The House wants even action.
Fact or Fiction?

How many times have you heard, "All the House wants to do is get even action on every game and make their 10% juice".  Now I've been hearing this for years, and what amazes me is the number of seasoned gamblers that still believe this.  Now I wasn't around when grand pappy Chan was running whisky during prohibition, but maybe back then that was the case.  But in the age of Supercomputers and readily available information, you better believe the "house", whoever you want that to be, is taking a stand on a good number of games every card.  If we the bettor can get all the information we can digest off the internet, and I can write a computer program to crunch stats in my spare time.  You better believe Roxy's boys over at LVSC are feeding HAL Jr. every stat imaginable and when it comes to sports are as close to clairvoyant as it gets.   Now I'm sure the lines makers do calculate the number they think would bring even action, however more importantly they also calculate the number that will bring a 50% split in the actual ATS outcome of the game.   Then they put out a number some place in between.  Now I’m sure you’re saying “Why would they do that Larry Chan?”    Well Larry Chan is going to show you.   For instance,  later today (November 15, 2003), the Cincinnati Bearcats will travel to Fort Worth to play the Horned Frogs of Texas Christian.    Now let’s look at some possible scenarios on setting the line for this game

 Now let's assume there is a finite amount of money to be bet on this game (yes I know in real life we cannot make this assumption, but bear with me).    So let's say that on today's game there will be $20,000 bet on this game.  Now let's also assume that Roxy's super computer spits out the number of -10 to get even action on both sides.   But the computer also says that the true line should be -4.  In other words if they played this game 100 times the line of TCU -4 would result in each team covering 50 times.  Now where should Vegas set the line to ensure the most profit? Someplace in the middle of course.  Well I'm sure Vegas takes into account some very interesting differential equations to find this exact number, but I'm going to make some more simplistic assumptions of possible outcomes.

So for our first possible line of TCU -10.  We know this will result in equal action of $10,000 on each side of the game.   So no matter the result of the game the house will make $1000 dollars per game, and in our example, if this game was played 100 times the house would end up making $100,000.   Now scenario number two.  Vegas set's the line to the calculated true line of TCU -4.   So now if at -10 the action was split evenly, with such a drastic departure from that number we know that most of the action will come in on TCU.  Exactly what percentage?  Well it doesn't really matter as you will soon see.

For example if the money came in 75-25 or 

15,000 on TCU    and   5,000 on Cincinnati the expected return for the house would be as follows.

if TCU covers.   -15000 + (5000*1.1) or  -$9,500

if Cinci covers     -5000 + (15,000*1.1) or  + 11,500

With each outcome occurring 50% of the time if we play 100 games...

(-9,500 * 50)  = -475,000  and (11,500 * 50) =  575,000   

or profit of $100,000.    The same amount of profit as if we had even action on both sides. 

18,000 on TCU    and    2,000 on Cincinnati the expected return for the house would be as follows.

if TCU covers.   -18000 + (2000*1.1) or  -$15,800

if Cinci covers     -2000 + (18,000*1.1) or  + 17,800

With each outcome occurring 50% of the time if we play 100 games...

(-15,800 * 50)  = -790,000  and (17,800 * 50) =  890,000   

or profit of $100,000.  

So as you can see when you have a true line where the outcome against the spread will be evenly split, the amount of money on each side is irrelevant to the bottom line profit in the long run.

Now scenario number three.  If we know that at a line of -4 each team will cover 50 times out of a 100.  We can also assume that any departure from that number will increase the likely hood that one team will cover a higher percentage than the other.  For the sake of example let's say that at our even action line of TCU -10 the actual results will be 40 ATS wins for TCU and 60 ATS wins for Cincinnati.  So we know at -4 each team will win 50 percent of the time and the profit will be $100,000 for the house in the long run, and we know at TCU -10  Cincinnati will cover 60 percent of the time and the profit will be $100,000 for the house in the long run.  Let's see what happens when we pick someplace in the middle.   So let's say we split the difference and select a line of TCU -7, and make a hasty interpolation of Cincinnati covering 55 percent of the time at that number.  Now let's do the math for the profit calculation.   We don't know in what proportion the money will come, but we do know that more money will come in on TCU than on Cincinnati due to the fact that if we are getting even money on the game at a line of -10.   So for example let's assume that the money comes in 65-35 or  13,000 on TCU and 7000 on Cincinnati

If TCU covers.   -13000 + (7000*1.1) or  -$5,300

if Cinci covers     -7000 + (13,000*1.1) or  + 7,300

TCU will cover 45 out of 100 (-5,300 * 45)  = -238,500

Cinci will cover 55 out of 100 (+7,300*55) =  +$401,500

for a profit of +$163,000.  

So as you can tell the lines maker would strive to set a number that will be someplace between the line that will bring equal action and true line.  

 Now I know with my above examples, I've pretty much made a mockery of the complexity of the actual equation that must be used to find the optimal line, but hopefully I've somewhat illustrated that the lines maker will attempt to find a line where public perception will allow him or her to set the line someplace other than the true line, and even at that "incorrect line"  more action will come in on the team that will come in on the losing end of the spread a majority of the time.