Quote:
Originally Posted by big bwalz
Coming soon, currently the question is how exactly to analyze them and what is a reasonable/acceptable sample size to make a determination one way or another, if there is one. If there isn't one mathematically which I don't think there is then what would the industry standard be? the standard 100k cash game hands, 10 milly? It's being worked out but any suggestions are welcome and I realize it's a touchy issue and there will be widely varying positions.
How many you need depends on what you're testing. Let's use one of the examples from the OP, because it's easy to test and doesn't even need hole cards:
Quote:
Originally Posted by GreenBliss420
Ok, so considering nobody has touted this network more highly than me, I think I have every right to create a thread about my experiences after about 20k hands on WPN in the last three weeks. All @ 5NL, all with PT4 running.
[...]
The board? I'm sorry - I play thousands of hands on other sites. There is absolutely no question that there are more low board runouts on this site than I have seen anywhere else, on TV, any hand histories, anywhere. At all. I can't tell you how many times I'm holding broadway suiteds and the board runs out 447 or 223, monotone or four flush. Over and over, and over. Again, I'm sorry but ...I just don't see any resemblance between how the boards run out on BCP and how they run out anywhere else I've watched or played poker.
So there we go, easy thing to test: How often do the boards run out "low" on WPN, compared to how often that should happen? OP was able to spot it just casually playing 20,000 hands, so we can say that's a big enough sample to check for it. Easy peasy.
Quote:
Originally Posted by big bwalz
It's a little more complicated than that lol but thanks. There's also the fact of hands being mucked @ showdown and how that may or may not affect the EV #'s.
EV is basically meaningless, but I'll admit that WPN's ****ty policy on mucking at showdown makes it slightly harder to test hands than sites that aren't terrible software wise. You can just use hands that got all in at some point before the river instead, though.
Quote:
Originally Posted by big bwalz
According to what/who? See where the problem lies? It's high level math stuff I don't have the background for, as I said it's being looked into but suggestions are welcome.
Great news for you: there's a
forum right here on 2+2 that has many posters who are incredibly well versed in statistics and probability.
Quote:
Perhaps a poll should be taken in the general online forum to get an agreed upon industry standard, just a random thought.
A poll is pointless. All you need to do is figure out how many hands you need to be within a certain confidence interval. Generally 5 standard deviations is considered out of the ordinary enough to be definite proof of something.
Last edited by otatop; 04-10-2015 at 11:40 PM.
Reason: forgot an s on boards