Probabilities and Value (1440 Views)
Posted by:
Catalin (IP Logged)
Date: June 20, 2002 09:23PM
Much has been made on both boards lately about probabilities (i.e. “true odds”) and value based betting. A prime example was Saturday’s 9th at Churchill where Spain was bet down to 2-1 second favoritism. Having seen both TG and Ragozin for Saturday’s Churchill card, I have no problem going on record that Spain’s true chances of winning were far short of the 1 in 3 being offered. If memory serves she was no better than co-second or third fastest going in, was giving weight to the two ML favorites, and had drawn a post that guaranteed at least a length of ground loss. Despite a pretty good pattern on both sheets, she looked at least to me no better than 3rd most likely to win. I believe I recall Mall stating that he pegged her at upwards of 9-2. Soupy, he of specious reasoning makes the argument that the 2-1 looked pretty good when she won. EVERY winner looks like a good bet AFTER they cross the line. Unfortunately most of us are unable to get a bet down once they break the tape. The fact of the matter is that if we ignore true odds, the law of large numbers will quickly whittle away our bankroll, no matter how good we think we are at picking winners.
Consider the following example. Suppose I were to tell you ahead of time that the coin I was flipping was weighted such that it came up heads 70% of the time, and tails only 30% of the time. Now suppose I established a book on the outcome and offered the following odds:
Heads 1-5
Tails 9-2
What’s your wager? Is it the favorite (Heads) who looks far more likely on paper than Tails, or is it the 9-2 second choice? Anyone choosing heads is doomed for long run failure. As n approaches infinity where n is the number of trials, coin flips, races, etc., you would wager
$1 on Heads and receive 1-5 or (1.2) * .7 (probability of heads) or only 84 cents for every $1 wagered.
On the other hand while Tails wins far less frequently, the bettor would expect to receive a payout of 5.5*.3 or $1.65 for every $1 wagered. So now who do you like? This is may be a pretty simplistic example, but the concept isn’t any different from the Saturday’s 9th from CD re Spain.
Betting horses that pay below their true probabilities will ALWAYS result in a long run loss. That’s why success is predicated on your ability to accurately estimate true probabilities. Those who bet selections rather than PRICES have NO SHOT whether you’re betting the horses, wagering on a coin flip, or playing the biggest crapshoot of all, the stock market.
Sorry, only registered users may post in this forum.