River Card tested for evidence of bias - 6 million hand sample from major sites networks

    • LaughingAssassin
      LaughingAssassin
      Basic
      Joined: 15.03.2012 Posts: 2
      I saw that some of the tests run at Online Poker Watchdog were mentioned in another thread in this forum a while back.

      A new series of tests has just been published and I thought they might be of interest:


      The tests are an analysis of heads-up, turn all-ins designed to check for evidence of bias in the deal.


      These tests are similar to the pre-flop and flop all-in bad beat tests that were previously carried out except the all-in occurs on the turn. Effectively the tests are designed to detect bias in the fall of the river card.


      The hand samples are the same as those used in the pre-flop and flop tests and consist of approximately 1 million hands from each of these major sites:


      PokerStars
      Party Poker
      Ongame Network
      Entraction Network
      Bodog Poker
      Merge Network


      Results showed that all tests were within 2 standard deviations of expectancy and therefore it has been concluded that "no evidence of rigging" has been found.


      As with the flop results, the general scatter is slightly to the right of the mean (i.e. in favour of the underdog) - 4 of the 6 sites tested have shown a deviation that is greater than 1 standard deviation but less than 2 standard deviations in favour of the underdog. This could easily be explained by variance.


      It has been suggested that a slight bias may occur naturally due to card removal effects. It is possible that this could be the case in these tests also. However, it should be noted that the exact effect of card removal is not fully understood and further research is needed before conclusions can be drawn in this respect.


      I have included the effects of card removal in the discussion section.


      Full results and explanation of the analysis can be seen here.
  • 1 reply