500 samples is nowhere close. Actually, you first must define the accuracy at which you want to know your ROI, then I can tell you what's the probability of you having this roi +/- % with so and so many samples.
For instance. Assume you are winning 56 games out of 100 on average. That gives you 6,67% roi. Now after playing say 7000 games, in 95% of cases you will have achieved ROI somewhere between 4 and 9.5%.
get the idea? It is my understanding, that you will never know your "true" roi at lower stakes simply because if you are winning you will never play 10k games at the level of $5. As you move up and still win, you will settle down at some level and collect more samples you will be closer to know you ROI .
1000 samples will give you from 1 to 13% ROI some 95% of the time. Nowhere close to seeing your TRUE ROI. See, the buffer around the theoretical value of 6,67% is much narrower with 7000 samples vs 1000 samples.
at the beginning of your poker carreer though, you can't make any assumptions about this theoretical 56 game win out of 100. You just don't know anything. See, even if you assume you know your true ROI it will take a long while to get there in practice. But at the beginning you just play some samples, if you win, you move up. And if you can't win, you become a sub at husng.com :))
Also, you must bear in mind, that as you move up, your ROI expectations will go down.
It's also worth noting that even if you play 2k-3k games at any given level, you'll still be improving a tremendous amount over those sample most of the time.
So you could start out with a 4% true ROI, then as you improve it climbs to 6-7% over a few thousand games. This is especially common for newer players that have a lot more leaks to improve and skills to develop.
My advice for true winrates has always been to be realistic about your skill set and always look to improve your game. As long as you don't risk going broke with bankroll management, the worst thing that can happen is that you'll find out you're not running or playing so well in a level and you have to drop down a buyin level or two. Then you just move back up as you continue to grow as a player, so it's not such a bad scenario, and it's basically the worst case if you are responsible and realistic.
Would 500 HU Sngs give a reliable figure?
500 samples is nowhere close. Actually, you first must define the accuracy at which you want to know your ROI, then I can tell you what's the probability of you having this roi +/- % with so and so many samples.
For instance. Assume you are winning 56 games out of 100 on average. That gives you 6,67% roi. Now after playing say 7000 games, in 95% of cases you will have achieved ROI somewhere between 4 and 9.5%.
get the idea? It is my understanding, that you will never know your "true" roi at lower stakes simply because if you are winning you will never play 10k games at the level of $5. As you move up and still win, you will settle down at some level and collect more samples you will be closer to know you ROI .
1000 samples will give you from 1 to 13% ROI some 95% of the time. Nowhere close to seeing your TRUE ROI. See, the buffer around the theoretical value of 6,67% is much narrower with 7000 samples vs 1000 samples.
at the beginning of your poker carreer though, you can't make any assumptions about this theoretical 56 game win out of 100. You just don't know anything. See, even if you assume you know your true ROI it will take a long while to get there in practice. But at the beginning you just play some samples, if you win, you move up. And if you can't win, you become a sub at husng.com :))
Also, you must bear in mind, that as you move up, your ROI expectations will go down.
settle down in comfort, it is a looong ride ;)
Hope that helps.
It's also worth noting that even if you play 2k-3k games at any given level, you'll still be improving a tremendous amount over those sample most of the time.
So you could start out with a 4% true ROI, then as you improve it climbs to 6-7% over a few thousand games. This is especially common for newer players that have a lot more leaks to improve and skills to develop.
My advice for true winrates has always been to be realistic about your skill set and always look to improve your game. As long as you don't risk going broke with bankroll management, the worst thing that can happen is that you'll find out you're not running or playing so well in a level and you have to drop down a buyin level or two. Then you just move back up as you continue to grow as a player, so it's not such a bad scenario, and it's basically the worst case if you are responsible and realistic.