Win % Tables v Slots

On another forum that I frequent the link below was posted so that folks could look at the New Jersey gaming numbers. I have also read here about numbers available for Vegas on an area-wide basis (areas defined by the gaming commission) as opposed to numbers from individual casinos listed in the Jersey report.

http://www.njccc.gov/casinos/financia/mthrev/

What I have not been able to wrap my mind around is the "win %" for slots v table games. The "win %" for most slots is right around what we have seen as the house advantage on them...10%, give or take a bit, based on the denomination. For table games, the "win %" seems much higher than the "house advantage". Can someone remind me of the reason that the "win %" is higher than the "house advantage", please.

Quote: RonC

On another forum that I frequent the link below was posted so that folks could look at the New Jersey gaming numbers. I have also read here about numbers available for Vegas on an area-wide basis (areas defined by the gaming commission) as opposed to numbers from individual casinos listed in the Jersey report.

http://www.njccc.gov/casinos/financia/mthrev/

What I have not been able to wrap my mind around is the "win %" for slots v table games. The "win %" for most slots is right around what we have seen as the house advantage on them...10%, give or take a bit, based on the denomination. For table games, the "win %" seems much higher than the "house advantage". Can someone remind me of the reason that the "win %" is higher than the "house advantage", please.

For table games the win % is normally the %age of money held by the casino as compared to the amount brought in for. It's not the per hand %age, but the total for the table over the course of the day/month/year. Say there's $1000 chips brought in for, and $900 paid out, that's a win% of 10%.

Slots don't look at the coin in, but the amount bet. It's kinda misleading, if you ask me, but no-one does (quite rightly).

Put another way, win % in slots is based on total handle (a.k.a. coin-in, wager volume, etc). Win % on tables is based on drop (the money in the cash box) because you can't calculate total handle manually. There are technologies that attempt to do so but they're not widespread. If a player brings $100 to a 90% RTP slot game, makes $500 worth of wagers and loses $50, he's hit the 10% EV as $50 (win) over $500 handle. If that had happened at a table, the casino wouldn't know that the player had made $500 worth of bets, so they'd just calculate $50 (win) over $100 (drop) and get 50%. In other words, the player played through his bankroll 5x.

Sometimes you can make reasonable estimates of the playthrough ratio. On roulette, basically all of the bets have the same EV and you know what that is. In that case, you can divide the actual win% by the EV and get playthrough factor. It's harder for games like blackjack or craps, where player strategy or bet variations make things too variable to estimate.

For what it's worth, Internet casinos use slot-style win% because they can accurately track wagers. I've seen numbers around 1.3% (for a game with an optimal EV of maybe 0.5%), indicating roughly 0.8% of suboptimal play.

It seems to me that this is all moot... the actual take home per table/machine is surely the most important part?

In this case, someone was implying that the slots (based on "win %" reported) were a "better" bet than table games...the apples to oranges way the calculations are done make it look that way if you take the "win %" at face value.

Quote: RonC

The "win %" for most slots is right around what we have seen as the house advantage on them...10%, give or take a bit, based on the denomination. For table games, the "win %" seems much higher than the "house advantage". Can someone remind me of the reason that the "win %" is higher than the "house advantage", please.

The two figures would be very close to each other IF everyone bought in for his bankroll and gambled it in one session at and one table and did not walk away with any chips and also played with utterly relentless optimal strategy. Players are not playing optimal Basic Strategy if they are half drunk and don't even know what basic strategy is. Players calling out various center bets at the craps table are not playing basic strategy. And players do not BuyIn at one table, play there for one session and then leave town.

As others have said, comparing the slot and table return percentage is comparing apples to oranges. For slots it is relative to total amount bet, and for table games total buy in. Internet casinos, which have the benefit of logging every single bet, do both types of games on a total money bet basis.

I also think there is too much emphasis put on hold percentage in table games. What should matter is the hold, which is the buy in less chip refills.

Quote: thecesspit

It seems to me that this is all moot... the actual take home per table/machine is surely the most important part?

The most important part is actually profit per square foot. At the end of the day, the one thing casinos can't get more of is floor space, so maximizing profits based on their existing area is (or should be) the true financial goal. The average dice table in Nevada won over $100k in August (it was a good month for craps). The average slot machine made less than $3500 during the same period. However, you can pack many slot machines into the same space as a single craps table, and overhead for slots is significantly lower than for a dice table - which requires pit oversight, higher surveillance effort, and a crew of at least 5 people to operate rotating in/out as shifts of 4.

With the combination of cashless ticketing and lower-maintenance video slot products, slot overhead basically fell to the floor, so over the past 20 years you've seen a gradual (or sometimes dramatic) shift in floor % toward slots. What's happening now is a slight resurgence of table games, for a few reasons. First, there are the social aspects of the games that you simply don't get by playing 1-on-1 with a computer. Then there is the decrease in total property revenue from gaming, which at certain places is less than 20% (in the past, it may have been 75-85%+). Baccarat in particular has seen an amazing resurgence, winning over 160M in August, and Pai Gow tiles has seen its win go up by over 100%.

If everyone bought a pile of chips, played through them one time, turned them in and bought more chips,
then the win percent would resemble the house average for each game.

The win percent varies the most for baccarat. It has been as low as 3.51% (June 2010) and as high as
17% (February 2010) this year.

My assumption is that if you did a racial profile, you would find a lot of Americans playing when it was 3.5%.
They only play through their money a few times. When it was 17% was during Chinese New Year's. My assumption
is that the celebrants would buy a pile of chips and stay at the table for many hours playing through their
chips multiple times. They may even store them overnight and play the same chips the next day.

If you cash the chips in after a while, and then buy new ones, the win percent will be much lower.

I imagine that the spread of machines to automatically cash in the Ticket In Ticket Out machines have changed win
percentages. You can turn your ticket into cash and then redposit the cash back into the slot machines after lunch.

Since many of these actions are arbitrary and don't really affect the bottom line, the Wizard feels that "win percent"
is a meaningless metric. However, casinos have been collecting this data for a long time, and I am sure that a
pit boss or a slot machine manager is judged by how well his casino does against the appropriate regional metric.

True. Square footage doesn't change or atleast it doesn't change very rapidly.
Yet even if you buy slot machines with a small "footprint" and really jam 'em in close to each other and do everything else to maximize square footage usage, its really a question of number of sheep and quality of the shearing.
I've been in casinos early in the morning when a bored craps dealer has chatted with the stick about how attractive robbing a convenience store was looking to him. Conversation designed to elicit tips from me? I think not. As the dice dealer had said, you could start shooting in the casino and be hard put to hit any player.

Its just that it is hard to develop a metric for shearing. We all know Harrah's does it well.

We all know that some casinos are packed, though not necessarily with gamblers. Some have sports figures and entourages, some have gamblers, some have the famed fanny-packed half-drunk "gambler" who is more robot-like than the slot machine he is feeding.

Just as some restaurants judge themselves by turnover rate, I guess casinos have to also.

You've all heard the expressions about a hooker on pay night. Sometimes casinos have to work fast. It can't just be square footage. Too many of those tiny little places make good money for it to be just a matter of Dollars per sq ft.

Table game hold percentage (please stop calling it win percentage, this really bugs me) is an emperical number that casinos have resorted to use in lieu of information on actual handle. Casinos use these numbers to measure the performance of the game based on historical hold percentages. However, many casino managers/execs do not consider how long it take for hold percentages to level out and start "sweating" unjustifiably.
Hold percentage for each game varies; not only by game, but also by jurisdiction. I've seen hold percentages on blackjack swing by 10% from one jurisdiction to the next. This is simply due rule variations, player skill, chip purchasing habits, dealer speed, how long the player is willing to play, what their win/loss tolerances are,....you get the point.
The fact is, there is a weak relationship between the amount players purchase in chips and the amount that the player bets in chips. So the hold percentage, is more like a hold ratio e.g. a 15% hold means a ratio of $0.15 to every $1 in the drop box. If you typically open a table late when there is significant demand, you can expect that the long-term hold will be very high, becasue you will get players migrating with chips from other tables. A lot of casinos that look at performance apply statistical analysis to the hold percentages on a monthly/weekly basis to determine of the hold is outside a specified range (say 95%). to me this makes no sense, as they are two distinct variables. I've always found that the win itself tends to distribute more evenly than the hold, becasue the cash drop, for reasons I can't fully comprehend, tend to show a skewed distribution.
"Win Percent" would be a term I would use if I knew the total amount of every $ wagered on the table. This included bets that are pusshed and rebet, bets that made from winnings etc. Win percentagre would be the total win/total wagers. This would be a meaningful statistic, as it would eventually trend toward the the theoretical house advantage (providing itis not a strategy dependant game, in which case player will make mistakes resulting in a higher %).

From a player perspective, I wouldn't bother looking at hold % on tables.

Quote: Kelmo

Table game hold percentage (please stop calling it win percentage, this really bugs me) is an emperical number that casinos have resorted to use in lieu of information on actual handle. Casinos use these numbers to measure the performance of the game based on historical hold percentages. However, many casino managers/execs do not consider how long it take for hold percentages to level out and start "sweating" unjustifiably.
Hold percentage for each game varies; not only by game, but also by jurisdiction. I've seen hold percentages on blackjack swing by 10% from one jurisdiction to the next. This is simply due rule variations, player skill, chip purchasing habits, dealer speed, how long the player is willing to play, what their win/loss tolerances are,....you get the point.
The fact is, there is a weak relationship between the amount players purchase in chips and the amount that the player bets in chips. So the hold percentage, is more like a hold ratio e.g. a 15% hold means a ratio of $0.15 to every $1 in the drop box. If you typically open a table late when there is significant demand, you can expect that the long-term hold will be very high, becasue you will get players migrating with chips from other tables. A lot of casinos that look at performance apply statistical analysis to the hold percentages on a monthly/weekly basis to determine of the hold is outside a specified range (say 95%). to me this makes no sense, as they are two distinct variables. I've always found that the win itself tends to distribute more evenly than the hold, becasue the cash drop, for reasons I can't fully comprehend, tend to show a skewed distribution.
"Win Percent" would be a term I would use if I knew the total amount of every $ wagered on the table. This included bets that are pusshed and rebet, bets that made from winnings etc. Win percentagre would be the total win/total wagers. This would be a meaningful statistic, as it would eventually trend toward the the theoretical house advantage (providing itis not a strategy dependant game, in which case player will make mistakes resulting in a higher %).

From a player perspective, I wouldn't bother looking at hold % on tables.

I guess that's the whole point. Unless you know the total amount of all bets made at the table(s), you have no way of calculating what percentage of that total the table retained/won. Lacking the ability to track total bets (for now), table game supervisors have resorted to trying to find some kind of correlation between the amount of cash stuffed in the box, chip count, fills, etc. As you say, these correlations are weak. Yet, I've seen many dealers fired simply because their table holds were below the house average for too long (two weeks, apparently in many cases, being too long). I remember spending a whole weekend dealing to one fish who evidently thought I was his "lucky" dealer---he bet 2-5 blacks and made plays like hitting hard 14 vs. a 6 and splitting threes vs. a 9. Somehow, he kept winning. I was starting to sweat, as this was a grind joint, and having my rack down two or three grand for three shifts in a row was going to look bad. Fortunately, he tanked it all late Sunday night. My point in relating this is that players like this can skew the hold, as well as players that buy in for cash, players that carry chips from the crap table to the BJ table, etc. etc. It's a statistic so tied to other events that it's virtually meaningless by itself.

I'll bet they envy slot managers because for them, win percentage is trivially easy to calculate: coin-in minus coin-out.

It was easier in the good old days when Max and Vito just stuffed a couple of suitcases full of money and took the next plane to Chicago.

The numbers from New Jersey seemed to point out to some folks on the other board that "slots are a better bet" than table games because of the "win %" shown for each. Yes, the table game % should probably be labeled differently for a more accurate description of what that number represents.

So...the next question is...does the "win %" (hold %) shown accurately represent the casino's profit for that month on those games or not? The bottom line of these reports is a total number that is taxed. If the % (whatever it is called) is not accurate, the taxes would be off. Or is it the correct profit and simply bad terminology?

Quote: RonC

The numbers from New Jersey seemed to point out to some folks on the other board that "slots are a better bet" than table games because of the "win %" shown for each. Yes, the table game % should probably be labeled differently for a more accurate description of what that number represents.

So...the next question is...does the "win %" (hold %) shown accurately represent the casino's profit for that month on those games or not? The bottom line of these reports is a total number that is taxed. If the % (whatever it is called) is not accurate, the taxes would be off. Or is it the correct profit and simply bad terminology?

The win is the win, regardless of what the hold % is and it iy should be taxed accordingly (Note: I'm not familiar with NJ Casino Tax structure).

Quote: mkl654321

Yet, I've seen many dealers fired simply because their table holds were below the house average for too long (two weeks, apparently in many cases, being too long).

Sad, but true. This is because most managers (and executives for that fact) don't understand hold % flucuations and assume it has to do with cheating dealers/players for two reasons: 1) because it is a shortcut to thinking. 2) becuse the security/surveillance industry makes a ton of money eggagerating the threat of and cost of cheating, so they can justify the costs of their elaborate systems. In my 26 years in the industry, I do not recall a situation where surveillance has actually caught a cheater in real-time. Verified, yes, but actually caught in real-time, no. Typically, it is the floor staff that will notice something is wrong and alert management or the culprit is caught during random reviews of footage. In this sense hold can be a useful indication that something may be off and may lead to something, but this only warrants looking into the matter and is by no means conclusive unless the footage and a subsequent investigation reveal something.

I have a craps table that consistently holds 20%, simply becasue we open it late. Much of the play comes from players who have already bought it on some of the other tables, so they carry in value chips to the game. Guess what happens to the hold on the tables that the chips moved from?

an aside on surveillance: they often do not want to reveal where they are looking so they may not report something as being from surveillance. I don't know if surveillance actually does catch someone in real time. I am told that a small California poker "casino" that had someone put weed in the betting area didn't catch it until the call FROM the pit. Only the fixed cameras were on it initially.

I can understand that a late opening table gets "overflow" from other, more-crowded tables and people probably lug chips with them. Perhaps that table has consistent percentages, perhaps they fluctuate. I can't see a dealer getting fired for something that is not his fault at all but if you were worried about it then I assume it happens.