| |
BGonline.org Forums
cube him to death #2
Posted By: Daniel Murphy In Response To: cube him to death #2 (Leonardo Jerkovic)
Date: Tuesday, 8 June 2010, at 7:25 p.m.
In this position, I don't know who is Creamy and who is Brownie in real life and it doesn't matter. Brownie is winning 1-away 4-away post-Crawford but Creamy for some unknown reason hasn't doubled this game yet. Yet Creamy is a 200 Elo favorite? Ok, stranger things have happened. So let's say that Creamy who is trailing and doubling is the better player by 200 Elo points.
Your question is: should Brownie take this easy drop against an equal opponent because Brownie is -200 Elo points worse than Creamy?
No, I do not think so.
actually 200 elo diff is 35% to win whole 7pt match (according to elo formula)
Ok, let's say Brownie's MWC was only 35% when the match began. But the score is no longer 7-away all. The score is (or will be, after Brownie drops) 1-away 3-away post-Crawford. What does the Elo difference say Brownie's MWC is in a 3-point match? 40%. But that's for 3-away all, not 1-away 3-away post-Crawford. Hold that thought, we'll come back to it.
What's Brownie's MWC if he takes? Brownie can (a) win the game and match or (b) lose the game to lead 1-away 2-away with 45% MWC -- the Elo difference says says 44% at DMP, plus another 1% for dropping some obvious free drops. So Brownie's double/take MWC is 0.31 + (0.69 * 0.45) = 62%.
What's Brownie's MWC if he drops? Brownie will lead 1-away 3-away post-Crawford. Now Brownie can win the match by (a) winning a single game or (b) losing a single game but winning the next game.
Let's suppose that in the 3-away 1-away game, Creamy is a 56% favorite to win the game and 25% of his wins will be gammons. So we have
- Brownie wins single: 44%
- Creamy wins single: 42%
- Creamy wins gammon: 14%
Brownie's MWC = .44 + (.44 * .42) = .44 + .185 = 62.5%.
That's better than Brownie's double/take equity, so Brownie should drop.
Now, if we change Creamy's gammon rate so that it is higher than 25%, we can reach a point where Brownie's 1a/3a post-Crawford MWC is so low that Brownie should take the cube now. For instance, if Creamy's gammon rate in the 1a/3a game is 27%, then
- Brownie wins single: 44%
- Creamy wins single: 40.88%
- Creamy wins gammon: 15.12%
and Brownie's 1a/3a MWC = .44 + (.44 * .4088) = .44 + 0.18 = 62%. Then it might seem like Brownie might as well take the cube now, and if Creamy's gammon rate is even higher, Brownie should definitely take the cube now.
But what else is happening at 1-away 3-away post-Crawford? Why is Brownie a 200 Elo underdog? Does he make a lot of bad doubles? No problem here, he won't be doubling. Does he take a lot of bad cubes? No problem -- nearly every cube at 1a/3a pCr is a take. Does he play too timidly to avoid gammons? Again no problem -- avoiding gammons is a good strategy leading 1a/3a pCr.
Another point is that we don't know -- I don't know -- how much of Brownie's error rate is due to losing too many gammons, generally, or how well Creamy plays for gammon, or how much of Creamy's superiority is due to winning extra gammons against all opponents and against Brownie. So I don't know what Creamy's usual gammon rate is, or whether it's higher or lower than "bot-normal." I'll guess that Creamy's gammon rate against Brownie is higher than his gammon rate against an equal opponent. But I'll also guess that Brownie is aware that leading 1-away 3-away post-Crawford is not a score at which Brownie wants to lose a gammon!
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.