Perfect Pairs

[Katweezel]

The general public standard for blackjack play in Australia is at the usual low level as I notice a guy stand on A4 V 3. (Why on earth would he refuse a free hit and stand on soft 15?) A woman doesn't double her 55 V 3, even though she has plenty of chips. Another guy stands on his pair of 7s V 5. It does not occur to someone else to split 33 V 6. Double 9 V 4? Nope.

The common trend seems to be stand on stiffs, no matter what the dealer has... Bad play must be setting clueless players back into the realm of 5,6, maybe as much as 7% negative expectation for their game. But the CSM tables are all pretty much packed at weekends with hordes of people playing like this. Is it a similar tale of generally poor public play in the US?

Perfect Pairs are a very popular sidebet here. Casinos pay 30/1 for a winning PP. In a 6-deck CSM, what are the correct odds against scoring a PP? How is this calculated? My guess is somewhere around 60-something to one. PP punters think they are doing great at 30/1. I'd appreciate the math for this, so I can tell a couple of ploppies I know what idiots they are for getting ripped off so bad! :cat:

 

[rollem411]

2 Words...Scavenger Blackjack.

As for the PP side bet. In a 6D game the probability of hitting that is .016077 or a 1.6077% chance. I've explained the math behind this in another thread. I'll try to dig it up and post it.

 

[davidpom]

Perfect Pairs is a ridiculous side bet that offers an edge of circa 15% depending on the rules. There are various paytables available to be used that affect the total edge.

In CSM's the odds will be affected, but if you work out the logic on a hypothetical game (i.e. six decks of cards used) then it works as follows:

I have a 2 of diamonds. What are the odds of me getting another?

Well, there are 312 cards in total. I have one already. So that leaves 311 cards. There are 5 other 2 of diamonds left in the shoe. So I need to get any of 5 cards out of 311 = 0.016%. Thats about a 1 in 62 chance. So you get paid either 25 to 1 or 30 to 1 on a 1 in 62 chance!!! Yuck!

There is some compensation due to coloured and standard pair payouts - but the odds are still rubbish. The edge in this game is far too big to play seriously.

My advice: avoid Perfect Pairs, unless you want to give your bankroll away!

 

[Sonny]

In a 6-deck CSM, what are the correct odds against scoring a PP? How is this calculated?

The probabilities for that side bet (and pretty much any other one) can be found here:

http://wizardofodds.com/blackjack/appendix8.html#perpairs

The house edge is usually around 5.79% for the 30:1 payout table.

-Sonny-

 

[garnier]

Here in Portugal, at Casino Estoril, we can play for the aces. If we get 2,3 ou for aces in a row we win the "progressive jackpot". We bet 1€ for play for the aces.

If we get 2 aces we get 25€, if we get 2 aces of the same kind we win 100€, and so on to the maximum prize of getting the for aces (i think it must be of the some color, i dont remember) that will be the value of the jackpot. Another day a chinese man win almost 200.000€...

 

[EasyRhino]

You know, a local joint just started offering the perfect pairs (on a double deck game, which might make it worse). 30 for suited pair, 10 for same color pair, and 5 for pair.

And the 10 minute stretch where people at the table got THREE perfect pairs sure would make a gambler thing it was a good bet.

Keeping a count of all pairs would seem impossible. Too bad the perfect pairs doesn't pay on "any ten"... that might be doable.

 

[Katweezel]

PP chances improve?

You know, a local joint just started offering the perfect pairs (on a double deck game, which might make it worse). 30 for suited pair, 10 for same color pair, and 5 for pair.

And the 10 minute stretch where people at the table got THREE perfect pairs sure would make a gambler thing it was a good bet.

Keeping a count of all pairs would seem impossible. Too bad the perfect pairs doesn't pay on "any ten"... that might be doable.

Presumably, at certain times, such as very high positive or negative true counts, your chances of scoring a PP are improved somewhat. :cat:

 

[Sonny]

Presumably, at certain times, such as very high positive or negative true counts, your chances of scoring a PP are improved somewhat. :cat:

It depends on how many of each card are left. In many cases the probability of getting a PP could be smaller when there are less ranks available for the PP. For example, in a DD game if we remove one of every low card (2-6 of each suit) the probability of a PP drops from .97% to .92% and the house edge is still a whopping 22.72%. But when we remove both of every low card the probability of a PP raises to 1.59% and the player gets an advantage of 22.22%. This bet depends a lot on the specific composition of the decks.

-Sonny-

 

[actuary]

It is clear that since PP is played from a shoe, there are times where the composition of the remaining cards will give the player the advantage. It's easy to come up with examples where the player would have an advantage.

This also goes for the Match the Dealer bet found on some bj and nearly all Spanish tables.

But the real question is how can we, the AP, know when we have the edge so that we can exploit the bet? Obviously, a computer could do this very easily. Since those are illegal in must jurisdictions, we need a counting system. But is there one simple enough for us to use?

 

[Sonny]

...we need a counting system. But is there one simple enough for us to use?

More importantly, is the advantage big enough and frequent enough to be worthwhile? I haven't run any sims yet but that would probably be the next thing I would look at. If it ain’t worthwhile, forget the counting system.

-Sonny-

 

[EasyRhino]

http://wizardofodds.com/blackjack/appendix8.html#perpairs

Dear lord, two decks yields a 25% house advantage?! And I place a single red chip bet on it once just to be a sport! That's $1.25 I could have tipped a cocktail waittress.

I feel like I need a shower.

 

[actuary]

More importantly, is the advantage big enough and frequent enough to be worthwhile? I haven't run any sims yet but that would probably be the next thing I would look at. If it ain’t worthwhile, forget the counting system.

-Sonny-

I don't have an answer to this very important question. However, the house edge in PP increases (worsens) for a "balanced" shoe as we play into a shoe. By "balanced", I mean a shoe were the cards that have been played are distributed evenly among all ranks (2 Jacks, 2 6s, etc.)

So intuition tells me that in order for the player to have an advantage towards the end of the shoe, the remaining cards have to be really out of balance. It wouldn't happen often, but when it does, the advantage can be big.

I also believe that this bet is vulnerable to shuffle tracking, with a perfect shuffling dealer.

 

[Sonny]

Well, the simulation results are in. Even with computer-perfect play and good penetration you will make the bet less than once every 200 hands. The EV is almost nonexistent. This is not a good bet for a card counter.

-Sonny-

 

[moo321]

Well, the simulation results are in. Even with computer-perfect play and good penetration you will make the bet less than once every 200 hands. The EV is almost nonexistent. This is not a good bet for a card counter.

-Sonny-

This is assuming we count every single card, rank, suit and all?

 

[Sonny]

This is assuming we count every single card, rank, suit and all?

Right. This is assuming we look at every card dealt, enter it into a computer and determine the exact house edge at the beginning of every hand. No estimations, no approximations, no simplifications just exact calculations using a full combinatorial analysis. Knowing the deck composition alone will not give you a worthwhile advantage.

-Sonny-

 

[EasyRhino]

Well, the simulation results are in. Even with computer-perfect play and good penetration you will make the bet less than once every 200 hands. The EV is almost nonexistent. This is not a good bet for a card counter.

a) How did you do the sim?
b) intuitively, the result suddenly makes sense. Any deck depletion is going to also reduce the availability of cards that you could pair with. So you'd probably have to get to situations where roughly half of the cards have been deplted from the deck, but they were all the same value. Or you're dealt to the last four cards in the deck, and they're all the same value.

 

[ohbehave]

We all have our stories but one night recently I'm playing pitch and 2 players at the table are making this bet almost every round. The guy hits it twice (for $250 each) that evening and the lady hits once... the suited pair... and I would have hit it once had I been placing the bet. They were also getting unsuited pairs by the boatloads. Almost made me start making the bet. Still it is a sucker bet in the long run.

 

[Sonny]

How did you do the sim?

First I whipped up a spreadsheet to make sure that my calculations were correct compared to the Wiz. Then I wrote a little simulation program that removes two random cards, calculates the house edge, removes two more cards, calculates the house edge, etc. It keeps track of how many hands give a positive advantage and how big each advantage is. In theory the program should give many more positive results because it only deals two cards per round (no dealer hand, other players, hit cards, etc.). Without all those burned cards you will get more rounds per shoe and more opportunities to find advantageous situations. The average advantages ranged from around 3-6% per hand but the frequencies are so rare that the overall win rate was tiny. The shoe games with good penetration had higher frequencies but lower average advantages so it almost evens out.

-Sonny-

 

[Sonny]

They were also getting unsuited pairs by the boatloads. Almost made me start making the bet. Still it is a sucker bet in the long run.

Yeah, the beauty of the bet, from the house's point of view, is that a player will get a match about 7-8 times per 100 hands. That really encourages people to make the bet because they see matches happening pretty often and they feel bad when they get one they didn't bet on. There's a lot of positive reinforcement even though the odds are against the players.

-Sonny-

 

[CasinoBlackjack]

Too bad the perfect pairs doesn't pay on "any ten"... that might be doable.

At Fallsview Casino in Niagara Falls Canada, they have a table called "Lucky Ladies"

Payout Table looked like this

Any 20 (This includes Ace/9) 4-1
Suited 20 9-1
Matching 20 19-1
Queen of Hearts Pair 125-1
Queen of Hearts Pair with Dealer Blackjack 1000-1

I was just playing regular, but most everyone else was playing these. Although people were down from this bet, the 17% house edge it gave didn't seem like a reasonable number, it didn't seem people were doing that poorly. Every time I would get a Blackjack I would play my $2.50 on there and I got it a few times and it seemed to even out by the time I was done my session.