| |
BGonline.org Forums
Is a .09 cube error "huge"?-- *Spoilers*
Posted By: Albert Steg In Response To: MCG on first Backgammon Video Podcast--link below (Phil Simborg)
Date: Monday, 15 August 2016, at 9:35 p.m.
I was intrigued by the conversation around 41:00 and forward, where Matt makes a comment about a .091 cube error being “not Huge” or somewhat “marginal,” whereas Phil is asking whether that isn’t a pretty substantial blunder, being over .08. Matt acknowledges that in the long run, equity is equity, yet he feels that the cube error in this position isn’t so severe or substantial compared to a checker play error of the same magnitude. He may simply mean that cube errors tend to express a much larger range of error than checker plays (it's a lot easier to make a .25 cube error than a .25 checker play error), but I think there's maybe more to it . . .
==== SPOILER ALERT (answer to the video problem may be implied) =========
This is one of my favorite backgammon topics along the line of ‘is all equity the same?’ and the question of theoretically correct ’robot’ play vs. practical, human play. When I’m going over cube positions I harvest from a bg session, I’m never too bothered by missed cube opportunities where I could have gained .05 or even .08 or so by cubing in a position that would only push my equity up to to something in the +.500 to .700 range. Against humans, there is always some possibility of your opponent making a cube error — and I like to give my opponents as many “opportunities for error” as I can. Positions that wind up under .700 after a Take tend to be pretty obvious takes and often only marginally profitable doubles.
Of course, if you’re in a high-volatility situation where the outcome will essentially be decided on your roll of the dice, you want to double even with a tiny improvement in equity -- and I really don't want to miss out on cubes that will gain .10+. But otherwise, if the doubling decision seems borderline, I ask myself “Will doubling give my opponent a substantial opportunity to make a mistake?” along with “Might I be making a mistake by doubling?” I’m generally pretty happy to give up a small equity improvement in exchange for a decent opportunity of inviting my opponent to Drop a scary Take or Take a steamy/optimistic Drop on a subsequent turn.
In the video position, the market-losing sequences are pretty few, so, being unsure whether it’s a proper double and seeing it’s a super-clear take, I’d be inclined to take a roll, and welcome the prospect of sticking my opponent with a tougher decision a roll or two down the road. So, like Matt, but maybe for a different reason, I'd shrug off a failure to double here against a human as 'not that big a deal.'
This is different from a checker play, where if you make a sub-optimal move, you’re stuck with it, whatever your opponent does (leaving out the often chimerical idea that making a sub-optimal play will ‘complicate’ a position and lead your opponent to blunder). Usually the immediate/subsequent dice rolls will punish you for a bad checker play, without the opponent's help. (Typically only a smallish subset of the 21 possible rolls will give your opponent a genuinely enhanced "opportunity for error").
On the other side of the coin, I’m always happy to see my opponent puzzling over whether to cube or not in a position I know is an easy take. For once, at least, I know *I* won’t be the one making an error, and the error monkey is on his back.
Albert
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.