[DvdBonan]

How do you calculate expected value in a game and how does a bettor use it to his or her advantage

[suicidekings]

EV = (P[win] * amount to win) - (P[loss] * amount wagered)

Example:
Your projections show a 55% chance that a team will win a game in the NBA, which would correspond with a theoretical ML of about -130 (-1.5 fave), but the line opens up showing that team as a PK with a corresponding ML of -110.

EV = (0.55 * $100) - (0.45 * $110) = $5.50

Placing a wager at -110 when you feel it should be -130 means you're getting a better deal. To take advantage of this, you would need to project lines for each game and compare them to the opening lines. When you see discrepancies between your lines and the opening lines, there may be an opportunity to profit. However, this depends heavily on your ability to produce quality lines for games. If your work is off, you'll see a lot of "opportunities" that could lose you a lot of money.

Conversely, if you wagered on a team at -130 when it's true value was closer to -110 (a PK, with a 50% chance of winning), then:

EV = (0.50 * 100) - (0.50 * -130) = -$15

It's like buying anything.. You can buy the same product (probability of winning) at different places and different times (ie: For $25 at one store, or for $15 at a different one). You'll be happy with the product 80% of the time no matter where you buy it, but when it does fail you (1 out of 5 times), you're out more money if you bought it at the first place instead of the second.

[suicidekings]

An example from tomorrow's NBA games (2009-11-10). There's two ways to use the expected value of a wager to help you profit:

Beating the Closing Line
Houston Rockets opened up at +7.5 (-110), and early action was placed on the game such that the books adjusted the line down to HOU +6 (-110). You can use the SBR half point calculator to see what the actual value in this change was. It turns out that buying HOU +6 (-110) is equivalent to buying HOU +7.5 (-143). If the line closes at HOU +6 (-110), the early buyer would have beaten the closing line by 33 cents. If two players lay the same wager amount for +7.5 (-143), buying 1.5 points, and at +7.5 (-110), the early buyer obviously stands to receive a greater return if the team wins, per unit bet.

Line Shopping
If, instead of buying the 1.5 points to match the +7.5 at the higher price, the bettor decides that he still likes the bet at +6, he can still maximize his profits by the same concept by having accounts at several different books. As I write this, the line is available at The Greek at HOU +6 (-110). However, if the bettor has an account at Pinnacle, he can get the same HOU +6, but for only (-102).

The bottom line is it's all about getting the best price you can and having a good understanding of how to compare the value of different spreads and prices.

[IrishTim]

Good answer to a ridiculously general question. The think tank here is what separates this forum from all the others.

[Wrecktangle]

I'm with Irish here...nice concise answer from suicide.

Dvd, you ought to dump your points on him.

[patswin]

good explanation very concise

[talnted]

bump for the good post on explaining EV.

[Nuggz]

Thanks for the bump, I posted a thread with a similar question. Not about calculating EV and using it, but about finding a win% in a discrepancy between lines.

Say I calculate Boston to be +4.5 dogs, and the books have them +6, how can I calculate how much those 1.5 points increase my win% over 50%?

[Justin7]

Thanks for the bump, I posted a thread with a similar question. Not about calculating EV and using it, but about finding a win% in a discrepancy between lines.

Say I calculate Boston to be +4.5 dogs, and the books have them +6, how can I calculate how much those 1.5 points increase my win% over 50%?

If 4.5 is the "fair" line, you need to know what the push value of the 5 and 6 are. If each of these hit 4% (they do not, this is just an example), your wager at +6 would hit at 50% + 4% (for crossing the 5) + 2% (for capturing half the "6") for an effective win rate of 56%. This is a monster edge.

[Nuggz]

Understood. I knew the answer wasn't complicated I was just having trouble figuring it out. Makes sense now, thanks.

[hustla]

EV = (P[win] * amount to win) - (P[loss] * amount wagered)

Example:
Your projections show a 55% chance that a team will win a game in the NBA, which would correspond with a theoretical ML of about -130 (-1.5 fave), but the line opens up showing that team as a PK with a corresponding ML of -110.

EV = (0.55 * $100) - (0.45 * $110) = $5.50

Placing a wager at -110 when you feel it should be -130 means you're getting a better deal. To take advantage of this, you would need to project lines for each game and compare them to the opening lines. When you see discrepancies between your lines and the opening lines, there may be an opportunity to profit. However, this depends heavily on your ability to produce quality lines for games. If your work is off, you'll see a lot of "opportunities" that could lose you a lot of money.

Conversely, if you wagered on a team at -130 when it's true value was closer to -110 (a PK, with a 50% chance of winning), then:

EV = (0.50 * 100) - (0.50 * -130) = -$15

It's like buying anything.. You can buy the same product (probability of winning) at different places and different times (ie: For $25 at one store, or for $15 at a different one). You'll be happy with the product 80% of the time no matter where you buy it, but when it does fail you (1 out of 5 times), you're out more money if you bought it at the first place instead of the second.

Can U tell me how to calculate this: P[win] or P[loss]? From stats or just predict a percentage?