Baccarat Follow the Last anyone?

Ohoh the kangaroo taxman Kometh......

When you fly here there will be a bucket of dark beers waiting, Cerveza Negra, and a tub full of.ladyboys , tanned skin (morenos)

 

I will run another cycle on the Wizard's Simulator. Near the end of the week. I am implementing some scheduled software changes over the next few days. I will remove the restraints and let it loose. I will remember to display the stats at the end. I never record any of this, I do not look at what it just played. I did not think that I changed any bets after a tie. If I did then that was my mistake on entering not the machines.
If the play called for a $200 bet and a 5 ties came out, the bet would have remained the same.

As for stopping it is arbitrary (stopping is more from eye strain & the lure of snacks). I did not want to make the video to long. In real life I stop constantly. Since there is no one else playing on the rng, it picks up right where I left off. I do not need a starting data stream. When I went to the Wizard that was beginning of a new start for the AI.

 

No you didn't change any bets after a tie.

I Will be looking forward to your full AI in action.

What has been your biggest session drawdown?

Your prior message when you said you keep to 1:1000 did you mean 1000 units as bankroll?

 

Part of the problem in running this particular software is that the hardware investment is about $240,000. Not you typical home computer setup.
If your desire is to utilize it, I would consider that. The more it is exposed to the better it learns. I can automate the betting process. I just have to add another blade and memory to the computer frame. No big deal. It will then run parallel programs. It will not affect my development side.

Put a pool of interested players together and have someone run the account and distribute the profits. I can direct the software at it to make the decisions. You can then use the profits to further develop what you guys are testing (helps reduce the costs).
If your designs already fully function then this conversation is moot. (Banker only Bets).

 

How much to add a blade and memory?

Are you saying that. We open an account and xyz online casino and you can direct your AI to it?

No.one has a fully functioning design, maybe soxfan,

 

Will your AI place the bets also or only provide instructions on bet side and amount

I would be interested in testing something like that and can reload my account at betonline of bovada.

I'm overseas now and also have access to.many online casinos

 

I try to keep the profits in proportion to the bank deposited and the betting. That gives me a greater lifespan at a particular casino. I have been asked to leave before. I do not try to build a bank to grow larger. I would deposit $15,000 then work off of that. I strip off the profits an maintain the same bank level. That would work on a $25 minimum to a $375 max that ratio is 600:1. If I am going down to $10 then I would maintain a $10,000 bank (better cover for the profit stream). I never finish a session down. It is always a profit just the amount varies. The casinos here give alerts after 60 minutes of play. (gambling problem laws). So I try to keep a session under 60 minutes and spread them through the day.
On the $25 min play I think my largest would have been $4,300 for the day. I try to adjust the software for consistency no highs or lows. Then if I want that to occur I would rather implements that myself to make it more natural. If more profit is desired than up the minimum to $50 or more (depending on the max)

 

I have extra hardware already here. I just cannot use it with one system running. The software use parallel processor and only so may threads run at one time. Adding more processors would not help an individual AI program run faster unless I redesign it to take advantage of that. (I hope that makes sense). So adding another blade is just another computer. A second AI running has no affect on the other system. (more electricity)

It does not matter who controls the account or money or where it is, as long as I can access it for playing then I can tie the AI to it and let it do all the work. I would just eliminate the step of my actual manual input.
I do that for development not for the cranking out profit (the profits are a by-product). The second system is a profit stream. It helps the AI learn faster but it does not need me.

 

Do you mean your largest drawdown was $4300 on a $25 min bet?

 

Unrelated question. Is Bovado the Costa Rico casino. If it is I was looking at their site and their web page said the any withdrawals had to be completely played thru 1x before it could be removed. Where they referring to bonus money or actual hard deposits? Or was I so tired that I just read it wrong?

 

Yes. But from a development point of view, that is not good. I am trying to stabilize the range from low to high tighter. Software orientation not profit. If I want bigger swings than I want to ask for it, I do not want it doing it on its own. It is trying to show off!

 

The more I learn here, it would be easier just to partner with soxfan and let him go wild! I think is overhead is a lot less.

 

Tunica is one of my favorites. I've been there many many times having spent a lot of time in Memphis.

 

$ 4 ,300.-- draw down on .$ 25 min bets puts any one into the amateur league .

 

Baccarat like all EC is 100 % Luck and 0 % strategy .

 

Are you able to provide any detail on the continuous learning approach you are using? As in what software libraries and tools you used to build and train the model?

 

Nonsense, that's only a -172 units down-draw. Even the proficient 21 cards counter get clipped for -1000 units from times to times, hey hey.

 

Because I am from an old generation of computer development I work primarily in C++ and JAVA, I use assembly for the heavy lifting.
Python, Scala, etc. are easier but I do not feel they are as robust (they are great for early introductions), plus I like working in heavy rule-based and structured languages, slower development and more work.
I use an HP enterprise with multiple Xeon chipsets and a modified Unix OS rather than Linux.
The library was developed in house when I owned a computer trading firm for currencies. I have just extended it to this game instead.
Up/Down or Player/Bank not much difference. The structure itself woks in a math we developed called packeting to identify patterns then I use a multi-dimension approach to identify pattern life span. The AI creates around 70,000 different imaginary worlds by twisting the original data stream until it can create the most stable forms, from there it untwists that set back into the real world and provides a decision and its relative accuracy. Right now it his still learning how to identify the best worlds to consider and resolve conflict when it has equal amounts on both sides.

 

Tell that to yer pal biden-spike, with his 88% hey hey.

 

I am sorry, when I gave the $4,300, that was my highest profit level in any one day at a $25 minimum. The largest downside was $1,050 at the same $25 minimum.