Handicap playing: "pass" and "pick".
Leo Bueno
Is anyone aware of a reasonable way to handicap a superior player when
playing a weaker opponent?
In Chess, for example, I recall in my early days playing vastly
superior opponents who would give me a material advantage (from a
queen at a the beginning to a pawn when I got better).
It occured to me that in Backgammon something like a handicap can be
accomplished by giving the weaker opponent the right to PICK a
favorable roll or PASS on a bad roll. Let me explain.
A "PICK" would allow the weaker player the option at any time in a
game (or match) to *select* a dice roll of his/her choice, i.e., pick
any roll he/she deemed most favorable.
For example, suppose both players were bearing off and your opponent
has 2 checkers on the 1 point and you have 3 checkers on the 6 point.
It's your roll and you have the right to a "pick"; you would of course
then and there pick a double 6 and win the game.
A "PASS" would allow the weaker player the right to *forego* playing a
particularly bad roll and simply hand the dice over to the opponent to
take his/her turn, i.e., the weaker player would pass on the bad roll
and not be forced to make moves which would weaken his position.
Say I were playing a 10 point match with someone of the caliber of Kit
Woolsey or Bill Robertie; they would, for example, spot me 5 PASSES
and 3 PICKS to make the match more competitive.
Note that the inferior player would still have to excersise good
judgment in deciding when during any given game to PASS or PICK, as
needed.
Does this sound like a reasonable way to even out players of
significantly disparate skills?
==================================================================
Snail-mail: P.O. Box 440545, Miami, FL 33144-0545 (USA)
Test De Cubanidad: http://www.accesspro.net/leobueno/cubanidad.htm
Castro Fall Poll: http://www.accesspro.net/leobueno/adiosfidel.htm
==================================================================
Michael J Zehr
In article <34918918...@news2.accesspro.net>,
Giving settlement odds is a more common way of evening out skill levels
in match play. This is how most BG rating systems work. (When two
players of unequal rating play on FIBS, GG, etc., if the lower-rated
player wins, the loser "pays" more rating points than the lower-rated
player would "pay" after a loss.)
This has the advantage that the settlement is in theory fair (if both
players have a rating near to their "true" rating, whatever that means)
and you're playing vanilla BG.
You could also spot the weaker player a point or two.
(In money play having odds for paying changes cube decisions, so it is
less frequently used. In match play it doesn't change the cube
decisions.)
In your description above, keep in mind that PICKS are vastly more
powerful than PASSES. Several times a game you're probably in a
situation in which there are one or two horrible rolls or root numbers,
other than a bad dancing number. (A pass doesn't save you from dancing
on a one point board since your options are to make no play or to use
your pass and still make no play. <grin>) The pass is only useful for
those times when you roll the root number. However several times a game
you have perfectas that have huge equity swings, and you get the chance
to use your PICK at every one of these.
The existence of picks completely throws off the cube action. At a
score of -4:-4, what's the right cube action if you're soon to clear
your midpoint against a 2-point anchor with questionable timing? If
your opponent has a pick left, that might make this no-double/take,
because the player with a pick left can double as soon as a shot
appears, even if it's a 1-17 shot.
Finally, what is the purpose of this? Traditionally in BG (and in many
activities) an inexperienced person is willing to pay an experienced
person for the chance to learn. But by having picks and passes, the
expert is no longer playing real BG, they're playing something else. So
playing them doesn't help you learn BG.
So, to answer your first question, yes there are reasonable ways to
handicap the stronger player, by paying odds instead of a 1-1
settlement. This evens out the expected gain/loss while keeping the
game the same. The handicapping method you suggest radically alters the
nature of the game.
-Michael J. Zehr
Kevin Bastian
I totally agree with Michael. Picks and passes would so alter the game
that:
A. You would be playing "picknpassgammon" not backgammon. (If you WANT to
make up a new game, go for it, but don't call it backgammon, and don't be
fooled into thinking your playing backgammon.)
B. The weaker player wouldn't learn how to play better backgammon, at least
not in all respects (although they might become a very strong
picknpassgammon player).
C. The stronger player just might get weaker (possibly even poorer when
later playing real backgammon against players who haven't had their
instincts fouled by all the adjustments required in picknpassgammon).
As Michael points out, FIBS (and I presume most/all of the backgammon
servers with rating systems) handle this situation fairly well by using
odds for the match settlement. If FIBS estimates that, based on our
respective ratings, Jelly is a 70% favorite to beat me in a 9 point match
(I'm making these numbers up, didn't calculate them), then if we play 100
such matches, I'm likely to win 30 of them, Jelly likely to win 70. The
rating system is set up so that after going 30-70 against Jelly, I'll walk
away with approximately the same rating as when I started, as will Jelly,
since my victories will gain me proportionately more ratings points than my
losses will cost me.
(Note: This all assumes that the player ratings are reasonably accurate and
that the ratings formula is a good forecaster of outcomes with players of
differing abilities. I have often seen both of these assumptions disputed.
If you agree, you might want to adjust either the formula or the players'
ratings "to taste." Adjusting the ratings differential and using the
existing formula would be the simpler way to go. But the approach is still
valid.)
The beauty of this system is that neither player has to change their
backgammon playing style. (Weaker and stronger players do sometimes adjust
their game styles against players of differing abilities, but the way they
do so now is just another part of the game of backgammon, not
picknpassgammon.) Imagine all the mental gyrations a world-class player
would have to go through when considering a double if s/he had to think
"well, let's see, he's got 1 pick and 2 passes left...or is that 2 picks
and 1 pass...or did he use a pass already...no, that was last
game...hmmm...he's 70% likely to dance here...oops...no, he can use his
pick to roll 61 and nail me...oh wait, if he has TWO picks left, then a 61
followed by a 66 and I lose...well, I can't double...in fact, if he's smart
he'll double me next turn and I must pass...gheesh"
I like FIBS/Michael's system. If you use it, you should take into account
the two factors that the servers use in determining the match odds: match
length, and relative rating differential between the two players. An
article explaining these factors in some detail is at
http://www.northcoast.com/~mccool/fibsrate.html if you're interested, and
it includes a link to François Hochedé's FIBS rating calculator. Thus 55-45
odds might be appropriate for a 1 pointer, yet it might take a 75-25 split
to be fair in a much longer match, the theory being that the longer the
match, the more likely that skill will outweigh luck in determining the
outcome.
An alternative approach would be to spot the weaker player one or more
points at the start of the match. For example, if I'm going to play a
stronger player in a 7 point match, letting me start ahead 1-0 or 2-0 would
help equalize it. This would be relatively simple to do equitably. First,
estimate the favorite's winning chances in the match of the given length,
perhaps using the FIBS rating calculator. Then check a match equity table
and find a score with that approximate winning percentage for the underdog.
Then the two factors will roughly cancel each other out and the outcomes
over time should be fairly even.
At least with either of these two approaches, you're still playing
backgammon in all its glory. Neither player is playing under any different
circumstances than they will normally encounter. The underdog can still
knock off the world class player -- not with a bizarre pick or pass -- but
with a combination of BACKGAMMON skill and luck, the same two factors the
champ needs to knock off the novice.
Good dice!
Kevin Bastian
KevinB on FIBS
Morten Daugbjerg Hansen
>>
>>Does this sound like a reasonable way to even out players of
>>significantly disparate skills?
Well, instead of writing something myself, to explain why this is tottally
useless, ill repeat Michael Zehr's words:
*****************************************************************
>The existence of picks completely throws off the cube action.
> by having picks and passes, the
>expert is no longer playing real BG, they're playing something else. So
>playing them doesn't help you learn BG.
>the handicapping method you suggest radically alters the
>nature of the game.
*****************************************************************
Play the expert to learn, not to invent some other game...
IF you were to alter the game, a better choice would be to rearrange the
starting position, something like adding a checker to the newbie, but being
a newbie he would probably not feel it was an advantage, nor be able to play
it to advantage....other choice could be to move one of his men from the
rearest point to the midpoint, or other rearrangements....then the game
would still resemble backgammon quite a bit except for the opening ..
My $0.02
Morten Daugbjerg :-)
Julian
In article <01bd0809$9f6d24e0$4a78...@REMOTE.sctcorp.com>, Kevin
Bastian <kba...@don.t.spam.me.ibm.no.spam.net> writes
>I totally agree with Michael. Picks and passes would so alter the game
>that:
>
>A. You would be playing "picknpassgammon" not backgammon. (If you WANT to
>make up a new game, go for it, but don't call it backgammon, and don't be
>fooled into thinking your playing backgammon.)
Indeed.
Incidentally, wouldn't 'picknpassgammon' be a pretty trivial game with a
doubling cube - if only one player has any picks left then he can cash
the game as soon as he has a freak roll available...?
---------------------------------------------------------------------------
Julian Hayward 'Booles' on FIBS jul...@ratbag.demon.co.uk
+44-1344-640656 http://www.ratbag.demon.co.uk/
---------- if you cut here, you'll probably destroy your monitor ----------
Stephen Turner
Julian wrote:
>
> Incidentally, wouldn't 'picknpassgammon' be a pretty trivial game with a
> doubling cube - if only one player has any picks left then he can cash
> the game as soon as he has a freak roll available...?
>
Doesn't that depend on whether the opponent might have a freak later? So
you're probably pretty reluctant to use the first pick, leaving yourself
with fewer than your opponent.
--
Stephen Turner sr...@cam.ac.uk http://www.statslab.cam.ac.uk/~sret1/ Statistical Laboratory, 16 Mill Lane, Cambridge, CB2 1SB, England
"The Bishop of Huntingdon and Postman Pat each opened new school extensions"
(Cambridge Weekly News, 28-May-97)