“J” Backgammon Terms

A elimination event, usually with a large entry fee, in which only the winner and runner-up receive prize money.

The fact that an improvement in the opponent's position can make redoubling correct in a position in which the player on roll owns the cube and has one remaining chance to redouble.

[Named for Oswald Jacoby, who proposed the rule.]  A rule popular in money play which says that gammons and backgammons (2) count only as a single game if neither player has offered a double during the game. The Jacoby rule is not used in match play. The rule speeds up play by eliminating situations where a player avoids doubling so he can play on for a gammon. See post by Daniel Murphy.

A game once popular in France in which players start at diagonally opposite corners and move around the board in the same direction.  See: How to Play Jacquet.

A formula devised by Rick Janowski for estimating match equity (1) at a given score. If d is the difference in match score and t is the number of points (4) the trailing player has to go, then the probability of the leading player winning the match is .5 + .85_d_ / (t+6).  See also: Neil's Numbers and Turner's Formula.

A collection of formulas devised by Rick Janowski for estimating cubeful equity from cubeless equity. The basic formula for cubeful equity (between take points) is:

CF = CL\*(1 - x) + CE\*x

where CF is cubeful equity, CL is cubeless equity, CE is cubeful equity assuming all doubles are perfectly efficient, and x is a number between 0 and 1 that measures the cube efficiency. Typical values for x range from 0.55 to 0.8.   See also: Janowski's Takepoint Formula.

A formula devised by Rick Janowski for estimating your take point given your cubeless probability of winning the game. The basic takepoint formula is:

       2L - 1  
TP = -----------  
     2W + 2L + x

where TP is the cubeless equity of your take point, L is the average value of your cubeless losses (e.g., −1, if you can't lose a gammon), W is the average value of your cubeless wins (e.g., +1 if you can't win a gammon), and x is a number between 0 and 1 (typically 0.55 to 0.8) that measures cube efficiency. See Janowski's article, Take-Points in Money Games.

The first commercial neural-net backgammon program (1994) after TD-Gammon.   Website: Jellyfish Backgammon.

Potential for awkward rolls on a future turn.  See also: Double Jeopardy.

The standard deviation of the difference between two rollouts: JSD = sqrt(SD1*SD1 + SD2*SD2). A measure of how statistically significant the result is.

An exceptionally good roll, especially a roll that reverses the likely outcome of the game; a roll much luckier than average.

Joint standard deviation.

Affectionate name for a player's farthest-back checker.