>>16919489
A language model would know

You need to be able to count cards in your head

>>16919490
There is actually a benchmark for LLMs playing poker. But they're not close to being as good as humans

>>16919493
Wrong game

>>16919489
I already did this over 20 years ago. I have over a million dollars in lifetime winnings in poker. I used to coach my exgf after I met her streaming poker on twitch and wrote a document for her on how to be a winning player. If you want a copy let me know and I will post it

>>16919624
>over 20 years ago
The winning strategy back then was don't be a full-fledged retard.

>>16919624
>I have over a million dollars in lifetime winnings in poker.
okay but how much did you lose?

>>16919489
If you want to apply mathematics to make money, the first and most important step is not gambling. Every single game, no matter what strategy you employ, has a negative expectation value over a long enough playtime.

>>16919655
Pssh, details, details.

>>16919493
>>16919698
>I don't know what poker is.
Why do retards like you insist on contributing? You waste time solving the captcha only to post something that serves no purpose than to demonstrate your idiocy.

I watched casino royale and didn't understand a single thing about this game other than it's gay and retarded and only useful in mega-scale money laundering.

>>16919655
if you claim to have "winnings" that means minus any losses you may have incurred. People who arent morons just know this without having to be told

>>16919489
btw knowing the math is only the beginning, only half the game. you should start with a book like play poker like the pros by phil hellmuth after you finish learning the math. Also the math is pretty simple if you understand statistics, you dont need a book to teach it to you.

>>16919640
it isnt any harder today then it was then. this is what is known "young dipshit" bias, where young people think nothing existed before they heard of it. Professional gamblers have existed for 1000's of years broccoli headed zoomtard. Granted more people play today than ever before and the math is easier to learn today than ever before but there are always fish anywhere you go. There may be more good players, but there are also more fish

>>16920186
thats not how gambling works larper

>>16920240
1. you are stupid
2. LARPing what? Being a winning poker player? If you think that is something worth LARPing to strangers about you must be really fucking stupid. Which you obviously are since you cant figure out a statistical advantage to be a winning player yourself

File: IMG_20260226_054617651_HDR.jpg (3.7 MB)

3.7 MB JPG

>>16919489
>poker
tangential project. picrel. think about your data streams and what they really are. for craps and dice control, if I write some super elaborate program to run a bunch of test strategies, what am I really doing? those throws are not from my hand. the fixed focus usb cameras are for my craps table. the idea is to do a 1-hour session, throwing dice in a specific set, maybe a 1212 set, and have a video recording of the set, arc of throw, backspin, rebound energy, axis of rotation stability, and twists either off-axis or invert-axis.

the thing is, paper charting is not enough, because I have noticed already more backspin on my right die due to my pointer finger on my left hand having more "power" than my ring finger, which causes a difference in the two, therefore causing a skew in the statistics of my "digital hand" after the throw results are extracted from the video file because fuck me I am not writing a realtime analyzer to do that. just record 4 cameras, super simple image capture of the dice and the pips to derive the motion into a 3d cube under computation. that way I can also tell if my right die has an unstable rotation axis compared to my left die, and adjust. it also tells me warmup time, running time, and cooloff time for casino play.

the point of all this blathering is this- you can math the fuck out of everything, and even play robots in a game, but all that will do is make you good at playing nonhumans, and not even showing what your actual poker stats are compared to the ideal statistical poker stats you'll get from calculations. good for calculating your "digital poker hand" just like my "digital craps hand" to be used in strategy building and simulation, also practice. I figure I need about 100k throws to get a good idea.

you might want to take a more data based approach instead of a mathematical idea approach.

>>16920738
>cheating
also try to figure out how to log all the games you play in competition. for craps, most players fail to keep as close a record of their casino throws as they do their practice throws, because the same difference in paper trading the stock market vs. money trading the stock market. for me, a 1213 is different than a 1312 set because I take into account the left or right die in my setting, making even more permutations, but really just making a playbook for myself of what I can actually throw. in casino play, logging throws on paper is more or less impossible alone, and most casinos won't let you have someone do that for you. one good way, without using a cheat device is to record audio. in our wondrous age, pulling text from audio everytime I saw "two six" or "one one" or some other slang to indicate the dice in a left/right format, so long as I don't cross them in the throw or lose track.

that way, I can capture that data bias too and exploit that edge. from there, I can prune the playbook of throws and strategies which work in practice, but for whatever reason simply do not work in the casino. stuff like that. I am totally new at this, but the greats I read, like Scoblete and Wong give the same advice. analyze your own logs. look for your own patterns, play to your personality. you might find that for whatever reason, your poker playing falls apart once you start playing above $100, forcing you into a new low risk strategy...

>>16919490
>muh ai
kys

>>16920186
>if you claim to have "winnings" that means minus any losses you may have incurred
We get this, but there are a huge number of degenerate gamblers who excitedly bag about their 'wins' while not subtracting losses. These people are mentally barely children (the typical child understands basic math well enough to see why this is silly) but since they exist you do need to clarify that you aren't one of them.

File: 1770659834078544.gif (745.7 KB)

745.7 KB GIF

>>16920738
>image capture of the dice and the pips to derive the motion into a 3d cube under computation
I'm interested. Care to share any more about this?
also interested in the software or programming side of things. I'm a software dev

anyway, I like the way you think!

File: hesback.jpg (35.7 KB)

35.7 KB JPG

>>16919489
how to play poker tho.
I'm a mad scientist but not interested in those type of games.

>>16920795
>software dev

peep these while I write up my software plan

https://www.youtube.com/watch?v=zFiubdrJqqI
https://www.youtube.com/watch?v=YZkLQsv3huo
https://www.youtube.com/watch?v=vtyYRMK_oy8

>>16920795
>software

cameras- I bought fixed focus cameras with a rigid mount to do easier image subtraction between frames to get

https://www.youtube.com/watch?v=NSS6yAMZF78

a 2d bounded image of what the voxel extraction needs to be. this also has the added benefit of super easy vertex extraction from the cube as it flies through the air, since the pips are all identical, and color variance from transparency is an easy filter out. having 4 cameras means total 3d coverage for the 3d voxelized photogrammetry extraction.

from there, the voxel itself can do approximated planes to define the bounds of the dice, per frame, and then use that to extract the mathematicle center of the dice, being precise cubes, is isotropic, making super easy calculation of the "ideal" thrown arc, and then calculate the "ideal" vertices from the approximated planes on top of the voxels. from there, the axis of rotation can be easily analyzed for any nutation, or wobble of the axis... among other things like rate of backspin.

what is a plus is having different colored dice in the setup means easier for tracking finger influence automatically. no need to do any paper logging. my plan is to setup a batch of 50 presets, like 1212, then 1213, then 1312. the software catches it all via simple 2d image recognition on those frames, but can just as easily extract the full motion no problem.

I used to get paid to do this, so it is now my retirement thing.

>>16920795


more details about the overall concept here

>>16909769

I am building the image frontend right now, with acoustic cameras to follow. writing the baseline software for acoustic and visual cameras to synthesize their flock input from a set of orbiting drones above a city into a 3d photogrammetric realtime voxel scene at the base station, with acoustic extraction via 3d sound voxels from the acoustic cameras flying on the same airframe. sort of like ShotSpotter, but in the air. you can gain the capability to see, for example, if a car starts up in a parking lot somewhere from altitude due to the sound mapped onto that in 3d from the acoustic camera alongside the visual camera on the drone. that is faster than a thermal can pick it up, and can be offloaded to an autodetect function in software, better than a wide stare infrared solution.

also looking to install this in a nearby airport which is undergoing huge renovation right now. that makes a simpler setup because the acoustic and visual cameras are fixed in place, simplifying the ground station calculations which are done on the GPU as a voxel shader.

gotta start with craps for shits and giggles, then get the voxelized photogrammetry pipeline on the gpu, then flying.

>>16920802
>mathematicle
kek