This week I’ve begun writing the ML framework on which to train the poker AI. I’ve implemented a system to run Monte Carlo simulations, and am working on implementing stochastic node selection for training. In addition, I’ve implemented the core of the CFR portion of the algorithm which will determine the model weights.
Andrei’s Status Report Feb 14th
Andrei worked on collecting knowledge on the rfid product classifications to make sure to purchase the correct types, In addition I am currently waiting to receive the autocard dealer to be able ot hack it to follow game mechaniccs. This coming week I will purchase the RFIDs to test them out and in addition I will be testing the auto card dealer capabilities.
Ben’s Status Report Feb 14th
This week I got into developing the base poker game software, and working on making sure that it is reasonably optimized to be able to run large scale simulations on it effectively. In addition, I spent a lot of time working on the design presentation for next Monday.
Team status report for Feb 14th
Right now our biggest risk is the reliability of all the pieces we have ordered that we haven’t gotten to test yet. The automated dealer may be very difficult to integrate or the RFID may not be reliable enough to track chips or cards. We have backup plans to build our own simpler dealer, and reading chips by color without RFID, but hopefully back up plans will not be needed once we get our items.
No changes to our schedule right now, we are mostly waiting for some critical parts to be delivered, and making progress on the things we can like the chip return system.
Part A: Since this project is to provide casual players at home with a realistic gameplay to either practice against ai bots or friends and bots there is likely no large psychological problems. That does not mean that on an individual basis users could not find themselves gambling using this product and therefore could pose a psychological impact on those who gamble using our product. There are otherwise no affects on a users welfare or safety using this product. (Written by Andrei Da Silva)
Part B: For social factors, we are hoping that this project allows people to connect with their friends over an in person game, rather than being glued to your phones. We want to make poker more accessible for groups that may not know exactly how to play or struggle to find a dealer, but are looking for this kind of activity for connection. (Written by Sam Istvan)
Part C: Our product will hopefully be economically much more reasonable for a majority of the target audience – people who want to have a functional automated home poker system. Currently, competing products are enormously expensive, running $5000+ per table, and don’t have physical chip and card integration. We hope that our product would be able to target more casual audiences that appreciate the physical experience of poker and don’t want to spend exorbitant amounts of money for the convenience of a dealer.
Sam’s status report for Feb 14th
This week I have been working on the design of the chip return. I have started printing small 3d printed prototypes to test how my mechanism will work. Below is an image of two of the pieces I’ve made. One is a track that lets a poker chip roll through it past the color sensor I am using to sort the chips. The other is a wheel that I will use to break up a bulk input of chips into a sequential order. I am currently designing the way chips will enter my sorting system, I am trying to do a similar idea to a basic coin separating machine but by color as all the poker chips are the same size. I am a little behind as we haven’t received our poker chips yet so I dont know they’re exact dimensions, which is important to some of my prints, so hopefully that will be delivered soon.
https://www.youtube.com/watch?v=7ILHtAPY29I&themeRefresh=1
Ben’s Status Report for Feb 7th
This week I went more in-depth researching advanced poker algorithms and their implementation, and made decisions on what I should actually implement. The goal is to have a model that can be competitive vs casual or amateur poker players, not necessarily to be able to beat professionals. Thus, the basic idea is to take portions of the Libratus and Pluribus papers and either simplify or implement features in a less granular way, so that the model can both be simpler to implement and take less time to train. At the moment, the simplifications are likely to be a much less refined subgame solving protocol similar to the Pluribus one, as well as rounding bet amounts, and other simplifications. I am on schedule, and next week I will be looking to implement the base of the poker game on which I can begin to implement monte carlo methods.
https://www.science.org/doi/10.1126/science.aao1733
https://www.science.org/doi/10.1126/science.aay2400
Andrei’s Status Report for Feb 7th
After completing the proposal, I begun doing research on auto shufflers and auto dealers to be able to find out which parts are better off purchase and which parts should be made by hand instead. As of right now I am considering purchasing the entire auto dealer and shuffler and hacking it to work the way we want but I am still not 100% sure.
Team Status Report for Feb 7th
The most significant risk is if RFID does not read the chips fast enough. We are going to do some testing this week to see if we can limit this problem by using color sensors.
We have no changes in our design or schedule as we are just starting, we are hopeful that we will face no slow downs from ordering.
Sam’s Status Report for February 7th
This week I did our Proposal presentation so a good chunk of my week has been preparing speaking notes and practicing. I’ve also started design for the chip return, first by printing some stl’s online to get an idea of how other people have created things to deal with poker chips.
I think my progress is on schedule, we haven’t made significant progress on any portions of our project but our schedule is aiming for the most intensive work to be done starting this Monday, which is when we will be ordering and assembling.
This week I want to do testing on if RFID works fast enough and accurately enough for this project, and how much color sensors would help. I also want to start designing the 3D file for printing the chip collection, so I need to at least create how chips will be fed into my sorter.
Testing the Website
hello world!