F***, F*** (yup, Five, Four)

LuvChild · Joined: 28 Jun 2007 Posts: 30 Location: Philadelphia, PA

Ok... think I got it figured out-- the imaging that is...

In the posistion below I am black playing a 5-4 from the bar; down 1-0 in a 3-pt. match... how would everyone out there play this?

Apparently I made a big mistake here by hitting but I don't understand why.

Does GNU think that I need to be committed to a back-game at this point? I don't think so, but if so, doesn't even that start to go out the window if the piece coming off my mid-point gets hit? I know timing is crucial to a back-game, but so is having more than the 6 and 8-points made. and I think sending a second white checker back spreads out their position and gives me a chance to win this going forward. Even if they make my five point they can't hold it long with so many pieces already rolling forward... still don't believe this (GNU) is correct and I know the bots have the occasional "glitch" with a back-game... am I just whistlin' dixie?
_________________
I judge myself a success not by my accomplishments, but rather by the number of times I fail, but continue to try.

puzzle_t0rico · Joined: 20 Apr 2007 Posts: 72

Hi there

Indeed, when you make GnuBG analyse to a 3-ply, the best move is bar/16

I do not know why this is the best move.
Though, there is only a -0.033 equity difference
_________________
If the brain were simple enough to be understood, we would be too stupid to understand it.

LuvChild · Joined: 28 Jun 2007 Posts: 30 Location: Philadelphia, PA

I'll admit the play I made isn't a serious error in equity, but I still find the play fascinating simply because it is so far from what I would consider a "conventional" play-- and because I'm of the opinion that it's because of the back-game possibiities so I was wondering if anyone out there is familiar with a similar condition.

I'm used to running things through only a one-ply analysis since I'm still learning the game; so one-ply at a time makes sense for me personally at this point. I still try to think about these things multiple rolls ahead however. I'm surprised still that hitting isn't the right play; espically when you still have the opponent's 5-point held and running makes sense if the back-game theory is correct because if hit it doesn't send you back as far, keeps you 13-pt. checkers safe for future building, and offers the possibility of a double-shot on the return.

Maybe this is just one of those oddballs I need to revisit in 6-12 months one I understand things a bit better.
_________________
I judge myself a success not by my accomplishments, but rather by the number of times I fail, but continue to try.