CORDIAL MINUET ENSEMBLE

Marvelous. I hope to see those photos!

The 6th of June? That's today! You didn't specify a timezone though...
(Funny, I just found this thread)

With the inactivity, we need to manually organise games now. And these forums just don't seem to work for that. Occasional (regular?) reunions would be a fantastic idea.
I notice the Chatzy forum has been deleted due to inactivity. A Discord channel may work much better, since plenty of people use it, and if they have a page open to any other discord channel they would see messages from the CM one too. And it keeps a message history, like Chatzy and unlike IRC.

That really seems like dodging the maximum risk which was the whole point of the offer!

I uploaded a copy here: http://pastebin.com/X10ZfJbQ
It expires in one week.

Haha, I guess losing all your money to cullman proves something and is an admirable action!

OK, I'll see it later I guess. Whenever it's available. Good luck!

Are any of those live-streamed games viewable online?

I guess a community of early adopters isn't a long term one, as they probably move onto other things quickly as well.

I still enjoy the game but I assumed I would always be able to play it later, so sort of saw no reason to actually make an effort to join tournaments and so on rather than chase other interests. I would be quite sad to not be able to play anymore, but I don't think it's impossible yet.

I would join another tournament if the hours are alright and it's not in the next few days. I think it could still easily attract a lot of people.

I feel guilty for saying that I was working on an android port, in the face of a couple other people interested in doing the same, and then dropping the project half-finished. I've got to put something up.

The CM bot known as String Corn (this constitutes a public announcement, I guess) has not been playing many games lately because it runs on my netbook which has been quite unreliable; going down because of wifi and router problems, accidentally suspending, etc. If CM is at these lows I might as well release the entire source code.

Well, that's a wrap up.

Not to be mean-spirited, but I'm pleased to see Expression Bosom didn't win, because he played for only 40 min and then walked away. That's still a huge problem with the tournaments. If there were many active players you would have to stick around, but as it is it anyone who's ahead is incentivised to stop playing. But I couldn't blame Jason for not wanting to spend time implementing a different format now.

Yes, I'm surprised I hadn't heard of tineye! I figured google and yandex 'search by image' were good enough.

The "P" on the board you're referring to is actually the Hebrew letter Qof, ק, which looks little like the letter on the amulet. And this is their meaning:

http://www.triple7center.com/Codes_SYMBOLS.html

26 letters are excessive. Many ciphers both recent and more historical use fewer letters for English. For example the Playfair cipher used in WW1 had 25 letters to fit a 5x5 grid. Many historical ciphers both added and removed letters. You can see on that second amulet that J has been omitted, which is normal.
As another example, the cipher used in Book III of Trithemius' Steganographia (supposedly the first book on cryptography; Book III is disguised as a book on magic and demonology), which I was reading about recently, took a 23 letter Latin alphabet, dropped one letter, then added 3 for German sequences like sch, ending at 25 codes.

Hello.

The lack of videos is extremely dire, I would place a lot of blame on that. Last I looked, there wasn't a single video giving you a hint of what double guessing and betting is like. Needs narration, not just explanation. Is it even possible to explain all the guesses and doubts that pass through your mind? Maybe not.

...

Interesting article and scheme. His expected ROI would be 10% if he stops when he hits either -$500k or $500k. I believe that when he keeps playing while he's ahead, his RoI gets asymptotically closer to 20% as the amount that the casino is willing to lose increases. Suppose that the casino stops the game once he's won $1500k and he's got $500k to invest. Consider:
Game A: He starts with $0k profit, and has even odds of hitting either -$500k, at which point he stops, or $500k, at which point he enters Game B....
Game B:  He's initial $500k up. With even odds he would either reach $1500k or $-500k, so the expected profit starting from here is (1500-400)/2 = 550k.
So the expected payoff for Game A (the original game) is (550 - 400)/2 = $75k. On an investment of $500k that's a return of 15%!
Iterating again, if the casino stops at $3.5M loss then his expected RoI appears to be 17.5%, and at a $7.5M cap, it's 18.75%.

However if your odds were only something like 48-52 I would think that would take many multiples of 2% off the RoI.

Also, that's not really a Markov chain (unless you consider one with one state per possible balance, though it doesn't seem interesting), it's a random walk.

The only progress I've made is that I finally figured out how to modify the source to get CM to run at less than 666x666 (the android port was a special case, because SDL-android gives you a screenbuffer larger than the actual screen).

I'm sure the server will run fine on any low end machine. Without databases full of thousands of users and games the only computationally expensive thing will be generating magic squares. On my netbook, with a very slow Atom CPU, that can take over 3 seconds, and I think the 1st gen RPi ARM CPU is even slower. Easy to fix: getMagicSquare6 isn't being compiled with optimisations turned on! Turning them on makes it a non-issue.