Theoretical House Edge vs. Actual Casino Take

I'm new to this forum, so forgive me if this topic has been addressed before.

Looking over the Nevada Gaming Board's 2011 statistics for Strip casino revenues, I noticed that they list the casinos' Win Percent for various table games and the different denominations of slots. My understanding is that the Win Percent is the actual percentage of bets won by the house, with the rest going to the players.

The Win Percents listed for Strip casino slots are about what you would expect, with a house Win Percent of 9.92% for penny slots, decreasing to under 4% for $25-$100 slots.

The Win Percent in 2011 for blackjack was 10.85%. Since blackjack has a theoretical house edge under 1% using basic strategy, I assume that the actual 10.85% casino win is due to lousy players who don't know basic strategy.

What puzzles me is the Win Percent of 15.72% reported for roulette. Roulette is supposed to be purely a game of chance with no skill involved and therefore no way to diminish one's chances by playing poorly.

So my question is: with most roulette bets having a house edge of 5.26% (or 2.7% on a single-zero wheel), how is it that the casinos can report a Win Percent exceeding 15% of all bets?

- Walt

I'm not sure, but the state Montana calculates hold as 1 - money paid out/drop (drop is money inserted into the machine). This number would be higher if players play longer and lose more of their original cash inserted.

Table game win is reported not as a percentage of all bets but as a percentage of the drop. Since players rewager their bankroll more than one time, win / drop is greater than win / handle. Casinos typically can't get total handle for table games so they use drop instead. What the roulette numbers tell you is that on average, players played through their buyins three times.

I'm pretty sure that percentages on fruit machines are based on bets made and bets paid out - since in the UK and I assume elsewhere - the machine has all the figures. Thus as 96% machine would, on average, keep 4% of all money put in - obviously a sole $100 machine that paid a huge jackpot might mean the house lost for that month's figures, but with smaller valued machines, there are many machines and a few jackpots here or there doesn't effect the bottom line much.

With table games, they don't count the total bets but only how many notes are exchanged for chips at the table - that is why you usually can't buy chips at the cage. What it's saying is that for every $100 of chips sold, the house keeps $10.85, $15.72 or whatever.

In practice suppose in a given period 10 players each sat down and bought in for $100, played several hands, some won some lost, and got up; the table would have taken $1000, on average each player would have lost $10-$15, and the table made on average $100-$150. From each player's view, they made many bets before their money ran out or they gave up playing - for instance if I buy in for £20 at a £2 blackjack table, it can all go in a few unlucky hands (I once had £16 running, 4 splits, all doubles, all lost!) or actually last quite a while.

Okay, I think I get it. The house win on slot machines is reported as a percentage of all bets, whereas the house win on table games is reported as a percentage of all buy-ins. The table players playing thru their buy-ins multiple times in one sitting is what increases the Win Percent for the house.

Thanks.

- Walt

Quote: walt702

What puzzles me is the Win Percent of 15.72% reported for roulette. Roulette is supposed to be purely a game of chance with no skill involved and therefore no way to diminish one's chances by playing poorly.So my question is: with most roulette bets having a house edge of 5.26% (or 2.7% on a single-zero wheel), how is it that the casinos can report a Win Percent exceeding 15% of all bets?

15% is low for roulette, a better average is 18%
to 23%. People reinvest their winnings over and
over till their money is gone. And its not 'exceeding
15% of all bets.' A casino can only win what the
players bet, where would the extra 15% come from?

EvenBob, the 15.72% is the reported house win on roulette by Strip casinos for all of 2011. As I've learned here, that's 15.72% as a percentage of player buy-ins.

You suspect correctly that the house win for roulette could be higher. For downtown Vegas casinos, the 2011 win percent was 20.08%, and the win percent for North Las Vegas casinos was a whopping 24.05%. The Boulder Strip was slightly lower than the LV Strip at 15.65%.

Extending the discussion to race and sports betting, I suppose that since those wagers are all ticketed and machine-tabulated, they are reported using the same method as slot machines, i.e., the reported win is a percentage of all bets. For all of 2011, LV Strip casinos reported a racebook win of 15.67% and a sportsbook win of 3.84%.

- Walt

Quote: walt702

...The Win Percent in 2011 for blackjack was 10.85%. Since blackjack has a theoretical house edge under 1% using basic strategy, I assume that the actual 10.85% casino win is due to lousy players who don't know basic strategy....
- Walt

Walt, no, this actually relects strong Basic Strategy play. With good/optimal play, the table hold is about 10 times the house edge, so a 2.5% HE game would hold about 25% against strong play, and 35% or more against poorer play.
For Blackjack, in newer markets with weak players, the 1% BJ game holds closer to 20%; for mature markets with strong and consistent players, a 1% game does hold about 10%, - reflecting strong play.

Paigowdan, in that first post I was still thinking that the house win for table games was calculated the same way as slot wins. Now that I understand the difference, I agree that a 10.85% blackjack hold reflects a lot more favorably on the players.

Apparently the sharpest blackjack players go to the Boulder Strip (9.47%), while the weakest players can be found in North Las Vegas (17.68%).

- Walt

>Apparently the sharpest blackjack players go to the Boulder Strip (9.47%),
Not necessarily. One might say that they are already there or simply don't respond to glitz and glamor and do not desire traffic jams and gawkers. Casinos seem to attract locals and some locals clearly park themselves at "their" casino whereas many other locals seem to flit about sampling the various "toppings" that are available on their Plain Vanilla ice cream.

> while the weakest players can be found in North Las Vegas (17.68%).
Not necessarily the weakest players just the weakest statistical base since there are so few licensees reporting.
Its probably best to use Downtown figures instead of North Las Vegas figures, particularly for slots.
Also by weakest player you might mean in skill but weakest probably should embrace those who are most under bankrolled to survive a dip in the Variance Pool. That is likely to be North Las Vegas as well.

I once had a boss who could not understand why he almost always lost as Blackjack in AC. He always played at a $25 minimum table. And usually had a $300-$400 bankroll. LOL

Quote: walt702

Apparently the sharpest blackjack players go to the Boulder Strip (9.47%), while the weakest players can be found in North Las Vegas (17.68%).

The casino's baccarat win percent always rises dramatically over Chinese New Years. It has nothing to do with weaker players, it's just that they are far more obsessive, and they re-play their winnings over and over.

I take offense at the news reporting the variable win percents as "casino luck". While there is obviously going to be some variance, it is much more a function of repeated play.

The NGC reports sometimes subdivide the categories to the point, that you would expect large variance. If you look at the category of $100 machines on the Boulder strip, they will show that the casinos sometimes lose a larger percentage. Obviously that has nothing to do with luck, but the machines must pay out soon or later.

Quote: walt702

Apparently the sharpest blackjack players go to the Boulder Strip (9.47%), while the weakest players can be found in
North Las Vegas (17.68%).

It's tough to compare two casinos just by hold. For example Casino A and Casino B have a $10 bettor sit at a BJ game. Both casinos deal 100 hands/hour and have the same rules. A young player at Casino A buys in for $200 and plays for an hour; he's very bad so let's call him a 2% player. A retiree at Casino B buys in for $200 and plays for three hours at a 1% edge.

Casino A ($10, 100 hands, 2% edge):
Expected Win: $10 * 100 * .02 = $20
Hold: $20/$200 = 10%

Casino B ($10, 300 hands, 1% edge):
Expected Win: $10 * 300 * .01 = $30
Hold: $30/$200 = 30%

If you didn't know all of the information about play time, average buy-in, expected win, etc., it'd be easy to think that Casino B will be the more profitable casino and/or the one with the worst players. However, maybe Casino A is a strip casino that might see of 500 of these short term bad players a night while Casino B is an out of the way locals casino with only about 100 retirees a night. Casino A will hold less of more while Casino B will hold more of less. It's almost pointless to try and compare the two. It's better to look at total drop and the win (in dollars) separately.

All helpful comments. Thanks, folks.

the hold in example B would be 15%

Quote: WongBo

the hold in example B would be 15%

Yup. Whoops, I changed the buy-in amount in a hasty edit and forgot to change the hold%.