Imagine that we were living in the age when Raquel Welch
was wearing a loincloth bikini. I don't mean in the "One
Million Years B.C." I mean in 1966 B.J. (Before Jackoffski).

Imagine that, being a well trusted gambler mathematician
among his ilks, Axel dreamed up and published his "cube
skill theory" that became widely adopted and which said:

- If your last roll was a prime number and you are ahead by
more than 10 but less than 20 pips, then double.

- If the cube value is a multiple of the absolute value of the
square root of your pip count, then take; else drop.

Imagine that all the bots incorporated his skill formula and
that all the mentally ill gamblers strived to play like the bots
and achieved very low error rates as calculated by the bots.

With everything else being the same as today, i.e. checker
skill of the gamblegammon bots and the gamblegammon
human giants being the same as they would be in the year
2022 A.D. (Anno Domani), they all would have the same
ER/PR's and would get ranked in the same order as today.

None of you would know any differently and if Murat said
that the so-called "cube skill theory" was bullshit, none of
you would believe him.

MK

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

MK <[email protected]> writes:

Imagine that, being a well trusted gambler mathematician
among his ilks, Axel dreamed up and published his "cube
skill theory" that became widely adopted and which said:

- If your last roll was a prime number and you are ahead by
more than 10 but less than 20 pips, then double.

- If the cube value is a multiple of the absolute value of the
square root of your pip count, then take; else drop.

Will not happen, because

1. I do not publish bullshit.

2. If 1. is false, then still others will not believe my bullshit.

if Murat said that the so-called "cube skill theory" was bullshit,
none of you would believe him.

Compare and contrast with above.

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On Sunday, June 19, 2022 at 7:35:01 AM UTC+1, Axel Reichert wrote:

MK <[email protected]> writes:

Imagine that, being a well trusted gambler mathematician
among his ilks, Axel dreamed up and published his "cube
skill theory" that became widely adopted and which said:

- If your last roll was a prime number and you are ahead by
more than 10 but less than 20 pips, then double.

- If the cube value is a multiple of the absolute value of the
square root of your pip count, then take; else drop.

Will not happen, because

1. I do not publish bullshit.

2. If 1. is false, then still others will not believe my bullshit.

if Murat said that the so-called "cube skill theory" was bullshit,
none of you would believe him.

Compare and contrast with above.

The Axelisation theory does make sense and works well.
It deserves to be better known.
It is not good backgammon to Axelise in every racing position --
ideally intuition should be used if a player's experience is sufficiently well developed.

It is an interesting exercise to actually Axelise in your own games.
There's a small element of the theory that doesn't quite make sense,
but things that don't make perfect sense are to be expected.
What doesn't make sense about it?
Suppose that one of the players has only an acepoint stack with no other checkers.
Then having 2n checkers in the stack is the same as having 2n-1 checkers in the stack.
So it makes no sense for the method to ignore the above equivalence and use the exact
number of acepoint checkers.
But that's a very small issue and it's just as much of an issue with racing algos generally
apart from EPC perhaps.

Paul

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On June 19, 2022 at 12:35:01 AM UTC-6, Axel Reichert wrote:

MK <[email protected]> writes:

Imagine that, being a well trusted gambler mathematician
among his ilks, Axel dreamed up and published his "cube
skill theory" that became widely adopted and which said:

Will not happen, because
1. I do not publish bullshit.

Whether they themselves believe in them, many people do.
You may not publish but you subscribe to bullshit of others.

More than any other reason, I used your name in my example
in order to draw you into the discussion. ;)

2. If 1. is false, then still others will not believe my bullshit.

Don't be so sure. People believe all kinds of bullshits, especially
if they are elaborate and come from prophets they naively trust.

I wonder if I can talk you into doing another simple experiment
to show this time that even the cubeless equities calculated to
pretentious levels of accuracy are mostly bullshit (more so at
the early stages and less so at the late stages of games).

This experiment would be even easier than your Murat mutant
cube skill experiment but more fundamentally important since
cube decisions, as well as checker decisions, are based on the calculated/estimated equities.

You can use my approach of mutant making the worst move in
its first turn, either as its opening or reply roll, and call it Murat
mutant-2 if you wish.

https://montanaonline.net/backgammon/xg.php

https://montanaonline.net/backgammon/xg/Worstfirst/Worstfirst.txt

https://www.youtube.com/watch?v=CPhBQl5ttmo

If I remember right, you had said you didn't know "C" language
and I dodn't know how you modified Gnubg for your experiment
but this would require adding only one line of code to the mutant
to say "If first move then pick the worst move" and let it play as
usual afterwards without any other changes.

In a past thread you had suggested that I should play a session
of cubeless games which didn't make sense for my purpose of
debunking the very "cube skill" but it would make sense in this
new experiment which would focus on early cubeless equities
as well as early adjusted cubeful equities being all bogus. I bet
mutant will win more than expected by its checker error rate. I
wouldn't be surprised if it even wins close to 50% if not more.

In cubefull games, mutant will drop most cubes immediately
following its initial huge blunders, (losing only 1 point, with no
risk of high cube SPP), but may and most likely will still manage
to win at least more than expected by its checker+cube error rate.

Wouldn't you all want to know the results of such experiments?

If you would dare, of course, since the implications may end up
proving devastating to your faiths in position equity bullshit...

MK

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)