A Guide To Betting Trends & Stats
As Disraeli said “There are lies, damned lies and statistics.” Statistics is the collection and analysis of numerical data. Trends are movements in the general direction.
If you toss 10 coins and get HTHTHTTTTT (H= heads T= tails) then statistically 70% of heads won and the trend is for tails to win. Neither of these is any use for predicting the result of the next toss, with it equally likely to be a Head or a Tail.
Trends are indicative of change, such as if Trap 1 wins the first four greyhound races at Walthamstow this might be because a sudden downpour has flooded half of the track, or if a football team keep 5 clean sheets at the start of the season it might be related to the skills of a new goalkeeper.
Beware of “expected trends” however. A widely voiced expected trend is that because the Grand National fences have been made “easier” then more favourites should win than did in past years. Two joint favourites have won in the last 10 Nationals, accompanied by other winners at 100/1; 66/1 and 2 at 33/1.
Statistics can be useful to the punter, but only with a little appreciation of what they are. The first thing to recognize is that you can soon try to check too many statistical variables: horse; trainer; jockey; course; going; draw; weight etc and either give yourself a headache, or increase the confusion.
The value of statistics lies purely in the sample size. Since betting racing began, that one-third of favourites win is true. The number of races that this is based on, (i.e. every race ran) validates the statistic. It doesn’t mean that two of today’s favourites will win the 6 race card at Chepstow. None of them might win, or they all might win. It does mean however that it is likely that one third of favourites have won all the races at Chepstow since it was opened in 1926, why, because it is a massive sample size. A similar football statistics would be that home teams have the advantage.
Another statistic validated in time and proven in the UK; Ireland; Australia; the States; S Africa and South America, is that the chances of a horse winning are reflected in its price, with an advantage the shorter it is priced. If you bet every horse over time that started at Evens, you would have lost less money than if you bet every horse at 6/4, which would have lost less money that every horse at 5/2, and so on. You would have still lost money, but you can rely on the statistic continuing.
So where does this leave us with statistics and betting? The past is no guide to the future but there are certain statistics that hold true over time because of large samples. Therefore statistics that say “Jockey A riding for Trainer B has a 75% hit rate in handicaps at X racecourse over the last 2 seasons”are meaningless if the two only coincided on four occasions. The sample size is too small and the winners might have all been on the same horse. A useful statistic based on a large sample size is draw bias. Particularly at courses like Chester and Beverley, horses drawn low have a tremendous statistical advantage. Having reviewed this statistic before striking your bet at Chester you may choose to check your selection’s draw and increase or decrease the size of your bet because of this statistic.
Useful stats to consider in horse racing are:
Ten year record of favourites in the race (particularly good at Sandown “military” meetings, and festivals like Punchestown and Cheltenham.)
Horse form (Consistency; course distance and going preferences)
Racecourse (Draw bias, trainer record)
and in football are:
Home bias and competition record (for that team)
Always remember that the larger the sample size the more reliable the statistic.