by Bob Pitlak
When I began handicapping thoroughbreds, back in the 1970's, a cardinal rule often quoted was something like "never wager on a horse that hasn't raced within 21 days." While working on an upgrade of my HDCP-12 program, I decided to see if such a rule, to paraphrase Joe Pesci in the "My Cousin Vinny" movie, could "hold water."
To test this, I studied over 1800 racing cards from American tracks between Jan 1, 2013 and May 18, 2013. I checked the last race of each horse in all races on all cards, more than 148,000 PP lines. For each, I recorded the number of days since the horses' previous races. And, if the horse won the race, I recorded that fact along with the resulting payoff.
Initially, I was surprised to find that the average win occurred 44.3 days since the previous race. This seemed like a very long time, and it actually was. The distribution was badly skewed by small numbers of winners that ran after unusually long days away from the track. Although I stopped graphing data after 125 days since the last race, there were, in fact, winners that had not raced in over two years! These "outliers" badly skew the average. Look at it this way, a single horse that won after a 2 year layoff balances out 52 horses that won after 14 day layoffs! If you ignore these, it was clear that most winners ran within 50 days of their last race.
In fact, for the winners, the median number of days since last race was 26. Basically, this means that there were about as many winners with less than 26 days rest as there were with more than 26 days rest. And, the mode, the single number of days off that occurred most frequently was 14 days. There were 1,027 winners that ran after exactly 14 days rest.
In terms of the "cardinal rule" cited above, not quite 60 percent of the wins were recorded after more than 21 days since the last race. Clearly, the advice, "never wager on a horse that hasn't raced within 21 days" doesn't hold water
How about return on investment? This was the most interesting result of this study. Of course, with no handicapping involved, one would expect to show a negative return. That was the result in the 148,000 racing lines that I examined. However, while it was a bit erratic, the data shows a definite trend for the R.O.I. to improve with an increase in the number of days since the last race. Why is this? It may be as simple as the fact that the public is over-betting on recency!
Click a Link to Explore: NFL | Thoroughbred Racing | Poker/Lotto/AutoRacing | Useful Links | Free Link