Do You Have A +EV Methodology?
 

Reading some of the recent posts on this board, there seems to be some confusion between picking winners and winning $ (+EV). If you want to pick winners, just bet the favorite every race: these days, especially at tracks with small fields, you will average close to 40% winners. Bet horses going off 7/5 or lower, and you will average better than 40% winners. Bet horses going off less than 1:1 (even money) and you will average better than 50% winners! No handicapping needed, just watch the tote...


Unfortunately, picking these high probable winners won't actually increase your bottom line, although you will lose at something less than the roughly 20% takeout.


Of course, no one is actually advocating betting on chalk these days, but what if your methodology seems to pick mostly horses that pay less than $10? I would propose that you need to revise your methodology, since playing horses going off less than 4-1 puts you in direct competition with the large rebate players who will pound these horses in the pari-mutuel pool. Unless you are getting rebates of 8% or higher, it will be rare that any of these win bets offer +EV. The focus in these races will have to be to avoid the win pool, and focus on horizonal and/or vertical bets that combine these chalky horses with non-chalk. Of course, this requires other (and possibly new) handicapping skills such as finding the place horse, or classifying a race favorite as legit or vulnerable.


What about win bets? I propose that you test your methodology by a sample that contains only horses that pay $10 or more in races that you bet on. If, in this sample (30 races or more), you find you have a +EV, then you can conclude you have a winning methodology for betting in the win hole. If you are not making $ betting these races, you either need to adjust your methodology or pass on making win bets. Remember, when a horse pays $10 or more to win, you are unlikely to be sharing the win pool with the large rebate syndicates, and likely have +EV.


Bottom line, earning +EV from betting horses involves both the win percentage and the mutual payoff. Just be aware that they are negatively correlated, which can affect one's psyche when playing this game. In addition, your variance will be higher when avoiding favorites in the win hole. So why torture yourself win betting horses going off 4-1 or higher? Because, as the bank robber Willie Sutton once said when asked why he robbed banks, "That's where the money is..."


Richard

So basically, if you (or RX3 on RDSS) rate a runner at say a 30% chance and the expected win mutuel (using the live tote) is 3-1 ($8), your EV for the single Win wager would be positive because the $Net would be +0.30. Am I on the right track with this?

Use the formula EV = (Odds +1) x (Probability Estimate). Using your example, EV = (3+1) x (0.30), or 1.20. An EV =1.00 is breakeven, so you have +EV in your example.

Richard

That’s a lot easier, doing the $1 EV than the $2 one... proof positive is that I messed my outcome up (I posted +0.30, should have been +0.40 for a $2 EV).

What about calculating EV on multiple contenders? Say I narrow the field down to 3 horses, and Horse A has a 30% chance at 3-1, Horse B a 20% chance at 7/2 and Horse C a 12% chance at 8-1? If I am to bet two out of those three to Win, how would I do so?

Horse A shows a +EV of 1.20

Horse B shows a -EV of 0.90

Hose C shows a +EV of 1.08

Horses A and C are +EV

The optimal bet is to use a group overlay calculation of the Kelly Criterion which is beyond the scope of this discussion -- best to research it on your own depending on your math skills.

Next best would be to bet on both A and C, in proportion to their odds (assuming they don't change after the last flash before the gates open) -- that is bet 2.67 units on A and 1.00 units on C.

Or, you could use bastardized Kelly, which purports to bet your edge divided by your odds (again, this assumes you REALLY know your edge AND your odds): bet 0.067 of your bankroll on A, and 0.01 of your bankroll on C. Most will use half, quarter, or even one-tenth of Kelly due the bankroll swings.

Hope this help.

Richard

Group Overlay Method: http://www.cynthiapublishing.com/Mit...dm20001104.htm

Developed by Dick Mitchell in several of his books, particularly Commonsense Betting, pages 101 - 104 (excerpted at the above link).

Has long been something I want to incorporate into RDSS for multi-horse dutch betting. Some day!

Ted