Forum Archive :
Using a match equity table
Paul Epstein writes:
> As a mathematics student and a relative newcomer to the game, I
> would be interested to know how decisions about doubling are affected
> by the scores in a match. Suppose [match to 21, leader holding 4-cube
> at 15-8, no gammons possible.]
> How high does A's probability of winning need to be before A should
> decide to double?
> How high does B's probability of winning need to be before B should
> accept a double offered by A?
[Although I'm sure a lot of you know how to solve this, I'll post a
detailed description of how to solve such problems so that those of you
who don't can see how it's done. Please forgive any verbosity...]
You can't really answer this question without a feel for a player's
chance of winning at any given score. For this you use a match equity
table, which lists winning chances at a certain numbers of games away
from match for each player. For example:
Here's the one in _World Class Backgammon, Move by Move_ by Roy Friedman.
1: 2: 3: 4: 5: 6: 7: 8: 9:
1: 50 66 75 80 83 87 89 92 94
2: 34 50 59 64 72 77 82 85 88
3: 25 41 50 56 63 69 74 78 82
4: 20 36 44 50 57 63 68 72 77
5: 17 28 37 43 50 56 62 66 71
6: 13 23 31 37 44 50 56 60 65
7: 10 18 26 32 38 44 50 55 60
8: 8 15 22 28 34 40 45 50 55
9: 6 12 18 23 29 35 40 45 50
The numbers on the left and top are number of points needed by each
player, and the values in the chart are percentage of match wins.
For those of you unfamiliar with such a table, keep in mind that leading
3-2 in a match to 5 gives you the same chance to win as leading 5-4 in a
match to 7. So a position is usually referred to as x-away, y-away.
Back to our problem. A is leading 6-away, 13-away. This table doesn't
go that high, so there are two possibilities: gut instinct or another
My estimate at 6-away 13-away says something like 85%. Kit Woolsey's
new table says 82. He's put a lot of research into it so let's believe
him. Now as a matter of fact, the match equity at the current score
doesn't mean much. What we really want is to look at possible outcomes
of the match. 2-away 13-away gives a match equity of 96. 6-away 9-away
gives an equity of 67. (The table above says 65, but lets use all Kit's
So A's current chances are 96p + 67(1-p) or 67 + 29p, where p is the
chance of winning this game.
A should only double if A's match equity goes up after a double and a
take. (We've already ruled out gammons in our assumptions above.)
Now what if there's an 8-cube? The score will be a win for A or 6-away
5-away, which is 43% for A. So A's chances have gone to 100p + 43(1-p)
or 43 + 57p.
Where is the break even point? When the two values are equal. So we have:
67+29p = 43+57p
24 = 28p
24/28 = p
or about 86%! So A has to have a very big lead in the game to double at
this score with the cube at 4.
But, there's another factor we need to consider -- should B take, and if
so what should B do? If B takes, then if A wins the game, A wins the
match, but if B wins the game, B is only about even. Since a win for A
already wins the match, B loses nothing by doubling. So B should double
back immediately! Thus whoever wins the game wins the match.
After a redouble by B, A's chances are 100p. So we plug that in:
67+29p = 100p
67 = 71 p
67/71 = p = 94%
So A has to have a 94% or greater chance of winning to gain by doubling!
Now there's another question though, should B accept? If B drops, B's
match equity is 4. After a take and redouble, B's chance is 100-100p.
Let's see where these are equal:
4 = 100 - 100p
96 = p
If you recall that p always has stood for A's chance of winning, B
should acept with a 4% chance or greater, of winning...
So... A should double with a chance of 94% or better, B should accept if
A's chances are below 96%, and drop if A's chances are above 96%.
This is one of those cases when A can't use all of the value of the cube
by doubling. (A needs 6 to win, turning the cube gives B the chance to
win at leat 8, and in fact 16!) Whenever you're in this situation, you
have to be very very slow to double!
Hope this helps (and is mostly correct!)
-michael j zehr
- Constructing a match equity table (Walter Trice, Apr 2000)
- Does it matter which match equity table you use? (Klaus Evers+, Nov 2005)
- Does it matter which match equity table you use? (Achim Mueller+, Dec 2003)
- Does it matter which match equity table you use? (Chuck Bower+, Sept 2001)
- ME Table: Big Brother (Peter Fankhauser, July 1996)
- ME Table: Dunstan (Ian Dunstan+, Aug 2004)
- ME Table: Escoffery (David Escoffery, Nov 1991)
- ME Table: Friedman (Elliott C Winslow, Oct 1991)
- ME Table: Kazaross (Neil Kazaross, Dec 2003)
- ME Table: Kazaross-XG2 (neilkaz, Aug 2011)
- ME Table: Rockwell-Kazaross (Chuck Bower+, June 2010)
- ME Table: Snowie (Chase, Apr 2002)
- ME Table: Snowie (Harald Retter, Aug 1998)
- ME Table: Woolsey (Raccoon, Apr 2006)
- ME Table: Woolsey (Kit Woolsey, May 1994)
- ME Table: Woolsey (William R. Tallmadge, Jan 1994)
- ME Table: Zadeh (Jørn Thyssen, Mar 2004)
- ME Table: Zorba (Robert-Jan Veldhuizen+, Dec 2003)
- ME at 1-away/2-away (crawford) (Fabrice Liardet+, Nov 2007)
- ME at 1-away/2-away (crawford) (Ian Shaw+, Apr 2003)
- Match equities--an alternate view (Durf Freund, Oct 1994)
- Neil's new numbers (neilkaz, Aug 2011)
- Neil's numbers (Kit Woolsey+, Oct 1994)
- On calculating match equity tables (Neil Kazaross, July 2004)
- Turner formula (Gregg Cattanach, Feb 2003)
- Turner formula (Stephen Turner, June 1994)
- Using a match equity table (Michael J. Zehr, June 1992)
- Value of free drop (Neil Kazaross, Oct 2002)
- Which match equity table is best? (Martin Krainer+, Oct 2003)
- Which match equity table is best? (Ian Shaw+, Dec 2001)
- Why use a match equity table? (Kit Woolsey, Feb 1999)
- Worth memorizing? (Alef Rosenbaum+, Feb 2003)