A/E Value Index
The A/E index is an enhanced impact value type stat which can also be used instead of the ROI stat. The A stands for Actual whilst the E stands for Expected and thus the stat shows the index of actual winners to expected winners. Calculating an index this way shows 'value' and thus value bets, and value lays can be easily identified.
If a stat has 'more winners than expected' then we have found a good thing for backing. If a stat has 'less winners than expected' then we've found a weakness - an indicator of bad 'value' for betting purposes.
To calculate the A/E index you need to know the actual number of winners and the expected number of winners. The first part is easy as this is the number who actually won. Calculating the expected number of winners is a bit more tricky as you need to sum up the odds of all the runners.
First, calculate the probability or 'odds' of winning. This is not the odds of the horse such as 5/1 but the odds chance of 5/1. Use the following formula to work out the odds chance:
Odds of Winning = 1 / (Price + 1)
If a horse is Evens then his odds chance (or probability) of winning = 1 / (1 + 1) = 0.5
If a horse is 3/1 then his odds chance of winning = 1 / (3 + 1) = 0.25
Once you have calculated the odds of each runner you then add all those figures up to obtain the expected number of winners.
e.g. If we have 100 horses who went off at 3/1 then we would expect 25 of them to win. The odds of a 3/1 shot is 0.25 so adding all 100 of those together = 25.
Here are some real examples using information obtained from the Flatstats Favourite Stats Reports.
Flat Turf 2yo Maiden Stakes Class 4+5 (D-E)
| In the past 10 years there were 4407 races The actual number of favs who won was 1741 Adding the odds of all favs gives an expected figure of 1832 We expected 1832 of favs to win. 1741 of them actually did. |
| A/E index = 1741 / 1832 = 0.95 |
That is quite a good figure. The favs in this type of race nearly win as often as expected to. In fact, because of the slight error due to bookie over-round, these favs are probably winning 'as expected'.
Flat Turf Amateur Races
| In the past 10 years there were 350 races The actual number of favs who won was 59 Adding the odds of all favs gives an expected figure of 88 We expected 88 of those favs to win. 59 of them actually did. |
| A/E index = 59 / 88 = 0.67 |
This time the A/E index is much lower and indicates that Amateur races in general are bad for favourite backers. In this type of race the favourite does not win as often as expected to and thus this is a good race for laying the favourite.
| A/E Index Stat Summary |
| Pros: Useful for determining 'value', finding underbet horses or false favourites etc. |
| Cons: Bookies over round can skew figures but if you are comparing like for like (such as determining if one jockey is better value than another) then this is not a problem. |
Ten Year Trends
Ten Year Trend type stats are very popular with mainstream TV pundits. They like to state the number of winning favourites, or the strike rate of the favourite for the past 10 runnings of a particular race. Whilst it is sometimes fine to look at favourite stats for major races, many punters look at and rely on the 10 year favourite stats for minor races such as a seller at Catterick.
There are many systems based on following 'key races' where favourites have shown good or bad form. Relying on those stats for minor races is futile as there is no significance with the past 10 running's of minor races such as sellers at Catterick.
The reason being is that the sample size is too small and the timescale too large for that small sample. If punters want to analyse the fate of the fav for a particular race then they would be better off analysing all similar races to that one in question.
e.g. There have been 90 selling races at Catterick in the past 10 years. The 'fate of the favourite' in all those races is:
Catterick Selling Races
| Favourites: 88 Winners: 31 Strike Rate: 34% £1 Level Profit: £7.12 ROI: 8% |
This is a much more reliable stat than the 10 year trends favourite stat for the individual race. Punters checking their daily newspaper 'fate of the fav' stat will only see 10 previous races and that is too little information to be reliable.
The only time a ten year trend fav stat is useful is with the big races such as The Derby, Lincoln, Tote Ebor etc. Those races attract so much money, and a lot of punter attention. This makes the 10 year stat for the individual race a bit more reliable.
The 10 year trend type stat is useful for analysing the genders, ages, and weight ranges. This information will have a bigger sample size because there will be a lot more information to analyse. There can only be one favourite in the race, but there can be many colts, 3yo's, or horses carrying 9st 4lbs or more in the race.
Lincoln Handicap Ten Year Trends
| Colts 4 wins from 40 runs Weight up to 8st 7lbs 0 wins from 65 Non Distance winners 1 win from 81 runs |
Even though there are only 10 previous runnings of the Lincoln Handicap there are many runners of a particular group in the race and this gives a more reliable sample size. What stands out is that the low weighted horses (up to 8st 7lbs) have not won within the past 10 years despite 65 horses running from that weight range. Non distance winners do not do so well either. Out of 81 horses which had not won any race at the same distance as the Lincoln only one of them won the race.
| Ten Year Trends Stat Summary |
| Pros: Useful for analysing the major races of the year. |
| Cons: Useless when analysing minor day to day races. Best to analyse all similar races to get a bigger sample size. |
Summary
More and more punters are using stats for finding winners. Unfortunately many of them are using the wrong kind of stats or do not know how to use stats properly.
The biggest mistake punters make is relying on too low a sample size. If a punter says that "A Jockey has had 3 wins from 4 rides - a 75% strike rate" then do not rely on such a small sample.
The second biggest mistake they make is not taking into consideration the profitability or value of a stat. If you hear a punter say "Sir Michael Stoute has an 18% strike rate with his 2yo's" then disregard it. That strike rate stat does not tell you if backing all his runners has returned a profit or a loss.
The most useful stat to have on your side is the A/E index. This stat takes into consideration the odds of all previous runners and thus is a very good indicator of 'valueness'.
You will not see the A/E stat in any newspaper or mentioned by media pundits. Newspapers will only print bland stats and media pundits prefer to dish out soundbite stats.
Flatstats displays A/E stats for every horse, trainer, sire, jockey, favourite etc. Knowing this information will give you a big edge over other punters.
Related Articles and Resources
Article created 01-Nov-08. Stats may have changed since. Data analysed from Nov-98 to Nov-08
