\$EV below 0 or above the highest possible payout!

• Bronze
Joined: 26.04.2011
Hello I wanted to know why it's possible for a game to be -\$EV of for example: -\$15

In game that is not possible since you have a set buy-in so why does HEM not correct that to -\$0

Same goes for the possibillity of winning \$120 where the max win is \$100.

How should we interpretate these games and is there a way to adjust these varations?
• 4 replies
• Black
Joined: 03.07.2008
You got all-in with AK against AA and won. Then you got all-in with AK against AA and won.
• Bronze
Joined: 26.04.2011
I get the concept and how you can reach that -\$EV but Since it's not possible to really lose more than you put into from the start on shouldn't your \$EV not go above the max payout and not below the \$0?
• Bronze
Joined: 17.06.2010
I wrote an article on this phenomenon in backgammon called "Unbiased Nonsense." Backgammon analysis programs may report that you had more than 50% mwc in net luck so that you were a 120%:-20% favorite over your opponent in a match, which is logically impossible. Basically, making the "corrections" so that the adjusted results of individual tournaments make sense would introduce a bias which would mean that the adjusted results would no longer converge to the right value.

The all-in luck adjustment is only supposed to remove part of the luck in the game, and it doesn't need to remove any other luck to be useful, even though on occasions you can tell that the luck which is not removed is large. That's what is happening in those examples.

Since the adjusted result of 1 tournament or even 30 tournaments do not mean much anyway, just accept that it is an artifact of the correct method of luck adjustment, not a flaw.
• Bronze
Joined: 26.04.2011
thanks for your response Pzhon!