| |
Racing Forum
One last time, using your own numbers
Posted By: Dunbar
Date: 20 May 02, 2:23 pm
Dunbar, if a handicapper is betting 2-1 shots and applies optimal betting, he is only betting 2-1 shots because he is a 40% win capper. Optimal betting would mean your 1 unit example for the 2-1 shots represents 30%, and on the 20-1 shots the optimal bet would be 3%. If he loses all three of the 20-1 shots in your example and wins 4 of the 2-1 shots as in your example, that is a net loss. A reduced net loss from flat betting, but still a net loss.
Gary, I'm with you all the way until the last 2 sentences. If I bet 30% of a unit on a 2-1 horse, and 3% of a unit on a 20-1 horse, then I will have a win of (4 * 2 * .3) - (6 * .3) = 0.6 units on my 2-1 bets, and a loss of 3 * 0.03 on my 20-1 bets. Please note that 0.6-0.03 is +0.57 units. Thus, what would have been a net loss with flat bets has been turned into a net win by using proportional betting, contrary to what you said above.
Flat betting has always been, and will always be, the measure to determine if a system or individual is successful.
Flat betting is fine when you are betting on outcomes that have approximately the same odds. It is terribly wrong to flatbet when you are mixing 2-1 bets with 20-1 bets. It is wrong from a risk standpoint, and it is wrong from a testing-a-system standpoint.
No optimal betting can turn flat bet losses into optimal bet profits, it can only reduce flat bet losses.
Please see example above.
And, I would think Wong, or any other professional would agree.
No, Wong does not advocate flat betting. Here is a quote from his book, Betting Cheap Claimers: In a section titled "How Much To Bet" Wong writes, "Bet an amount equal to edge times bankroll divided by win odds." Note that when you divide by win odds, you get a very big difference between what you bet on 2-1 horses and what you bet on 20-1 horses.
Here is what SW writes about testing a system: In a section on Bet Size, SW writes, "You must also decide how much to bet on a horse. Here is a suggestion. Instead of betting (or pretend-betting) the same amount on every horse, bet an amount such that you will collect the same amount on each horse that wins."
Note that this suggestion, too, results in a very big difference between how much you bet on 2-1 horses and 20-1 horses. Using Wong's approach, I would pick a number like 1000, and bet $333 on 2-1 horses. My corresponding bet on 20-1 horses would be $47.60. I am betting 7 times as much on a 2-1 horse as a 20-1 horse.
BTW, this 7-1 relationship is correct, not the 10-1 I used earlier. I'm not going to go into why. Suffice it to say that Wong does NOT advocate flat-betting.
One last thing, and I do mean last, because I hope I can refrain from any further posts on this. Using a 7-fold difference between bets on 2-1 horses and 20-1 horses would still turn an flatbet net loss into a net win in the original scenario I presented--winning 4 out of 10 bets at 2-1 but losing 3 bets at 20-1. Compare these 2 approaches:
1. Flat bet $70 on each race.
Result: win 4*2*70, lose 9*70 = -$70.
2. Bet $70 on each 2-1 horse and $10 on each 20-1 horse (as suggested by Wong).
Result: win 4*2*70, lose 6*70 + 3*10 = $560-$450 = +$110.
If that isn't turning a net loss into a net win, then I have to give up.
Proportional betting is better for both profit growth and for analyzing results.
Done, Dunbar
- One last time, using your own numbers -- Dunbar -- 20 May 02, 2:23 pm
| |
Racing Forum is maintained by Pi Yee Press