draw bias is much less prevalent.
I checked through other courses to get a better feel for the data –
two courses that caught my eye namely Windsor and Hamilton. Both these courses
have had strong draw bias in the past but over the past few years the bias
has evened out.
Hence these two courses can be considered pretty "fair"; now. Their
results backed this up –
Hamilton clear top rated runners produced 14 wins from 58 (SR 24.1%) for a profit of 24.1pts (ROI +41.6%); at Windsor the figures were 12 wins from 47 (SR 25.5%) for a profit of 12pts (ROI +25.5%).
This draw based research is on going, but it currently looks like the future ratings for longer distances will be even better without the draw bias conundrum.
Betting Strategy
To date I have two solid profitable betting strategies to exploit.
Each shows a good strike rate, high profit on turnover and long term profits.
These are:
Strategy 1 - At all courses back any clear top rated horse with a rating of 105 or more.
Strike Rate 31.5%
Profit at level stakes SP 85 points
ROI 31.9%
Strategy 2 - At all courses back any clear top rated horse with a rating of 104 with a gap of 3 or more.
Strike Rate 29.5%
Profit at level stakes SP 99 points
ROI 30.5%
Eagle eyed readers will note that occasionally there will be double qualifiers (a qualifier under both systems) – these have a strike rate of nearer 34%.
Full results
If you would like a copy of my spreadsheet detailing historical research
results for Strategies 1 and 2 above you may download it from the link below:
http://www.racingtrends.co.uk/ratings.xls
If you can't read excel files there is a web page copy of the data and charts
at
http://www.racingtrends.co.uk/ratings-results/ratings.htm
(note the tabs of the bottom of the page that permit you to navigate between the sections).
Summary
- I have created a well researched method of rating the highly focused area of 5 and 6f all age handicap races;
- Horses are rated on ten key elements of form;
- The testing phase indicates the top rated runners vastly outperform Racing
Post Ratings and Racing Post Top Speed;
- Simply bet all top rated qualifiers and you would have made a profit at
SP;
- Two advised betting strategies have been devised to maximize strike rate
and profitability;
- These account for factors such as total rating score and how clear rated
a horse is of others in the race;
- Such strategy bets have made excellent past profits over the four year research
period.
- All quoted results are at starting price.
Starting Price is of course conservative compared to real life returns achievable via shopping around for best bookmaker prices or via exchange betting.
Each day I calculate my ratings and seek to identify qualifying selections for the two strategies.
Join now and you can be betting the same educated bets as I do.