| |
BGonline.org Forums
A discrete random walk
Posted By: Frank Berger In Response To: A discrete random walk (Maik Stiebler)
Date: Thursday, 23 July 2009, at 10:40 p.m.
sounds like a lot of work to answer your questions.
Related to continuous or discreet I trust your judgement, but I believe to reason about the possible lost due to a delayed doubling both may fit.
Do I overlook something or can I lose more due to a postponed double than to have lost the market. Let's assume that casher will be made, so you can't loose a game due to avoiding a cash. So my possible lost will be limited to the maximum market lost. But when I postpone my double several times I could get a penalty higher than the max lost due to a market loser.
And again. Lets assume I postpone my double for one roll and I'm 0.2 better off when I double now. Can someone try to calculate what the postponed double costs?
How much can I loose if I never double? When I use the continous modell I must reduce the winning probability of my opponent to 0 instead of to the cashing point. This amount is limited (and can be computed. I'm just to tired and lazy) If I add several penalties due to not doubling I can earn a greater penalty.
But even the model of "take the largest error" seem now not totally right to me. If doubling gains 0.2 I would penalize the player with .2 even if I double at the next turn. If we play this game several times how much equity I do give up? If I double at the next turn if the equity says double, I see only in the case of a market lost a smaller equity. So isn't the real error
The_cost_of_losing_the_market - (equity_where_I_should_have_doubled - drop_point)
I see that this has flaws too. If the equity drops and the player drops out of the doubling window and looses he would never loose his market and wouldn't be charged. That's bad too.....
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.