Thread Rating:

Mission146
Mission146
Joined: May 15, 2012
  • Threads: 133
  • Posts: 15280
January 14th, 2022 at 8:07:13 AM permalink
Looking at what I think we could all agree on:

1.) If DI were possible, then it would be a matter of employing precise physical mechanics that would cause the dice to operate...and they would still be operating randomly...but sufficiently and in a reliable way that the natural probabilities of a, "Random and Unbiased," roll of the dice is sufficiently altered to have a slight bias.

2.) When it comes to, "Precise physical mechanics," anyone who has ever attempted to play any sport in an organized, or even informal, way knows what I am talking about. It's the same reason that you can have a skilled foul shooter in the NBA, but nobody has ever been 100% lifetime from the free throw line. Other individual pursuits. Whatever example you want.

3.) So, you would have to have someone that has either proven or believes that the, "Natural," probabilities can be altered to one extent or another. Combine with that they would have to have the physical ability to do this in a reliable way long-term. It's obviously easy to toss a pair of dice, but to do it in a way as to sufficiently alter the probabilities such that the house edge (any bet) turns into a player advantage, even if someone wanted to assume that it's possible, to assume that it would be easy is fantasy land.

Like a foul shot, you're doing a precise physical action that is, in theory, roughly repeatable...but I think the precision would have to be even more than this. By a lot.

With all of that, I think it's maybe possible...in theory. Whether or not anyone actually does it successfully I have seen no evidence that would convince me as yet.
https://wizardofvegas.com/forum/off-topic/gripes/11182-pet-peeves/120/#post815219
Mission146
Mission146
Joined: May 15, 2012
  • Threads: 133
  • Posts: 15280
January 14th, 2022 at 8:11:31 AM permalink
Also, if someone were out to prove the theory...I think the best way to do so would be to start with a mechanical device and go from there.
https://wizardofvegas.com/forum/off-topic/gripes/11182-pet-peeves/120/#post815219
TumblingBones
TumblingBones
Joined: Dec 25, 2016
  • Threads: 29
  • Posts: 482
Thanks for this post from:
Mission146
January 14th, 2022 at 8:22:42 AM permalink
Those "grey areas" are the reason I suggested the digital twin approach. Those "invisible level of forces" would no longer be invisible.

Maybe we should take up the challenge as a WoV community project? I'm sure there's more than one WoV member with access to a craps table. Anybody out there with the hardware chops to build the sensor rig to collect the verification data? I could do the AI/ML part. We would still need somebody to program the simulation engine, maybe someone with experience writing video games and programming a game engine.
My goal of being well informed conflicts with my goal of remaining sane.
unJon
unJon
Joined: Jul 1, 2018
  • Threads: 14
  • Posts: 3146
Thanks for this post from:
Mission146
January 14th, 2022 at 8:23:29 AM permalink
Quote: Mission146

Also, if someone were out to prove the theory...I think the best way to do so would be to start with a mechanical device and go from there.
link to original post



I disagree. I like TBís idea of modeling it.

I would also break it into levels:

1) DI on a hard felt surface rolling dice a few inches.
2) DI on a hard felt surface rolling dice several feet.
3) DI on a springy felt surface rolling dice several feet.
4) DI on a springy felt surface rolling dice several feet and hitting a flat wall.
5) DI on a springy felt surface rolling dice several feet and hitting a diamond bumpy wall.

And see where DI breaks down.
The race is not always to the swift, nor the battle to the strong; but that is the way to bet.
TumblingBones
TumblingBones
Joined: Dec 25, 2016
  • Threads: 29
  • Posts: 482
Thanks for this post from:
Mission146
January 14th, 2022 at 8:41:26 AM permalink
Quote: unJon

Quote: Mission146

Also, if someone were out to prove the theory...I think the best way to do so would be to start with a mechanical device and go from there.
link to original post



I disagree. I like TBís idea of modeling it.

I would also break it into levels:

1) DI on a hard felt surface rolling dice a few inches.
2) DI on a hard felt surface rolling dice several feet.
3) DI on a springy felt surface rolling dice several feet.
4) DI on a springy felt surface rolling dice several feet and hitting a flat wall.
5) DI on a springy felt surface rolling dice several feet and hitting a diamond bumpy wall.

And see where DI breaks down.
link to original post



I don't think Mission's take is entirely incompatible with mine. Collecting real-world data is a necessary step or you can't validate the digital simulation has having the correct physics. I do, however, suspect that use of a mechanical device is not required, although it would be helpful.

The way I see it the goal is to eliminate, or at least minimize, all the arguments about sample size, the consistency of the sample throws, and is it all BS due to the variability of the human in the loop. Instead I propose a analysis in which the set of sample throws can be highly inconsistent and where there is no need to even make controlled throws. Instead the only purpose of the sample throws would be to collect enough data to validate the model. This would directly address unjohn's "levels" (e.g., dice hitting close to the wall vs several inches out). As long as the dice are closely monitored at the point of impact in terms of velocity, rotation, angles, etc., the data collected should let the AI/ML figure out the rest.
My goal of being well informed conflicts with my goal of remaining sane.
SOOPOO
SOOPOO
Joined: Aug 8, 2010
  • Threads: 113
  • Posts: 8467
January 14th, 2022 at 12:02:29 PM permalink
Quote: TumblingBones

Those "grey areas" are the reason I suggested the digital twin approach. Those "invisible level of forces" would no longer be invisible.

Maybe we should take up the challenge as a WoV community project? I'm sure there's more than one WoV member with access to a craps table. Anybody out there with the hardware chops to build the sensor rig to collect the verification data? I could do the AI/ML part. We would still need somebody to program the simulation engine, maybe someone with experience writing video games and programming a game engine.
link to original post



We did this already. Ahigh graciously offered his craps table for the challenge. A member, it's so long ago I don't remember his name, claimed he could roll less than 1/6 7's by some amount. I bet him he couldn't. I won the bet for I think it may have been $200? Ahigh made some sort of mechanical thrower which was also unable to alter the random nature of dice outcomes.

Someone can dig up that thread....
AlanMendelson
AlanMendelson
Joined: Oct 5, 2011
  • Threads: 130
  • Posts: 4197
January 14th, 2022 at 12:05:51 PM permalink
Quote: SOOPOO

Quote: TumblingBones

Those "grey areas" are the reason I suggested the digital twin approach. Those "invisible level of forces" would no longer be invisible.

Maybe we should take up the challenge as a WoV community project? I'm sure there's more than one WoV member with access to a craps table. Anybody out there with the hardware chops to build the sensor rig to collect the verification data? I could do the AI/ML part. We would still need somebody to program the simulation engine, maybe someone with experience writing video games and programming a game engine.
link to original post



We did this already. Ahigh graciously offered his craps table for the challenge. A member, it's so long ago I don't remember his name, claimed he could roll less than 1/6 7's by some amount. I bet him he couldn't. I won the bet for I think it may have been $200? Ahigh made some sort of mechanical thrower which was also unable to alter the random nature of dice outcomes.

Someone can dig up that thread....
link to original post



Ahigh was not and never was a DI.
ChumpChange
ChumpChange
Joined: Jun 15, 2018
  • Threads: 57
  • Posts: 2696
January 14th, 2022 at 1:52:13 PM permalink
I find I lose 5 to 10 PL bets while waiting for my turn to throw at the table so it is up to me to right this wrong when it is my turn. I really should just bet the DC while waiting for my turn.
tuttigym
tuttigym
Joined: Feb 12, 2010
  • Threads: 9
  • Posts: 706
January 14th, 2022 at 2:32:57 PM permalink
Quote: AlanMendelson

That's right... because you've never been at a table with a true DI.

You've never seen a player do it so you repeat what you've read by others who also have never seen it.

DI is not something supernatural. However it is a very rare skill in part because few have actually learned or practiced it.

If DI were taught in public schools beginning in the fourth grade there would be enough professionals to fill up the tables at casinos. But it's not taught or practiced and few have the basic physical abilities to even make a decent attempt at it.


Sorry, Mr.Mendelson, wrong answer. Those selling this "snake oil" of DC/DI were and are scammers fleecing the desperate, gullible, and easy entrapped to enroll in their classes and seminars for big $$$. Neither they or their "students" have ever been successful at real play, but they can tell some wild, tall tales of fantasy successes.

tuttigym

p.s. I played craps with DeMango who has the most consistent and accurate toss imaginable that I have ever seen, and his success is marginal at best. BTW he told me that he used to practice for hours. Not sure about that now. It has been a while
TumblingBones
TumblingBones
Joined: Dec 25, 2016
  • Threads: 29
  • Posts: 482
January 14th, 2022 at 2:38:32 PM permalink
I remember, as well as respect, the effort AHigh made to explore this (thread). What he did, however, is not the same as what I am proposing. He pretty much did the same thing that the UNLV folks did: build a mechanical arm and see if, after a sufficient number of throws, the statistics indicate that some degree of DI behavior is present. That is not at all the same as the experiment I am suggesting.

My main claim is that purely statistical experiments based on repetitive throws of dice, whether by humans or machines, are of little if any practical value:
  • the time and effort it would require to generate statistically irrefutable data set is to great
  • even if DI was validated, the experiment will not result in any strategy of use to a practitioner of DI (e.g., optimal trajectory, EV for a given set, strategy for adapting to an individuals shooting style/limitations)


Rather that a statistical experiment I am proposing the use of AI and machine learning. The key is the digital twin component:
Quote: wikipedia

A digital twin is a virtual representation that serves as the real-time digital counterpart of a physical object or process..... A digital twin also can be used for monitoring, diagnostics and prognostics to optimize asset performance and utilization. In this field, sensory data can be combined with historical data, human expertise and fleet and simulation learning to improve the outcome of prognostics.


Using the approach should allow us to get answers to questions the "mechanical arm" approach can't deal with, such as:
  • assuming someone is a "perfect" shooter, for a given combo of impact point, trajectory, and set, what are the various EV for each bet on the table?
  • what is the impact on EV for a given level and type of inaccuracy (e.g., shooter undershoots optimal impact point by 1" on 50% of throws)?
  • for a given shooter's individual style and capabilities, what is the optimal strategy?

Unlike either a human or mechanical shooter, an AI shooter can address all of these questions by making 100s of millions of throws and then analyzing the resulting data sets. The only use of data from human and mechanical throws would be to tune the model during the set-up phase.
My goal of being well informed conflicts with my goal of remaining sane.

  • Jump to: