![]() |
![]() |
![]() |
|
#20
|
||||
|
||||
|
Quote:
When you bet on more than one horse in the same race, you have increased the chance you will cash a ticket. That lowers your risk. When risk is lowered, you can and should bet more (assuming you are correct that you have an edge). A bettor finding 2 horses with value in the same race should bet more (total) on the 2 horses than he/she would have bet on just one of those horses. That reduced risk effect is dramatic if you use something like Kelly betting. The original poster mentioned horses that were 8-1 and 10-1. He didn't mention his estimate of fair odds, so let's just say 7-1 would be fair odds on each of those horses. Let's also assume that the total bankroll for betting horses is $1000. There's a Kelly calculator at Sportsbookreview.com: http://www.sbrforum.com/Betting+Tool...alculator.aspx According to that tool... If you only bet the 8-1 horse, your optimal bet for bankroll growth is $16, and your expected win is $2. If you only bet the 10-1 horse, your optimal bet for bankroll growth is $37, and your expected win is $14. BUT, if you bet both horses, your optimal bets are $21 on the 8-1 horse and $40 on the 10-1 horse; your expected win is $17, which is more than you'd have from betting either of the horses alone. --Dunbar caveats: 1. Kelly betting results in dramatic up and downs in bankroll. Most serious bettors DO bet proportionally, but they use a fraction (often 1/4 to 1/3) of full Kelly betting. This would apply both to singled bets and dutch bets. 2. I'm assuming the SBR Kelly calculator is correct without totally checking it out. 3. Obviously, if you are betting without an edge, the optimal bet is $0. So how you spend your money on a race is based on how you perceive the entertainment value, not on math.
__________________
Curlin and Hard Spun finish 1,2 in the 2007 BC Classic, demonstrating how competing in all three Triple Crown races ruins a horse for the rest of the year...see avatar photo from REUTERS/Lucas Jackson |