Twister Poker is the newest development in iPoker. It's a SNG where winner takes it all, games start at hyper turbo-alike 25bbs, 3 players enter. How it differs from regular SNGs is that you do not know how much you're playing for before the game starts. In case of 10$ SNG, you can be battling for 20$ prize or 10000$ prize. I'd presume this makes this format very popular with recreational players. On the other hand, when compared to hyper turbo headsups, rake seems to be on the high side in these games. iPoker is making prizepools in a way that for a 10$ SNG you're fighting for 27.94$ prize on average, which makes rake at the games somewhere around 6.7-6.8%. You should have higher edge when there's three players in the game and more opportunities in a longer tournament, but on the other hand game selection is harder than in heads up games. Compared to Sit n' gos, absence of ICM presumably makes good players have less of an edge against weaker players because ICM adds another layer in where you can make mistakes. For what is our current information, rakeback is counted as 8% or perhaps it is influenced by what kind of SNGs you hit. I don't know it yet for sure, but I'll update it here when I know how rakeback or VIP points for these games are calculated. That's it about introduction, now let's see some numbers. Odds in hitting a certain type of Twister SNG: 1 / 20000 = 1000x buyin for the winner (0.005%) 1 / 20000 = 200x buyin for the winner (0.005%) 1 / 10000 = 100x buyin for the winner (0.01%) 1 / 4000 = 50x buyin for the winner (0.025%) 1 / 1000 = 20x buyin for the winner (0.1%) 1 / 125 = 10x buyin for the winner (0.8%) - 99.8% of starting Twisters are between 2-10x 1 / 12.5 = 6x buyin for the winner (8%) - 99.1% of starting Twisters are between 2-6x 1 / 6.6 = 4x buyin for the winner (15.2%) - 91.1% of starting Twisters are between 2-4x 1 / 1.333 = 2x buyin for the winner (75.9%) So, the games are a mixture of poker and lottery. What does this mean for a aspiring poker player? I made a simulation of 100 million games of Twister poker to find out the standard deviation of the games. Then I did 3 simulations for 10 million games, one for a breakeven player, one for a 5% winner and one for a 10% winner. The standard deviation of the games is about 4.6 buyins a game. Let's sink this in a bit, the game that is most similar to Twister poker, hyper turbo headsup SNGs has a standard deviation of exactly 1 buyins a game. 180-man SNGs that are known to be variance monsters are between 5 and 5.5 buyins a game. If you want to play these games professionally, you better know you need to be playing a lot of games for your skill to be dominant. It will be hard to estimate your edge correctly, when the game type adds it own kind of variance. You should only look at how many games you're winning. If you're winning more than 35.8% of your games, you should be at least breakeven. More than 37.6%, you should have over 5% edge in the games. More than 39.4% and you should have more than 10% edge in the games. Another poker related site suggests 50 buyin bankroll management for these games. If you are playing these games as a breakeven player, you're 30% likely to lose 50 or more buyins during your first 1000 games. For 5% edge, probability is just 7% and 10% edge the probability is just 0.7%. The worst 0.1% run for 1000 games would be -200 buyins for a breakeven player, -151 buyins for 5% winner and -95 buyins for 10% winner. So a high edge pays in this format, but I'm not sure how high an edge you can get in Twister Poker. On the other hand, if you run good, you will run godly good in the games. Breakeven player, +5% player and +10% player all got the best run at about +2000 buyins. It's all about hitting and winning 2x 1000 buyin games during a 1000 game period. For this structure, you'll win your expectation or more 35% of the time. Regardless of your edge. I'd think that this has something to do with the previous chapter. If you win, you can win really big, but most of the time you're not winning. Median results for a breakeven player are -22 buyins, for a 5% winner +29 buyins and for a 10% winner +77 buyins out of 1000 games. So the games will quite likely feel more depressing if you're looking at them on a tracking software report. In case you have something you would like simulated for Twister poker, or you have questions or you spot an error in my math, please post a message to this thread.