Quote:
Originally Posted by jschell
How much do you discount a player's rating based on how many hands you have for them? Especially with the "tendency of the players at the table to be winners or losers" rating, we all know that good players can have downswings for thousands of hands. Thus, I'm concerned that rating a player based on their earnings will produce inaccurate results.
This is something we have thought about. We only give a positive rating or negative rating if the player in question is what we call a long term winner or a long term looser. We need a substantial number of hands to place them in one of these categories. The number we use is not large enough for us to give different ratings based upon different win/loss rates but its enough to nail down a basic trend.
Perhaps it might be a good idea to remove this rating if the player is barely winning or loosing over any number of hands since a small win or loss rate would not be statistically significant unless we had a huge amount of data on that player. What do you think?
Dave