Variance
Posted by Apoth
Video Player is loading.
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Resume Video
Start from Beginning
Watch Video
10
Train with the best
Learn from the Run It Once Pro Team with two new instructional videos every day.
Loading 10 Comments...
great stuff!
Thanks, glad you found it interesting. I know people enjoy seeing actual poker being played so I try to make the focus of most of my videos either live play or hand review but I do feel obliged to cover the odd topic like this (but they can be tricky to do effectively).
Really good content that I think will help a lot of players recalibrate their expectations on variance, myself included.
One idea that would be fun, perhaps better suited for the RIO YouTube channel, would be a video analyzing some of the hot runs in poker history and quantifying exactly how hot they were. A few that come to mind would be: Harrington back-to-back Main Event FTs, Hastings vs Isildur, Dwan on High Stakes Poker, Chidwick or Sontheimer in the recent US Poker Opens, etc.
Thanks for the video!
It would be very difficult to do this because we are missing a lot of info. I don't have a lot of information about winrates or field sizes and sample sizes are very small.
I'll keep your post in mind though and if I can get something interesting out at some point I will.
Great content, thank you!
Other than the major insight that getting out of small stakes tournaments (where fields are larger) is beneficial because our variance goes down, do you have other applications for this knowledge on variance? For example, is it just a good idea to play around with it and understand it, so that we have the correct mindset when losing? Maybe it's beneficial to apply this knowledge to our average buy-in/average field size and calculate our variance for the year?
Any other applications on how you personally use this knowledge would be great to hear.
Thanks again!
The applications tend to be a lot more general and not quite as specific as perhaps you were hoping.
For example: I pointed out the difference between the std dev in some games.
You might decide that you don't want to play tourneys professionally because you want something more reliable/lower variance.
You may also decide that you don't want to play PLO for that same reason and instead stick to NL cash because you need the semi reliable income (either because you need $ to make rent or because your mental game isn't strong enough or w/e).
You can also look at a bankroll of 10k and say "well, if I want to go broke < 5% of the time I can either play 200 NL or a combination of 100PLO/50PLO so I need to beat PLO for > 2x the winrate to make the same hourly" etc...
It also helps you when deciding to do things like hire a coach or buy mtt action.
You can determine whether or not their results are like variance. If I show you mtt results with a 30% ROI over 1k sample, how significant is that? 5k sample? 100k sample?
If I show you cash game results over 250 hands how significant is that? 1k hands? 10k hands? 100k hands? 1M hands? etc.
Hi good video, especially for people with no statistical background.
I just wanted to clarify one thing: you repeatedly mentioned that about 70% of observations fall into 1 standard deviation. That's not correct, assuming a normal distribution 68% of all observations fall into +/-1 standard deviation from the mean, so actually twice as much.
I really enjoy your content and look forward to the next one!
Yea, that's correct. Sorry I misspoke. 70% of all obersvations fall into 1 standard deviation in either direction. Thanks for taking the time to clarify this!
Hey Cris, great video, love math stuff!
Apparently, the calculator doesn't take into account the difference of variance between turbo or regular tournaments. How do you think it could change the results if our sample is only regular tournaments or if it's only turbos, etc..?
Turbo vs. regular would simply affect ROI--assuming the same player pool, whatever your ROI% is in a regular structure, it would go down by some degree in a turbo, and then down some more for a hyper. So, although it doesn't take structure variance into account directly, it does so indirectly via the ROI parameter.
Be the first to add a comment
You must upgrade your account to leave a comment.