So while I anxiously await Saturday's PP's I thought I might give a little update on the Handicapping the Handicapper project. Since the onset of this project, I have handicapped 161 races of various types. The table below shows the performance of my top 4 contenders.
The table shows that my top contender won nearly 30% of the races I handicapped. The 2nd almost 17%, 3rd 18% and 4th 14.5%. The reason each contender does not equal 161 is because there are occasionally races where I could not come up with 4, 3, or even 2 contenders I could make a legitimate case for (these races do not include any in which Rachel Alexandra or Zenyatta participated). These percentages, or probabilities of the contenders winning translates to the odds shown in the next to last column on the right. These odds indicate what my win expectancy for each contender level. The final column is my win expectancy + 50%. The extra 50% allows for an extra margin of error. Thus, the minimum odds I am able to take at this time is 7/2 and I require 9/1 for my Number 4 contender. It is easy to see that I have not been playing too many favorites to win.
In the original article, I suggested that by dividing races in categories such as Maiden Claiming, Maiden races, claiming and so on, one might be able to better refine the personal odds requirement for each type. In addition, I suggested the use of a moving average of 30 races for each type of race. Please check the archives for the original post, "Handicapping the Handicapper" for the details.
To build a quickie database to test while I was accumulating my personal data, I kept a record from races run at Calder Race Course. I discovered that 30 races were an insufficient base to work from as the odds didn't make sense in some cases. I expanded the number to 50 races, and found that at 50, the results made more sense in that the top choice of the collective betting public performed better than the 2nd, than the 3rd, than the 4th.
To date, I have not handicapped any single class of races 50 times yet. The largest category is graded stakes races with 26. This is because I have been trying to concentrate on better quality races over the Summer. In addition, I have been trying to become a little less parochial in my handicapping habits, so lack of personal time requires me to be more selective about which races I choose to handicap and play.
So for now I will continue to use the generalized table above although I think it is time to start a new one. As the number of races increases, the table becomes less responsive to my current handicapping prowess.
The 161 races yielded a total of 66 win or exacta wagers made. 42 win wagers and 70 exacta wagers on a total of 38 races. So using this method of creating a contender odds line has certainly had the desired effect of cutting the number of races played.
Of the 42 win wagers made a paltry 6 were winners for a 14.3% win rate. At an average 6.88-1 odds for winning wagers this resulted in $0.13 profit for each dollar wagered. Note this is not a $2.00 flat bet result as my wager size may vary depending on whether or not I'm including an exacta wager on the race. Also there are couple of winning wagers where the winner was obviously superior to the field and the odds sufficiently tasty that I couldn't resist and abandoned my system and played the low odds overlay (yes, there can be low odds overlays). Removing these from the database would have the effect of lowering the win rate, but the average winning odds would increase.
Surprisingly, the real money has been made playing the exactas. Of the 70 exacta wagers (2 to 6 individual exacta wagers per race) made 5 resulted in cashing a ticket for a 7.1% success rate. However, at an average return of 39.4-1 this results in a return of $2.01 for each dollar wagered. No confessions are required for this figure as all exacta wagers conform to the template of playing my overlay horse underneath the other contenders in the race. Seemingly, a very effective way of playing the races if one is able to contend with the inevitable losing streaks that are going occur in between the winners. For me the bottom line is am I showing a profit. The results here indicate that this method of setting acceptable odds is working.
A word about the exacta. I have found through experience that if my overlay horse is at odds of 5-1 or less, the payoff on the exacta is not usually worth the risk, especially if the other contenders are at lower odds. In this case, I just increase the win bet and leave out the exacta. Also, placing an exacta bet on such horses effectively lowers the odds on the overlay horse in such a way that should the horse win the extra money required for the exacta has destroyed the overlay price.
At this point, I am confident that once I am able to use odds for specific types of races, the method will really fine tune the odds I am able to accept on a race. For instance, after 26 Graded Stakes races, the required odds for the top choice is 5-1 based on a top contender success rate of 23%, but overnight stakes would require only 3-2 based on a 50% top rate after 20 races. In my opinion, this is where the real value of this exercise lays.
In conclusion, the purpose of Handicapping the Handicapper is to find an objective method of establishing an odds line for top contenders in the race. While there is still more work to be done before declaring the method successful, the interim results reported here show great promise.
A final note. It seems unlikely that I will be able to provide Public Handicapper selections for this weekend, so you're on your own.
Til Next time. Good Luck!
