Who goes first in Matches
Laury Chizlett
don't the opening throws for each game in matches alternate between the
players? I have sometimes lost the right to the opening move 6 times in
a row (nothing wrong about this - the probability, 1/2^5, is only about
3% - of the order of a 1/36 shot) but when it is so easy to smoothe out
this element of chance I wonder why it is not done.
-- ^ To Liverpool St
^ Station & City
Laury | ^ | 1.5 miles
________________________| |
TRP Ridley Rd Street Market |__
35 Colvestone Crescent __________________ _ | | Dalston
London / / | |__| Kingsland
E8 2LG / / |A10 | Station
________________/ / | |
Tel: 0171 923 0244 Colvestone Cres / | |
Fax: 0171 923 1471 ____________________/ | |
35
Chuck Bower
Laury Chizlett <la...@trpdata.demon.co.uk> wrote:
>I am new to playing BG matches, so I might be missing something, but why
>don't the opening throws for each game in matches alternate between the
>players? I have sometimes lost the right to the opening move 6 times in
>a row (nothing wrong about this - the probability, 1/2^5, is only about
>3% - of the order of a 1/36 shot) but when it is so easy to smoothe out
>this element of chance I wonder why it is not done.
Interesting question. Maybe you are the first person to think of this.
(I'm only half-joking. ;) Probably since match backgammon evolved out of
money backgammon, the individual games in the match were just kept close
to money play (where each game starts with no one having an advantage). Of
course you could also alternate opening roller in money play....
If someone decided to implement your idea, I would strongly recommend
that doublets NOT be allowed on the opening roll. (Doublets should be
rerolled.) The reason is that doublets are such a large advantage that the
person who won the opening roll in the initial game (who couldn't receive
doublets, by current convention) would be at a CONSIDERABLE disadvantage.
Allowing doublets on the opening roll of all but the first game would add
even more luck than the situation you describe. IMO that would clearly
make the cure worse than the disease.
Chuck
bo...@bigbang.astro.indiana.edu
c_ray on FIBS
EdmondT
>that doublets NOT be allowed on the opening roll. (Doublets should be
>rerolled.)
>bo...@bigbang.astro.indiana.edu
> c_ray on FIBS
>
The Middle East style of BG has players throw the dice to see who rolls first,
then the winner picks up the dice and throws them again for the first play.
This allows doubles on the first throw.
Steen Ladelund
>In article <84$zIOAUj...@trpdata.demon.co.uk>,
>Laury Chizlett <la...@trpdata.demon.co.uk> wrote:
>>I am new to playing BG matches, so I might be missing something, but why
>>don't the opening throws for each game in matches alternate between the
>>players? I have sometimes lost the right to the opening move 6 times in
>>a row (nothing wrong about this - the probability, 1/2^5, is only about
>>3% - of the order of a 1/36 shot) but when it is so easy to smoothe out
>>this element of chance I wonder why it is not done.
> Interesting question. Maybe you are the first person to think of this.
>(I'm only half-joking. ;) Probably since match backgammon evolved out of
>money backgammon, the individual games in the match were just kept close
>to money play (where each game starts with no one having an advantage). Of
>course you could also alternate opening roller in money play....
Actually, tjhis would be more natural in moneygame, since here you could
fix the number of games to be even. In match play the number of games may
vary from 1 to 2*n-1, n being the match length. So chosing who should be first
to start, even in a random fashion, migth not decrease the element of chance that
much in match play.....
> If someone decided to implement your idea, I would strongly recommend
>that doublets NOT be allowed on the opening roll. (Doublets should be
>rerolled.) The reason is that doublets are such a large advantage that the
>person who won the opening roll in the initial game (who couldn't receive
>doublets, by current convention) would be at a CONSIDERABLE disadvantage.
>Allowing doublets on the opening roll of all but the first game would add
>even more luck than the situation you describe. IMO that would clearly
>make the cure worse than the disease.
> Chuck
> bo...@bigbang.astro.indiana.edu
> c_ray on FIBS
--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Steen Ladelund * office H2.14
Dep. of Theor. Statistics, * tlf: (+45) 8942 3499
University of Århus. * E-mail: lade...@imf.au.dk
Laury Chizlett
>>don't the opening throws for each game in matches alternate between the
>>players?
> If someone decided to implement your idea, I would strongly recommend
>that doublets NOT be allowed on the opening roll.
>
> Chuck
I agree. I was only thinking of the two players each throwing a die into
their right-hand board as per usual, rethrowing doubles, but instead of
the winner going first, the right to first move alternates.
Arising out of this: what _is_ the edge gained by going first? ie the
equity of a player, after winning the opening throw - whatever this
throw is, but before moving. Is it just the arithmetic mean of all the
equities of the best plays of the 15 possible opening throws? Is this a
question!?
Laury Chizlett
>then the winner picks up the dice and throws them again for the first play.
>This allows doubles on the first throw.
>
>Edm...@aol.com
I was brought up with the old English way: the winner can either play
the first throw, as normal; or throw again, being compelled to take the
second throw, double or not. This practice, together with not being able
to have more than 5 men on a point (you should see me play a 65!),
ruined my BG progress for years!
Rodrigo Andrade
(Chicago or Flint?). I totally support it. But I think this change would
take years to stick.
RODRIGO
Rodrigo Andrade
13/7 13/8 or
24/18 13/8.
I've seen people playing both ways. Pretty successfully sometimes.
RODRIGO
Chuck Bower
Laury Chizlett <la...@trpdata.demon.co.uk> wrote:
>>>I am new to playing BG matches, so I might be missing something, but why
>>>don't the opening throws for each game in matches alternate between the
>>>players?
><snip>
>> If someone decided to implement your idea, I would strongly recommend
>>that doublets NOT be allowed on the opening roll.
><snip)
>>
>> Chuck
>
>I agree. I was only thinking of the two players each throwing a die into
>their right-hand board as per usual, rethrowing doubles, but instead of
>the winner going first, the right to first move alternates.
>
>Arising out of this: what _is_ the edge gained by going first? ie the
>equity of a player, after winning the opening throw - whatever this
>throw is, but before moving. Is it just the arithmetic mean of all the
>equities of the best plays of the 15 possible opening throws? Is this a
>question!?
I will answer you questions in REVERSE order. Yes, this is a question.
(In fact, it is a GOOD quesiton.) Yes, it is the weighted mean. ("Weighting"
is needed if you include doublets since a specified doublet happens half as
often as a specified singlet. But if you don't allow doublets, then there
are 15 singlets, all of which are equally likely, so you just add up their
worth and divide by 15.)
Now, back to the question you REALLY wanted the answer for: how much
of an edge does the opening roller have. In cubeless equity units (no cube,
but gammons count twice and backgammons thrice) it is worth 0.038. Playing
that form of BG (no cube, but counting gammons...) the opening roller wins
51.2% of all games.
These numbers come from Jellyfish rollouts of REPLIES to opening rolls.
I have chosen only results where JF is forced to make the correct reply to
the correct opening where "correct" means JF rollouts give the highest
equity result to that play. The standard deviations on the above numbers
are about 0.001 on the cubeless equity and 0.02% on the winning chances.
(These are only statistical uncertainties. They don't account for the fact
that JF level-6 may be misplaying one side more than the other.)
I have not done such an extensive study of games where doublets are
allowed by the opening roller, since this is not the form of backgammon that
I play. However, I just did a quick (432 trial) JFv3.0 level-6 cubeless
rollout and the results (with standard deviations in parentheses) was that
opening roller has a 0.071(0.015) edge in equity and wins 52.5% (0.6%) of
all games. From this it looks like, if doublets are allowed, the edge to
the opening roller is doubled. (Pun?) Since doublets occur only 1/6
as often as singlets, you can see that they are considerably more valuable
at this VERY EARLY stage of the game (which qualitatively should not come
as much of a surprise).
While visiting some Greek friends last summer I was told that in their
coffee-bar games, for a series of games the winner of the previous game is
(further) rewarded by being allowed to roll first. You could also let the
loser of the previous game roll first, if you so desired. (This would be
similar to basketball, and many other sports where the opponent of the side
which just made a goal receives possession of the ball.)
The bottom line is that custom or tradition tends to dominate the
rules. As long as both sides are playing under the same rules (and here I
include the assumption that both sides are AWARE of what rules they are
using) then there is no intrinsic advantage to one person or the other.
It is unlikely that anyone is going to convince US tournament players
(or online server players) to alter the current rules even this much
(i.e. alternate opening roller). But lest anyone think we "Westerners"
are particularly hard-headed, how likely would it be that you could walk
into a Greek (or Armenian or Arabic or....) backgammon gathering spot and
convince them to change their rules?
Claes Thornberg
backgammon like table/board. The rules for starting the first game in
a match is that both players rolls one die each. The one who has
rolled the lowest die starts by rerolling both dice. (In this game
opening doublets are almost always worse than non-doublets). In
succeeding games, the loser of the previous game starts by rolling
both dice.
Regards,
Claes Thornberg
--
______________________________________________________________________
Claes Thornberg Internet: cla...@it.kth.se
Dept. of Teleinformatics URL: NO WAY!
KTH/Electrum 204 Voice: +46 8 752 1377
164 40 Kista Fax: +46 8 751 1793
Sweden
svaber
Rodrigo Andrade wrote:
Both are horrible off course!
Thomas
Rodrigo Andrade
Yep, I know, but if you can get away w/ them... See:
>> 13/7 13/8
Roll any kind of 6 or any 1 next roll and you get your bar point. Roll a 21
and you get your golden point. A 32 also gives you your golden point.
>> 24/18 13/8.
Roll another 6 and you escaped your 2 runners. Roll a 11 and you get a
4-point prime.
As you stated, and I totally agree, both of these plays are awful, but in
the unlikely event that you get away w/ them, you're in pretty good shape.
RODRIGO