I would like to know the perfect collusion strategy so that I can try to simulate it.
Where can I find the perfect collusion strategy ?
Definitely a positive EV collusion strategy exists. I have seen teams in action, I don't understand what they're doing though. Anecdotally, I heard in the range of an 8% edge. Unconfirmed.Quote: ssho886 PLAYERS with perfect collusion strategy, possible to get +ve EV ?
I would like to know the perfect collusion strategy so that I can try to simulate it.
Where can I find the perfect collusion strategy ?
link to original post
Quote: TigerWuEven with six players colluding there are still two hands that are unknown; the dealer's hand and the dead hand.
link to original post
Knowledge that the dealer cannot have a teen or day is huge.
Knowledge that the dealer has a much higher than normal chance to have a teen or day is also helpful.
8% sounds a bit high, but I’d be very surprised if knowing 24 of the tiles (leaving 4/8 for the dealer) does not result in a player edge.
Quote: SOOPOOQuote: TigerWuEven with six players colluding there are still two hands that are unknown; the dealer's hand and the dead hand.
link to original post
Knowledge that the dealer cannot have a teen or day is huge.
Knowledge that the dealer has a much higher than normal chance to have a teen or day is also helpful.
8% sounds a bit high, but I’d be very surprised if knowing 24 of the tiles (leaving 4/8 for the dealer) does not result in a player edge.
link to original post
But what can you do with that knowledge? The bets have already been made. There's only three ways to set each hand. If you know the dealer has a good chance of having a teen or a day, what can you do but set your hand the best way possible, which is something you would most likely be doing anyway?
If this strategy was at all viable, you would think some Asian team somewhere would have capitalized on it already.
Edit - I just said the same exact thing as Soopoo. Sorry about that. Didn’t read all the replies close enough.
Quote: TigerWuQuote: SOOPOOQuote: TigerWuEven with six players colluding there are still two hands that are unknown; the dealer's hand and the dead hand.
link to original post
Knowledge that the dealer cannot have a teen or day is huge.
Knowledge that the dealer has a much higher than normal chance to have a teen or day is also helpful.
8% sounds a bit high, but I’d be very surprised if knowing 24 of the tiles (leaving 4/8 for the dealer) does not result in a player edge.
link to original post
But what can you do with that knowledge? The bets have already been made. There's only three ways to set each hand. If you know the dealer has a good chance of having a teen or a day, what can you do but set your hand the best way possible, which is something you would most likely be doing anyway?
If this strategy was at all viable, you would think some Asian team somewhere would have capitalized on it already.
link to original post
Unlike the dealer you can choose a less optimal setting for your hand. You would do this if you knew the most likely setting of the dealer.
This might help losses turn into pushes.
Also, knowledge of what tiles the dealer probably has means knowledge of how the dealer must set their hand. You could again choose a setting for your hand to counter the dealer's hand which must be set house way.
If you can minimize losses and maximize wins on a very low house edge game it might make a difference.
Quote: TigerWuQuote: SOOPOOQuote: TigerWuEven with six players colluding there are still two hands that are unknown; the dealer's hand and the dead hand.
link to original post
Knowledge that the dealer cannot have a teen or day is huge.
Knowledge that the dealer has a much higher than normal chance to have a teen or day is also helpful.
8% sounds a bit high, but I’d be very surprised if knowing 24 of the tiles (leaving 4/8 for the dealer) does not result in a player edge.
link to original post
But what can you do with that knowledge? The bets have already been made. There's only three ways to set each hand. If you know the dealer has a good chance of having a teen or a day, what can you do but set your hand the best way possible, which is something you would most likely be doing anyway?
If this strategy was at all viable, you would think some Asian team somewhere would have capitalized on it already.
link to original post
To give you a simple example, which I have used numerous times, if I have a choice of gong/3 versus 9/4, and I KNOW the dealer can’t have a teen or day you of course don’t even think and play 9/4. If you KNOW there are no teens or days left you will play the 11 with the high 8 to make a high 9 total on occasions where you would normally not. There are a bunch more subtle examples which can be used as well.
I am not sure it has NOT been taken advantage of. I’ve been at tables as the only native English speaker where I was asked repeatedly about my tile composition. I’ve also played at ‘loose’ tables where many players were clearly showing their tiles to the entire table while cursing!
Bottom line, knowing other tiles helps the player if he knows what to do with the information. Does it help enough to overcome the house edge? That is above my pay grade.
Quote: teliotDefinitely a positive EV collusion strategy exists. I have seen teams in action, I don't understand what they're doing though. Anecdotally, I heard in the range of an 8% edge. Unconfirmed.
link to original post
8% seems high to me too, even with optimal strategy. I've played with teams that invariably seemed to speak in Mandarin, Cantonese, or Vietnamese. When they speak in Mandarin, I can make out the numbers they are saying. Also, it seems I had an unspoken contract with these team to show them my tiles in return for their advice. Based on all the information I gathered, I think they were mainly counting the teen and day tiles.
This would certainly be a noble project to study collusion in tiles. However, weighing the pros and cons, I can't justify spending the time on it.
Quote: ssho88Agreed, asians mainly counting the teen and day tiles . . .I am looking for optimum collusion strategy.
link to original post
WOO has a calculator that you can enter your tiles and the known tiles. It will give you the optimal strategy for the specific house way it is programmed for. There may be tiny adjustments if the house ways at your casino are different.
Quote: SOOPOOQuote: ssho88Agreed, asians mainly counting the teen and day tiles . . .I am looking for optimum collusion strategy.
link to original post
WOO has a calculator that you can enter your tiles and the known tiles. It will give you the optimal strategy for the specific house way it is programmed for. There may be tiny adjustments if the house ways at your casino are different.
link to original post
I would like to know the strategy so that I can write it into a simulation program to check the average ev/round.
If it worth 8%, a 6 players team may able to get 0.08 x 6 = 0.48 unit per round !
Quote: ssho88Quote: SOOPOOQuote: ssho88Agreed, asians mainly counting the teen and day tiles . . .I am looking for optimum collusion strategy.
link to original post
WOO has a calculator that you can enter your tiles and the known tiles. It will give you the optimal strategy for the specific house way it is programmed for. There may be tiny adjustments if the house ways at your casino are different.
link to original post
I would like to know the strategy so that I can write it into a simulation program to check the average ev/round.
If it worth 8%, a 6 players team may able to get 0.08 x 6 = 0.48 unit per round !
link to original post
It is not worth 8%. Doubtful it is worth more than 1%. I’ve played for hours on end at tables where I had lots of this extra info. The number of times you make a play against what you would do without the knowledge is very small. And remember, you may just be changing a 53% win on a given hand to a 55% win. And in practicality, you won’t have all 24 tiles out available and ready to put in the WoO calculator.
I LOVE the rare hand that I play against ‘basic strategy’ due to the info. It’s just not that common.
Quote: SOOPOOIt is not worth 8%. Doubtful it is worth more than 1%.
link to original post
I agree. If it were worth even 1%, more people would be doing it.
Our own JB, who made my pai gow calculator, might be motivated to work on this if he were paid enough money.
Just curious, Mike, if you have ever witnessed a team in action. I have. It is something to behold ... they set their tiles, communicate somehow, then some re-set their tiles. It's lightning quick.Quote: WizardQuote: SOOPOOIt is not worth 8%. Doubtful it is worth more than 1%.
link to original post
I agree. If it were worth even 1%, more people would be doing it.
Our own JB, who made my pai gow calculator, might be motivated to work on this if he were paid enough money.
link to original post
It's not just the top tiles they care about. The choice between high-low and mid-mid is the difference between going for a win on the top value hand and a win on the bottom value hand.
I suspect the only way to analyse this would be to run simulations and work out the possible dealer hands (and player's ways to play) given the tiles you know about. In essence you just work through the last N in the deck to work out the possible dealer hands - I've done this for a 3-card poker based game I was thinking of. (I know that others have looked at this idea for various card games and shown some game are beatable and others aren't.)
Quote: teliotJust curious, Mike, if you have ever witnessed a team in action. I have. It is something to behold ... they set their tiles, communicate somehow, then some re-set their tiles. It's lightning quick.
It's not just the top tiles they care about. The choice between high-low and mid-mid is the difference between going for a win on the top value hand and a win on the bottom value hand.
link to original post
Yes, I have. I agree with your description of it. Seems to me some don't even look at their tiles but feel the dots.