Picking winners is easy - finding value winners is not.

Value Betting

22nd Apr 2001
Article is over 10 years old. Stats may have changed since.

If you want a system that wins over 50% of the time then just bet all odds on horses - but if you do you will not make a profit.

If you want to find big priced winners then just bet all 33-1 horses - but if you do the number of winners will be few and far between and again you will not make a profit.

In order to profit from racing you have to find not just winners but value winners. But what exactly is value?

The question of value is one that can not be answered easily. Where one person may find value another may not. One person may say an 11-8 favourite is value and should be odds on, another may say it is not value and it should be 2-1. This is the mechanism which makes betting profitable for some: finding value where others have the totally opposite view.

It does not matter if it is the price of a horse or the price of fish you are trying to obtain value for. Value is value and is as equally important to a punter as to a shrewd shopper.

False Favourites

This week the long term 2,000 Guineas favourite - Nayef - got turned over at odds of 4-6 in the Macau Jockey Club Craven Stakes. Some punters thought this was a price worth paying and backed him for that race. After all, he is one of this seasons talking horses after looking impressive at Ascot last back end. But using the sire statistics it could be seen that this was a poor price to take: for stats punters value could be had by opposing Nayef.

Hype can cloud your judgement and force you into betting a horse which offers no value and this looks to have happened with Nayef.

The same thing happened with another hot 2,000 Guineas contender - Tamburlaine - in the Lane's End Greenham Stakes at Newbury. Punters, thinking this horse is a Guineas market leader were overbetting him too and sent him off as the 7-4 favourite.

How can you tell if this is a good price or not? How much would he have to be to be a value bet?

Using stats it was possible to calculate that he should have been more like a 100-30 chance and clearly not a good betting proposition.

Creating A Betting Forecast

"Oddsline", "morning line", "tissue". These are all terms used to describe a betting forecast. Every newspaper racecard will have one and bookies will create their own or use the Racing Post as a guide.

If you create your own betting forecast you can see which horses are intune with the market and which are not. If you can identify a 5-1 horse that is showing 10-1 in the market you have found a value bet, or an overlay. If you can identify a 3-1 horse that is showing 13-8 in the market you have found a bad value bet that should be opposed.

Creating your own betting forecast is easy. All you have to do is use some kind of numeric rating to rate the chance of each horse in a race. Using these figures you can convert the chances into true probabilities, and as probabilities is a function of odds, you can convert the probabilities to odds.

How to Create a Basic Betting Forecast

  • Assign a numeric rating for each horse in the race
  • Total up all the ratings
  • Divide each rating by the total to obtain a probability
  • Convert each probability into odds

First, you have to use some kind of numeric rating to rate each horse. Here we will show you how to assign a rating from 1 to 10 using sire stats. This method works well in 2yo and 3yo stakes races where the horses have not had enough runs to create their own horse profiles.

A bit of good judgement is required which you will get better at after analysing a few races. All you have to do is look at the sires overall strike rate, divided it by 3, then adjust the value up or down depending if the stats look good or bad. Using your judgement you will assigne each horse a rating somewhere between 1 and 10. Here are some examples.

2.00 Newbury 21-Apr-01

Dubai Duty Free Millennium Millionaire Conditions Stakes
£ 9,282 3yo Conditions Stakes Class B 10f Good to Soft

Horse Rating Probability Odds RP Fcast
Welsh Border 8 0.308 9-4 2-1
Archduke Ferdinand 5 0.192 4-1 9-4
Forest Tune 3 0.115 15-2 6-1
Golden Wells 6 0.231 100-30 6-1
Regatta Point 4 0.154 11-2 7-2

First we looked at Welsh Border. He is sired by Zafonic who has an overall strike rate of 17%. Divide this by 3 to get 6 (rounded up).

Now we look at each of Zafonics stats to see if this is a race he is suited too. If a stat is a few points above 17% (his overall strike rate) we will increase his score by 1, if a stat is a few points below 17% we will decrease his score by 1. Any stat which is very strong positive or negative could be adjusted by 2.

Season Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
Spring 17 93 32 18% 34% 1.05 22% 60404-00601

No problem with the time of year - rating + 0=6

Going Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
Good to Soft 11 69 25 16% 36% 0.91 -39% 010030033-0

The going is fine - rating + 0=6

Course Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
Newbury 6 30 12 20% 40% 1.15 87% 3-143114601-

Shape Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
Oval 40 189 71 21% 38% 1.21 4% 460120011-6

Direction Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
Left 25 126 44 20% 35% 1.14 -31% 12150130-60

Surface Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
Flat 41 204 65 20% 32% 1.15 4% 1100000-001

Speed Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
Galloping 58 339 117 17% 35% 0.98 -22% 0001300-001

His runners do well at Newburys oval, flat, left handed track - rating + 1=7

Race Class Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
B 1 34 7 3% 21% 0.17 -96% -4023240300-

Bad stats for the race class - rating - 1=6

Distance Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
10.0 14 71 24 20% 34% 1.13 -21% 140120130-0

Good at the distance - rating + 1=7

Last Ran Wins Runs Plcs Win% Plc% I.V. ROI%  Last 10 Runs
57+ days 32 161 65 20% 40% 1.14 11% 010030-0001

Good after a layoff of 57+ days - rating + 1=8

Next we look at Archduke Ferdinand. He is sired Dernier Emperor who has only ever had 3 runs on turf at time of writing. It is not possible to calculate a rating for this one so we will assign him a middle rating of 5 (you could look at the trainer, jockey or combo stats to obtain a rating if you wish).

Forest Tune is sired by Charnwood Forest who has an overall strike rate of 14%. His starting rating is 14 / 3=5. Now we adjust his rating depending on wether the individual stats for Charnwood Forest are good or bad.

The sample size of 10 for the Spring is
too low so well ignore the 30% strike rate - rating + 0=5
Going stat for Good to Soft is bad - rating - 1=4
Stats for a recent run are bad - rating - 1=3

Do this for each runner until they all have a numeric rating between 1 and 10. It will take some practice to get used to but it does not matter if your judgement is not all that accurate. As long as you are consistent with your judgement in increasing or decreasing a rating after looking at each stat the final calculation will still hold true.

Now you have assigned each runner a rating you need to add all the ratings up to find the total.

8 + 5 + 3 + 6 + 4 = 26

Next divide each rating by the total to obtain the probability.

e.g. Welsh Borders probability is 8 / 26 = 0.308
Forest Tunes probability is 3 / 26 = 0.115

next you can convert the probabilities into odds

odds = (1 / probability) - 1

Welsh Borders odds are ( 1 / 0.307 ) - 1 = 2.25 or 9-4

Forest Tunes odds are ( 1 / 0.115 ) - 1 = 7.70 or 15-2 (rounded to the nearest bookies odds)

After we have worked out all the odds we can compare against what the bookies are offering (or the night before against the Racing Posts betting forecast). In this example it is clear that Welsh Border has the best sire ratings but should we bet on him?

Our forecast says he should be a 9-4 chance and as the Racing Post forecast him to go off at 2-1 there is no value there, but he is not necessarily a bad bet.

Archduke Ferdinand, whom we could not accurately rate, we make a 4-1 shot so looks bad value at the Racing Post odds of 9-4.

We make the complete outsider to be Forest Tune. He should be a 15-2 chance, yet is 6-1. A bad bet.

Golden Wells is second rated on 6 points and looks good value. We calculate his odds to be 100-30 but he is forecast to be 6-1. This is an overlay and should be bet.

Regatta Point should be 11-2 and is 7-2 in the Racing Post.


In this race we like the high 8 rating for Welsh Border but at 2-1 his odds are borderline. If he were to drift in the market we would bet on him. The second top rated on 6 is Golden Wells and is much more of an attractive overlay. It would be a good tactic to use a dutching method to bet on the both of them, or even bet the Exacta. Remember, we have found an overlay here, if his SP is higher than we expect the Exact pool should be higher too.


Golden Wells wins at a backed down price of 7-2. Welsh Border finishes 2nd at odds of 100-30. The Exacta paid £13.30.

Position Horse SP Rating Forecast
1st Golden Wells 7-2 6 100-30
2nd Welsh Border 100-30 8 9-4
3rd Archduke Ferdinand 2-1f 5 4-1
4th Regatta Point 9-2 4 11-2
5th Forest Tune 7-1 3 15-2

The next page shows the analysis of the Fred Darling Stakes and the Greenham Stakes.

Comments on Value Betting

Add Comment or Ask Question

Show random avatar image with my comment

Which is the Royal Racecourse:
Ascot, Newmarket, York ?

Big Priced Winner! Top Rated Dialogue Wins 11/1 | Subscriptions just £25 a month or see special offers  Subscribe Today!