Originally posted on 09/22/2012:

Thousands of ways to analyze this, but %'s are used to gauge many things.

% market. - This is the total market on a game. when you have a game with 2 teams at -110, you have two teams priced at 52.38% which when added gives you a total market of 104.76 %. Therefore, the theoretical hold is 4.54%, since 4.76 of the 104.76 = 4.54%. This market expects both sides to win 50% of the time, and as a player, you can expect to lose 4.54% of your wagering handle if this happens.

Now, lets look at the scenario above. You have bet the Jazz +3 vs. the Bulls at -110. The market then decides that the true market price is Jazz +2 -110. Using 10 cent per 1/2 point estimation, you're bet is now 7.32% reduction from its perceived market value. Jazz +3 is now perceived at -130 (56.52%) and you have it secured at -110 (52.38%). 4.14% off of 56.52% is where we get the 7.32% reduction figure.

You now have an edge. Lets say the new market price is accurate and they played this game 10,000 times.

Out of the 10,000 games, lets say the following results happen- Jazz lose by 4 or more 4400 times, Jazz lose by exactly 3 points 400 times, Jazz lose by exactly 2 points 400 times, and the Jazz either lose by only 1 or win the game outright 4800 times.

Lets look at the new price from the bookmakers angle, he has Jazz +2 -110 and Bulls -2 -110 and every game he has 1 player betting $100 to win $90.91 on the Jazz +2, and 1 player betting $100 to win $90.91 on the Bulls. If he is getting $200 bet on every game, and they play the game 10,000 times, he is getting $2 million in total bets. What is he paying out after those 10,000 games?

4400 games jazz losing by 4 or more he pays out $190.91 to the Bulls -2 bettors for a total of $840,004 (4400 x $190.91)

400 games Jazz losing by exactly 3 points pays out $190.91 to the Bulls -2 bettors for a total of $76,364 ( 400 x $190.91)

400 games Jazz losing by exactly 2 points pays out the original $100 bets as a push to both the Jazz and Bulls bettors for a total of
$80,000 (400 x $200)

4800 games Jazz win the game outright or lose by exactly 1 point only and pays out $190.91 to the Jazz +2 bettors for $916,368 (4800 x $190.91)

So the Bookmaker has collected $2,000,000 in bets, and has paid out $1,912,736. So he has made $87,264 profit on the games, which is a hold of 4.36%.

Now, lets look at that same scenario from the player angle who has bet the Jazz + 3 at -110. Lets say for those same 10,000 games, you bet the Jazz +3 $100 to win $90.91. What do you get back with those same "market results" ? You are betting $100 on 10,000 games for $1,000,000 in total bets.

4400 games Jazz lose by 4 or more, you get back 0
400 games Jazz lose by exactly 3 points you get back $40,000 since its a push ($100 x 400)
400 games Jazz lose by exactly 2 points you get back $76,364 ($190.91 x 400)
4800 games Jazz win the game or lose by only 1 point, you get back $916,368 ($190.91 x 4800)

You get back $1,032,732 which is a profit of $32,732 from your $1,000,000 in bets. A + 3.27% yield.

What if we calculated Jazz +3 at -130 ?, which means $100 bet to win $76.92

Now we get back a total of $959,984 which is a loss of $40,016 or 4% (which shows in this case, Jazz +3 at -130 generated the books desired hold range of 4%)

So in that scenario, Just betting the Jazz +3 at -110, instead of either Jazz +2 -110 or Jazz +3 at -130 results in a big difference between showing a profit of + 3.27% instead of a loss of 4% or 4.36%.

10,000 games at just $100 a game, thats a difference of around $73,000 just from being 1 pt better than market. Can you imagine if you were betting 5k or 10k a game?

That's why sports books need to be sharp, because it doesn't take much for them to be vulnerable.