This site uses cookies to improve your browsing experience. By continuing to browse the website, you accept such cookies. For more details and to change your settings, see our Cookie Policy and Privacy Policy. Close

Errors and Game Selection by Alcswab

    • StandardBacking
      StandardBacking
      Bronze
      Joined: 20.02.2015 Posts: 109
      In statistics there are 2 type of errors. A type I error is a false positive (believing something is true when it’s false), whereas a type II error is a false negative (believing something is false when it’s true). Depending on what kind of error you want to make it’ll influence how you approach and make decisions.

      Specifically, these 2 types of errors are very relevant to game selection. When you’re deciding whether to add in a tournament to your schedule the most* basic question you’re asking is “Am I +EV in this game”. The null hypothesis, or the base assumption, is that you’re -EV. Meaning you need evidence that you’re +EV to disprove the base assumption.

      If you misinterpret the evidence and think that you’re +EV in a tournament when you’re -EV then you’ve made a Type I error. Meaning you’ll play a MTT where you’re -EV.

      If you misinterpret the evidence and think that you’re -EV in a tournament when you’re +EV then you’ve made a Type II error. Meaning you’re skipping a MTT in which you’re +EV.

      Generally, if you’re conservative in your game selection you should be trying to minimize your Type I errors and setting a higher bar to prove that you’re +EV in a tournament.

      Whereas, if you’re aggressive in your game selection you should look to minimize your Type II errors, and setting a lower standard of proof that you’re +EV in a tournament.

      There’s no right answer when considering which way to go. Committing the Type I error will put you into -EV games and drastically increase your variance and lower your ROI, committing the Type II error will leave money on the table.

      You decide.

      [Edited by SDK1987]
  • 2 replies