repeated dice roll

[badgolferman]

I realize this would never happen in a game but I think it would be
interesting if an analysis has ever been done.

If there is was only one combination of dice you could get for the
entire game which wasn't a double what would it be? Conversely what
combination would you not want?

I would think a high/low combination would be best, or an even/odd
combination. Something like 52 or 63.

[Timothy Chow]

On the principle that backgammon is a race, my first instinct would
be to opt for 65. I would not want 21.

On further reflection, I see that one disadvantage of a 6 is that it
will often a bad number to roll when I'm on the bar. Still, I would
opt for 65.

I'm assuming that the opponent doesn't know that I'm always going to
roll the same number. If my opponent knows then I think I'm likely
to lose the game regardless. I can be closed out with just a two-point
board.

---
Tim Chow

[peps...@gmail.com]

A former boss of mine once said "You don't ask enough questions."
Here, you don't make enough assumptions/ questions.
As well as asking/ assuming about whether your opponent knows whether your roll is repeating,
you've also got to ask/ assume about whether you know that you will always roll the same number.

Even if your opponent doesn't know, your opponent is quite likely to make the 6 and 5 points anyway and then hit,
as this is a major opening goal.

Paul

[badgolferman]

Neither the opponent nor the player know.

In my view the roll would have to be something benign with a wide
range, that's why I voted for 63 or 52.

[Axel Reichert]

"badgolferman" <REMOVETHISb...@gmail.com> writes:

> Neither the opponent nor the player know.

But the opponent gets random dice? You could do some tries with manual
dice for one side and have the bot roll the other one.

In my opinion the roll must have more than average pips (8 1/6), so 54,
63, 64, and 65 are in the game for me. Not sure about the doublets,
perhaps blocked to easily, but we know from an older discussion that the
large doublets (even if not in consistent order) were evaluated as the
luckiest rolls on average. So after trying to predict the typical course
of a game, I will opt for 66.

Axel

[peps...@gmail.com]

That kind of reminds me of my trip to McDonalds yesterday (shame on me)
where I mistakenly went to the wrong counter, which wasn't the correct counter
to go to for customers placing their orders.

Bizarrely, instead of simply pointing out the correct counter, the employee
said "Are you ok?" so I replied "Maybe I'm missing something but what I'm actually
trying to do is to place an order. Is there a way I can do that please?"

badgolferman specifically rules out doubles in his OP.
At badgolferman's restaurant (which fortunately is much better than a McDonalds), we need to follow his procedures for ordering.

I wouldn't want to forgo the game plan of hitting and then priming or closing out.
From that perspective 65 is problematic in that only hits checkers at distances of 5/6/11
whereas, for example, 21 hits checkers at distances of 1/2/3 which is far more hitting.
However, the problem with 21 is that even after the priming and closing out has worked,
you have the big disadvantage of consistently rolling > 5 less than your opponent.
So you can lose races even though they look overwhelmingly won.
I want an ace for hitting. I like 61 which hits at distances of 1/6/7 and which doesn't lose
that much in the race. Since it's rare to have both the six and the ace made, we won't dance much
with this roll, and that's an important point.

61 for me (No, that isn't my age in case you're wondering. And nor is 16).

Paul

[Axel Reichert]

"peps...@gmail.com" <peps...@gmail.com> writes:

> badgolferman specifically rules out doubles in his OP.

Oops, overlooked this. Then it will be 65 for me.

> I wouldn't want to forgo the game plan of hitting and then priming or
> closing out.

"Backgammon, in essence, is a race."

> 61 for me

A sample of two games each (GNU Backgammon playing itself with manual
dice) led to one easy win and one closer loss for 65 and two gammon
losses for 61. Proves nothing, of course.

Axel

[peps...@gmail.com]

On Saturday, September 18, 2021 at 8:27:25 PM UTC+1, Axel Reichert wrote:
> "peps...@gmail.com" <peps...@gmail.com> writes:
>
> > badgolferman specifically rules out doubles in his OP.
> Oops, overlooked this. Then it will be 65 for me.
> > I wouldn't want to forgo the game plan of hitting and then priming or
> > closing out.
> "Backgammon, in essence, is a race."
>

...
Race, schmace!
My new plan is to select 21, do lots of juicy hitting and priming, and double the opponent out before
it gets to a race.

Paul