Quote:
Originally Posted by Josem
A couple of weeks ago, I posted in this thread asking for details of what sort of statistical analsysis people would like to be able to perform on this stuff.
I'd like to generate a Hold'em Manager report that you can conduct on your own hand histories - free of any interference from the relevant poker site.
What would you like tested?
Further, it makes no sense for the site to do the analysis themselves: the whole reason there is demand for the analysis is because you do not trust them. Therefore, it MUST be done by yourself (or some other third party) or it has no value.
It'd be really awesome if this discussion could progress at all: it's pretty frustratingly circular and hardly seems to develop.
I feel the same frustration, Josem.
Let me be the first to say that a discussion about what kinds of analysis could/should be done is deserving of its own thread and not jammed up in this 640+ page monstrosity.
The most obvious analysis could be done on all-in preflop hands. This should also be the easiest to implement by programmers.
I'd love to see some analysis on flop textures, from the most dangerous to the most serene. Are there 2 of a suit? 3? How many cards to a straight? Is there a pair on the flop? Trips?
Then analysis beyond that on how the turn relates to the flop. Does it complete a possible straight? Flush? Pair the board? And then the same analysis on the river.
Even further could analyze the known hands that saw all these stages and their relationships to the flop (gut shot draw, trips, 2 pair, etc), turn, and river.
Beyond that a way to analyze the money (or tournament chips) involved at each stage to see if that has any correlation with anything.
I understand I'm being a bit vague, but I think these points are worthy of discussion and if PT3 or HEM could do these types of analysis, we'd be entering into territory of being able to postulate on the effective fairness of the deal.
Last edited by smithcommajohn; 10-21-2009 at 11:34 PM.
Reason: typo