Articles >> horse-racing >>

Research into finding the worst horse racing system ever!

With the recent advent of the betting exchange, punters can play bookie and become layers. Surely, this is the way forward for the shrewd punter – find a selection process that gives the lowest possible strike rate and then lay, lay, lay!


I regularly use Betfair over the internet – for those who do not know, this is a betting exchange where basically you back against other punters. This is useful from a betting point of view, as you are more likely to find mistakes and better prices. A good example of this occurred on 6th Feb 2006 when I backed Monte Major at 79-1 on Betfair, and horse returned a 33-1 winner. I more than doubled by winnings, which at those prices is very handy! I only had £10 on the horse, but taking the Betfair price meant I gained an edge of £460 over SP.


However, perhaps even more importantly, Betfair has opened up a whole new avenue of betting – namely laying horses. Laying horses is when you “play bookmaker” and offer odds on specific horses or sporting events. You set the odds + the maximum stake, and hope that someone else will take the bet. For example, if you think the forecast favourite Bob’s your Uncle is unlikely to win the 3.40 at Southwell, you may wish to “lay” it. You decide to offer the horse at 5-2 for a maximum stake of £10. Assuming somebody takes this bet, there are two scenarios. If the horse loses you win the £10 placed by the backer; if the horse wins you lose £25 (£10 x 2.5).


Now, my idea is not to take on weak favourites, because you are still likely to get “stung” from time to time, but look at horses that have virtually no chance at all. My theory is this – if you can find horses generally offered at around 33-1 that only win 1 in 100 races ~ (1% of the time), then you should be able to offer the potential backer a bigger price, but still make a healthy profit. The reason for offering a bigger price is to hopefully guarantee your bet will be taken.


If you offer 40-1, you may be lucky and get some takers, but I would suggest 50-1 as a price about a 33-1 shot that some people will be willing to back at. Take the following scenario – a horse has a 1% chance of winning – over a series of 100 bets you should win 99 times (+ 99) and lose once (- 50). At 1 point per “lay” / bet, you should expect to make 49 points profit over 100 bets (ROI + 49%). Of course many 33-1 shots trade at bigger than 50s (like Monte Major), but there are plenty that do.


Right – time to find the system! I thought that looking at horses 10th or worse in the betting forecast was a good place to start. As with most of my general research I used data from 1990 to 2005. Here are the results:


Bets 167,469

Wins 3453

Strike rate 2.06%

ROI   - 53.0%


A good starting point, not surprisingly! A strike rate of 2.06% equates to a success rate of 1 win in every 49 races (on average). Thus a price of around 25-1 would be the type of price we would want to offer in order to make a profit of around 50% on investment.




I decided to filter these bets with a forecast price range of 25-1 or 28-1. Thus the system now reads:


1. Forecast rank – 10th or worse.

2. Forecast price 24-1 to 28-1.


Here are the results:


Bets 31308

Wins 522

Strike rate 1.67%

ROI   - 54.0%


Not much change in the overall profit / loss percentage, although the strike rate has been cut further. In recent years there have been roughly 3000 bets a season, so that pans out at around 8 to 9 qualifiers a day.


The stats suggest that qualifying horses will win 1 in every 60 races. Thus we have a margin to work with. Not as good as the 50% I had hoped for, but a margin nonetheless. This means that these horses quoted at 25 and 28-1 can be offered by us, “the layers” at a bigger prices and still make a tidy profit. At these prices, I would suggest a price of around 40-1 in order to interest potential punters. Thus a run of 300 bets offered at 40-1 should provide 295 winning bets and 5 losing ones – working on a point per “lay” / bet, we should therefore make 295 points on the horses that lose (which is what we want), and lose 200 points (40 x 5) on the horses that win. A profit of 95 points would be made, which equates to a profit of 31.7%. Not bad for a simple system! Of course, sometimes we will not get 40-1, and my strategy would be – if you can’t lay at 40-1 or lower, then leave it and wait for another opportunity.


Being greedy - like many punters I imagine - I decided to try and improve upon this. I thought the worst races for outsiders would be non handicaps. Traditionally, non handicaps are better for the head of the market in terms of strike rate; so one would assume the strike rate would be lower at the other end of the scale. Let us see :


1. Forecast rank – 10th or worse.

2. Forecast price 24-1 to 28-1.

3. Non handicaps only


Bets 10373

Wins 125

Strike rate 1.21 %

ROI   -64.7 %


We now improve matters showing that qualifying horses will win 1 in every 82 races. We now have an even bigger margin to work with. We could now offer a price as big as 50-1 in order to interest punters. Thus a run of 820 bets offered at 50 -1 should provide 810 winning bets and 10 losing ones – working on a point per “lay” / bet, we should therefore make 810 points on the horses that lose (which is what we want), and lose 500 points (50 x 10) on the horses that win. A profit of 310 points would be made, which equates to a profit of 37.8%. One problem however, that the exchanges tend to appreciate that outsiders in non handicaps are poor value and hence 50-1 prices will not always be obtained. I would stretch to 60-1 maximum, but no bigger.


My next idea was concerned with horses that have run poorly on their last run. Your last run is clearly the best guide to form and it is difficult to put behind you a bad run in a non handicap. The system now reads as follows:


1. Forecast rank – 10th or worse.

2. Forecast price 24-1 to 28-1.

3. Non handicaps.

4. Finished 6th or worse last time out.


Bets 6703

Wins 66

Strike rate 0.98 %

ROI   - 71.2 %


We now have even better margins to work with. This system pinpoints 1 winner every 102 bets. We could offer these horses at 66-1, and expect a decent return over time.


Lastly, let me include a factor that should be taken into account more when betting – the sex of a horse. Essentially, in horse racing, the female gender is outperformed by the male gender. Also, generally speaking, fillies and mares are more inconsistent. Hence it seems sensible to include a rule that stipulates laying fillies and mares only. The final system now reads:


1. Forecast rank – 10th or worse.

2. Forecast price 24-1 to 28-1.

3. Non handicaps.

4. Finished 7th or worse last time out.

5. Fillies / mares only.


Bets 2941

Wins 20

Strike rate 0.68 %

ROI   - 82.8 %


This gives us one winning horse every 147 races – offering these horses at 66-1 we would expect the following results : 294 bets would give us 292 losing horses (+292 points); 2 losing horses (-132 points).


Thus a profit of 160 points would be achieved, which equates to a return on investment of 54.4%.


Clearly, a losing bet (winning horse) will be serious blow to the wallet – but over time, this looks one of the more profitable betting options in horse racing today. You can expect around 200 qualifying bets a year.


Having tried this idea on pencil and paper over recent months, I have found that you need to offer these prices the night before racing to small stakes only. This is because some punters still “rush” in early, and take what looks like an “over the odds bet”. I have found that if you offer those types of prices nearer the time of the race, much bigger odds are on offer and hence the bet will rarely be matched. This is especially true for non handicaps, which is where this final “system” focussed on.



Free Registration
Free Horse Racing System
Racing Research
Racing Report Vault
Flat Racing Trainer Report
Horse Racing Research Articles

Copyright 2022