[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums

Sconyers database sez...

Posted By: Chuck Bower
Date: Saturday, 16 April 2011, at 9:22 p.m.

In Response To: DMP rollout (Timothy Chow)

I think bearing 2 gets us hit in about 2% more games...

Yes, according to the old (3 CD) Sconyers database, the probabilities of getting hit after each play are 0.09727 and 0.11957.

Looking deeper, the probabilities of getting hit on the first roll and probabilities of getting hit on a later roll are (0.08299,0.01557) vs. (0.10562,0.01560). Note that the probabilities of getting hit after the immediate roll are almost identical (difference is 3e-5 or one part in 500).

The six big rolls (where one play get hit and the other doesn't) lead to (4/9)*(6/36)*(11/36) = 2.2% advantage for the non-rip play. The 22 moves that leave a blot next turn regardless of how you play favor the rip by (4/9)*(22/36)*9% = 2.4%. (8/36 rolls don't leave a blot next turn for either play. I'm calling those a wash.)

So from this calculation it looks like ripping is (very slightly) favored at DMP. There are some assumptions here and a couple worth mentioning:

1) the CRP values for p(win) vs. # of checker remaining are correct.

2) those CRP values are for a closed board and this position isn't a closed board, so trailer can potentially try to get a second checker sent back.

However, to test the #2 I rolled out the original position (at DMP) with a closed board (acepoint covered by the 5-point spare) and it still gave the same result as your rollout, a difference of 0.009 in cubeless equity in favor of the non-rip play.

So I'm stumped. See a flaw in my reasoning/calculation?

Messages In This Thread

 

Post Response
Your Name:
Your E-Mail Address:
Subject:
Message:

If necessary, enter your password below:

Password:
 

 

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums is maintained by Stick with WebBBS 5.12.