Party Poker Flush Boards

One of the most common conspiracy theories among online poker players is about flush frequency. So we decided to take a sampling of over 30,000 five-card boards to determine how often a flush possibility is dealt.

Every so often in chat, you'll hear a player complain about flush frequency, particularly in Limit Hold 'Em.

After losing a hand:
CHUMPLAYER: Oh ****, another flush F***s up my set!
CHUMPLAYER: This *&^CHUMPLAYER: This *&^$&#ing site has too many #&$*#$ing flushes
amp;#ing site has too many #&$*#$ing flushes
CHUMPLAYER: They should call this place PartyFlushes
CHUMPLAYER: What a bunch of bull$#!@

The next few hands:
CHUMPLAYER: Oh, surprise another flush!
CHUMPLAYER: There's the heart on the river! I'm sure someone has it!
CHUMPLAYER: Of course you had it! The boards ALWAYS show a flush here.
CHUMPLAYER: I f@#@*&ing swear, top pair is never good here.
CHUMPLAYER: From now on, I'm just holding suited cards

There are lots of reasons people go on tilt from seeing flushes turned over by their opponent. Flushes usually win, and people tilt when their flopped set gets beat by a player calling bad odds to the river to win with the flush.

Here are some reasons why we may "feel" that there are too many flushes in an online poker game:
  • Flushes usually win so the hand is shown down instead of mucked. When you beat bottom two pair with your set, the loser mucks his hand unless he led the betting.
  • Four players seeing the flop in ten handed Hold 'Em tables usually have better hands than four players seeing the flop in your five handed Hold 'Em game at home.
  • Players online in limit Hold 'Em will often hold any two suited cards and will almost always hold any suited Ace. Lots of players are flush chasers.
  • Boards that have 3, 4, or 5 cards of the same suit actually aren't very rare.

Flush Boards = Theory vs. Reality

We wanted to do a study to determine if Party Poker had a significantly greater number of flush boards. We're looking for a percentage that might show some evidence that there are too many flushes on Party. If you averaged the Party Poker rigged conspiracies, you would expect at least a 10% increased chance of a flush board.

We took all the hands we have logged from the Party Poker engine and extracted only those boards that had five cards dealt.

Sample size: 31,889 Texas Hold 'Em 5-card boards

Total possible combination of cards on a five-card board: 2,598,960

Total percentage of boards that contain three, four, or five cards of the same suit: 37.11%

Given our conjecture above, if there is a 10% greater chance of a flush board dealt at Party Poker, we can expect over 40% of the boards to show 3-flush, 4-flush, or 5-flush. This would be a 10% increase over normal frequency.

All the theoretical math was done by Brian Alspach in a Poker Digest Article.

The results:

In our sample size, the total percentage of boards that contain three, four, or five cards of the same suit: 37.76%

Conclusion: In a limited size sample of 31,889 hands, there is no indication of any bias toward flushes on Party Poker.

In the coming year, we plan to do a further study of flush frequency on board cards to include a larger sample size.