Reply to thread

Message: [QUOTE="Funtast, post: 6876848, member: 1030354"] You need 2 things: - samplesize and - winrate With these two numbers you can calculate your true possible winrates and their %es. Easy example: - samplesize 25k - measured winrate 10bb/100hands Possible true winrates: 5-15bb/100 at 16.6% 16+bb/100 at 66.6% 4-bb/100 at 16.6% If the samplesize doubles to 50k hands it might look like this: 7-13bb/100 at 16.6% 14+bb/100 at 66.6% 6-bb/100 at 16.6% These are confidence intervals. You can never tell your true winrate with 100% accuracy but the bigger the samplesize the closer you get and the smaller the interval gets. Something like 95% 9-11bb/100 for example or 99,5% 10bb/100 +-.5. [/QUOTE]