Blackjack tilt: The hidden house edge

[Dan Druff]

I have played a good deal of blackjack this year, and it got me thinking about the game a bit more. I came up with a theory as to why normal basic strategy players lose so much money at it -- far more than the house edge of less than 1% at a standard-rules game.

The best of the non-card-counters are "basic strategy" players -- players who perfectly follow the by-the-book rules as to what to do with each hand. For example, a basic strategy player will hit a 12 versus a dealer's 2 or 3, but will stand versus a 4 through 6. Most players know basic strategy to at least some degree, and some know it perfectly.

But how does the house turn such a small edge on basic strategy players into a big one? A lot of this occurs from the compounding of the edge over hundreds (or in some cases, thousands) of hands. As the player keeps playing, his bankroll will (mathematically) keep dwindling, with higher expected total losses as time passes. Some other additional losses occur from a player deviating from basic strategy (sometimes based upon "feelings"), which can change the house edge to as high as 5%. These are all old concepts that have been discussed over and over. While I believe both of these explanations are valid, I think there is one other explanation for losses that is often overlooked: Tilt.

To understand my theory, you need to first understand the basic principle behind card counting, which by itself is quite simple:

- If there are a lot of tens and aces left in the deck (compared to the number of 2s through 6s), the player has an advantage. Therefore, it is wise for the player to raise his bet substantially.

- If there are a lot of 2s through 6s left in the deck (compared to the number of tens and aces), the dealer has a greater-than-usual advantage. Therefore, it is wise for the player to either lower his bet or avoid playing the hand altogether.

Of course, well over 99% of all blackjack players don't count cards. Therefore, they are unaware of the composition of the remaining shoe, and they are also unaware of the effect it has on them. Of course, that doesn't change the actual effect of the deck's composition on the player -- he just doesn't realize it.

Blackjack players tend to go on tilt because of two factors:

1) Mounting losses
2) Losing hands they expected to win, such as a dealt 20 versus a low dealer's upcard, where the dealer draws to a freak 21.

But how can a blackjack player tilt? While tilting blackjack players may deviate some from basic strategy, they still tend to stick to it for the most part. The greatest form of tilt comes from bet amounts. A player betting $50 per hand may suddenly bet $300, then $500, then $1000 per hand, in a sad attempt to chase his losses.

But how does the house have a greater advantage here, other than simply taking bigger action? I have come to believe that such tilt occurs at the WORST time for the player -- when there is a disproportionately small number of tens and aces left in the shoe.

Why do I believe this? A player is most likely to go on tilt when he is both losing and taking bad beats. This is most likely to occur when the "count is low" -- i.e. a lot of small cards left in the deck. For example, if the dealer shows a 5, he's much more likely to bust when there are a lot of big cards left in the deck than when there are a lot of small ones. Thus, while freak 21-drawouts can happen at any time, they are most likely to happen when the count is low. This, of course, is the time that the player is at the greatest disadvantage, whether he realizes it or not. If a player suddenly raises his bets at this point, he's not only giving the house more action, he's giving them the most action at the time when they have the greatest edge!

I have watched this happen to my table-mates time and time again. I feel like telling them, "No, it's not a surprise at all that the dealer didn't bust after showing a 6" or, "I expected your double-down on 11 to end up with a small card", but obviously I can't. Usually their response to such frustrating losses is to slam down a much bigger bet, which of course the house quickly eats up.

Apologies if this has been discussed before, whether here or elsewhere.

Comments?

[greg nice]

but, if the dealer draws out a luckbox 21, it would mean that the low count just used up maybe 3 or 4 low cards, swinging the count more towards the middle or high end. no?

edit/
and therefore, a player going on tilt might not be making such a bad mistake. and, if the count was high to start, its now even higher, so hes playing even better by increasing his bet

[Dan Druff]

Good point, and I actually thought of that.

However, high negative count situations won't be affected all that much by 3 or 4 low cards rolling off. For example, a running count of -14 will take a long time to neutralize back to zero. I wasn't referring to situations where the count is close enough to zero to where one hand can fix the whole thing.

[greg nice]

well if you are referring to situations with a huge low count, where one hand wont affect it, it seems like your thesis is highly situational. a large low count is just as likely as a large high count. although, i guess it would be more likely for a 21 drawout to occur during a large low count.

but, most of the time, the count isnt going to be largely skewed in either direction, right? so 3 or 4 low cards might swing it back where the affect chasing losses tilt is negligible.

[Double Down]

Interesting thought, but I've also seen players tilt by becoming "scared" when a dealer pulls out a miracle, and they'll lower their bet to the minimum, or sit out a couple hands, or even wait til the next shoe. Oppositely, players when doing well often tend to raise their bets to take advantage of a "hot" streak.

I think the reason for the losses of otherwise perfect bs strategy players is because of the increasing of the bets, thus
1. Increasing the overall average bet and
2. Greatly increasing his risk of ruin

I think you're also right on about how tilt will make a player sway from bs.

[stigmata]

Very interesting post Dan.

But I'm also not convinced for the reasons given above. Furthermore, tilt is most likely to be induced by a *series* of beats, rather than just a single hand. Thus, on average, any unfavourable count should be substantially reduced/reversed.

[CORed]

Having played a fair amount of blackjack, mostly card counting, I would like to point out that bad runs and bad beats can happen at any count. Yes, they happen more often with negative counts, but it is entirely possible to get clobbered in very good counts, and it's also very possible to have a hot streak in horrible counts. If you're counting, the bad runs with favorable decks are the ones that really hurt, because you're losing bigger bets.

This is why making money by counting cards takes a big bankroll and lots of discipline. If you don't have a solid grounding in math and statistics, it's very easy to conclude that basic strategy and card counting are a lot of BS and decide to go with your gut.

My experience is that typical players don't really know basic strategy. Most of them know they should hit stiff hands against big dealer up cards and stand against small ones, but few really know the correct rules for doubling or splitting. I used to see a lot of people double soft hands against dealer 7's or 8's, fail to double 11's against a dealer 10 or double a 10 against a dealer 10. I would venture to say that only a small minority of typical players would split 9's against a 9 or hit a soft 18 against a 10. Many will sometimes stand on 15's or 16's against big upcards, especially if they've busted a lot lately.

[MrBrightside]

[ QUOTE ]
Having played a fair amount of blackjack, mostly card counting, I would like to point out that bad runs and bad beats can happen at any count. Yes, they happen more often with negative counts, but it is entirely possible to get clobbered in very good counts, and it's also very possible to have a hot streak in horrible counts. If you're counting, the bad runs with favorable decks are the ones that really hurt, because you're losing bigger bets.

This is why making money by counting cards takes a big bankroll and lots of discipline. If you don't have a solid grounding in math and statistics, it's very easy to conclude that basic strategy and card counting are a lot of BS and decide to go with your gut.

My experience is that typical players don't really know basic strategy. Most of them know they should hit stiff hands against big dealer up cards and stand against small ones, but few really know the correct rules for doubling or splitting. I used to see a lot of people double soft hands against dealer 7's or 8's, fail to double 11's against a dealer 10 or double a 10 against a dealer 10. I would venture to say that only a small minority of typical players would split 9's against a 9 or hit a soft 18 against a 10. Many will sometimes stand on 15's or 16's against big upcards, especially if they've busted a lot lately.

[/ QUOTE ]

I agree. I see a LOT of deviations from basic strat. My wife and I get crap all the time for hitting with a 12 against a dealer 2 or 3. Invariably, If I say "that's the book play" some jerk responds "who do you think WROTE the book" while nodding their head toward the pit boss....

I'm not a died in the wool card counter. I go to vegas once a year with my wife and I'm there to gambooool, but I play perfect basic strat. Also, I've hit upon a compromise. I don't want to card count the whole time, and I tend to play the two deck game at the plaza downtown (where we stay). I either do one of two things.. Count the first two hands out. If the count is very positive, I up my bets. Sometimes I count aces instead, and bet more into ace-rich decks. Occasionally, I do both.

I had a beatiful run on this last trip. Several times while just counting the first couple hands, I had a serious plus count and an ace-rich deck, I pushed bets agressively in these situations and got some good blackjacks/soft doubles and had a +700 session for the nite. We were talking about card counting/blackjack over lunch with some guys at work the other day and I told them of this system for an easy intro and a way to give a SLIGHT edge over just plain basic strat. What do you guys think?

[NL Rounder]

The reason is not tilt, it's simply the short-term, normal VARIANCE inherent to the game.

[CORed]

I think you also need to look at the psychological impact of the way the game is structured. When you bust, the dealer immediately scoops up your money and your cards. The message this sends is "boy did you screw up". For the player who doesn't have a clear concept of basic strategy, or has doubts about whether it is really the correct way to play or a massive scam perpetrated by the casinos, I think this tends to discourage players from hitting when they should.

Also, of course, there is the fact that Blackjack is such a high variance game. You often get incorrect feedback from the game, either because your correct plays repeatedly lose or your incorrect plays repeatedly win. This is true of all gambling games, but maybe a little more severe in blackjack then in other games.

Add to that the hindsight factor. Very often it will become apparent after the fact that if you had made a different decision, you would have won instead of losing. If you are sitting at third base, very often the other players at the table will point that out to you, sometimes very rudely. Most blackjack players have very short-sighted hindsight, and are incapable of realizing that the first base player's actions have just as much influence on the outcome as the third base player's. They just know that if that idiot at third base hadn't taken the dealers bust card, they would have won.

All in all, blackjack, if you don't understand probability and math, offers all kinds of inducements to make the wrong decision, perhaps more than any other gambling game. This is one of the reasons that, in spit of ithaving a pretty low house edge for the basic strategy player, it remains a very profitable game for the casinos.