CORDIAL MINUET ENSEMBLE

Well I wouldn't be so quick to give up on a suited game mode before tweaking, players gaining more experience, and seeing what effect having the score graph has. Either way, rapid iteration.

Sure, but that adds an extra tactical consideration: try not to get into that situation after the reveal, and if you've got great numbers ideally you wouldn't give the opponent horrible numbers either, or they'll just fold.

What you seem to be saving is that CM has too much of a good thing, as you don't need to play a lot of rounds hoping to get that position of holding a sure winner and getting to squeeze your opponent. You make your own luck, and plenty of it, rather than facing the bad odds of multi-way poker. But also there's nothing that can happen that's remarkable. Sure you can get a perfect score, but it doesn't matter what the score is just as long as it's the best possible on the board.

Here's another idea: randomly generated magic squares with interesting properties, like that 1-2-3-34-35-36 example I showed. When generating those I noticed the other numbers also formed unusual patterns. There could be whole classes of 6x6 magic squares with special properties that affect game strategy which are never seen when generating without special constraints.

IIRC you added the reveal step to CM after some early testing. So removing it would get closer to the poker experience? Do you think an alt mode with it taken out again would be a worthwhile experiment? I doubt it would be a better game, but is relevant for this debate.

Would it be possible to separate Elo calculations by game type? Seems to be messing with the Elo leaderboard right now.

I also found it hard to see my picks; I actually thought at first that Jason had changed the ink blending in alt mode. The fact that I keep my monitors set quite dim could be part of the problem.

Another option: change the first column picking round. Both players still pick two columns, but both squares selected by the intersections are revealed and shared. Then two normal rounds of picking, for hands 4 numbers total. But I guess there's no way to order the two columns you pick, so it would make more sense to do it over two rounds... pretty inelegant.

Considering that I didn't even complete most of my favourite games before losing interest I'm pretty surprised I'm still playing CM, though not that much before the contest and I've played fewer than 200 games. But it definitely gets a bit same-ish for me too. It's most fun when you're both trying to double-guess each other, but you might be playing on the side rather than concentrating, or your opponent is doing the same or just acting randomly, so the minds of the two players aren't meeting.

My take away is that what keeps people interested in CM is when something changes. A contest or challenge, an experimental game mode, new ante settings. Maybe you could add additional game modes or adjustable game settings to CM (not necessarily radically different games).  Currently the game very much lacks variety. Don't always use the "don't split the playerbase" argument without considering whether doing something will increase the playerbase to compensate. But I guess this should go in the other thread instead.

I don't think it matters that turning a significant profit is hopeless right now. It would matter to the poker players, but it shouldn't matter to the normal gamer. Having a few cents staked on a game is enough to make the game more dramatic and with purpose even if the amount is insignificant. This aspect of the game works perfectly!

Regarding the 'terrain' of the magic squares you mentioned, I think this comes down to thinking ahead one or two moves which I find very hard to do aside from some easy patterns; there are too many possibilities. This is the main point on which a computer could outplay a human.

There isn't. It would be nice to have everything in one place.

There are keyboard controls but I don't use them and don't know what they are.

The idea with clicking on a square is that it lets you then hover over the score graph to see what the individual possible scores are, but I've never had a reason to do that. I'm not aware of any other hidden features, other than that you can hover over the squares at pretty much any time, which shows a few different things.

All play sessions are recorded and saved in the recordedGames folder. To play a session back, copy a file into the playbackGame folder. While playing back a game there are additional keyboard controls, but you must first enable them by editing settings/enableSpeedControlKeys.ini to contain a "1". They are: ^ to slow down the FPS, & for normal FPS, * through ) to speed up. (Note, on Linux vsync seems to be automatic, and has to be disabled externally (export vblank_mode=0) to increase framerate). Interrupting or reaching the end of the playback will cause DENIED BY SERVER errors, nothing to be concerned about.

There are also various other things you can adjust in settings/, like hue shifting for colour-blindness.

Dollar bills with sigils drawn on them.
(Because it would be too funny to mail people cash)

I'm especially amused that there were patches for Space Shuttle for the 2600; just like the real thing! That game was apparently an achievement of realism at the time.

I'm jealous, he beat me to it. I think it would be a nice game for serious academic research (poker is overused).

Upvote count doesn't actually matter at all, I wasn't talking about it; I have read that it's literally meant to trick shills into thinking their votes count. Actual ranking is calculated using non-blocked upvotes.

The story at the start is the most important part.

I meant that I was voting for articles a few pages behind the main page, and their ranking did not improve despite changing; I strongly suspect my votes count for nothing until it decides I'm not a shill.

That is a fantastically written article. Hard to improve. The detour into the game description seems too early before much story happens? More significantly, the ending of the piece is a bit dull. When I hit "So that's my story" my thought was "What! That story didn't have an end!". I would remove those words, to the end. Also it would be good if there was something else to put in the conclusion. More information about or quotes by cullman? Also, I don't think you explained that all players are anonymous, which is very important and explains what the whole point of the twitter bot is (without that, the bot seems irrelevant). Note that the bug bounty is still running.

I resisted for years, but created a HN account for the purpose of voting up Jason's submission the other day. Luckily now I can use the noprocast setting.

Damn cool to read that you were using the twitter bot to wake you up in the middle of the night. Did you really have a profit ratio of 4 at one point?

Edit: wasn't expecting it to get posted so fast; upvoted. Edit: It's nowhere to be seen; I imagine Jason and I are heavily penalised for being new users. It's always possible to resubmit later

I'm not joking, looking at that gif makes my eardrums hurt and my forehead tense up.

Jason, that seems to be wrong. PlayGamePage::mLeavePenalty is set when the client receives the get_game_state response, and it's not set anywhere where else -- not calculated in the client.