??????
You are not logged in.
Well, it doesn't look as cool as a board completely smeared with black by the end, but...
Have you ever been to Babycastles?
Right... you mean because you can't fold when you're not up against a bet at the moment? So you can't fold while making a move or while picking your reveal. You can only fold if they bet more after that. You're forced to either reveal or leave if you haven't folded yet. Hmm... I'll think about that.
As far as the bluff hail mary goes... I'm still not getting why that can't happen before the reveal. Push all in before the reveal, for example. Your opponent has much less hard info about your score at that point.
There are lots of things like this I could do that would help. But I don't want to do too much---I want to let the game board just be the game board.
One kindof interesting thing about this game is that "good play" is hard to pin down and define. There is a huge skill gap between a beginner and an expert (I pretty much never lose these days, and have stopped playing because I feel bad for new players), but I can't put what a skilled player is doing into words the way that I can with League of Legends or even The Castle Doctrine. I can't list techniques to try like I could for TCD.
I think the same thing is true for Poker. You get advice like "fold more." But that's not very concrete. How much folding is enough? How much is too much?
In CM, there are some more concrete advanced things that you can do in terms of which information you're taking into account, but I'm not sure how much that stuff helps compared the the more basic mind-reading skill that is harder to spell out.
It looked like this before:
http://www.gamasutra.com/view/news/2281 … Minuet.php
But that was before any kind of public release. Watercolors have been in place for a long time now. v10 made the black "mask" strokes less dark so that the board was still mostly-visible by the end (in v9, the black would accumulate so much that you couldn't read the ruled-out squares at all by the end).
Thanks, glad you liked it!
Fold and leave in the first round just seems like a strange thing to do. Yeah, so you don't get a good number on turn 1, so you fold. Now you're down to 99 chips. Is it really a huge advantage to leave?
There was some talk on here before about "deep stack play" allowing a strong player to exercise more skill, and how the smaller stack at the table is the limiting factor on the possible skill spectrum, but 99 compared to 100 doesn't seem significant.
Maybe that player was looking for a certain type of opponent, and if you were raising after the first pick, then maybe they were just outta there.
I've seen this happen in poker rooms too.
Yeah, it looks like there was some kind of database overload (too many connections) on the server this morning.
Maybe one transaction got hung, locking the tables and blocking other transactions, though I didn't get any deadlock messages.
In your case, Jere, it looks like a ton of your waiting_game_start requests got backlogged, and in the mean time, the flush cleared out your game as stale, and then after that, all of those wait_game_start requests came through (with some dropped, too, it seems), and the final one (the one your client was still waiting for a response from, that it had not abandoned) was denied.
I did get some connection overload messages, but they happened 10 minutes later, so I don't know what was going on. Sadly, in my experience with PHP/MySQL, sometimes this kind of thing just happens.
It seems to me like the big bluff would simply have to happen BEFORE the reveal move, right? The game without the reveal move (where the final bet happens where each player has 6 possible scores) is identical to the current game up to that point, except in the current game, you know the reveal is coming. So, all the potential for deception at that point is still possible. I don't think having the reveal come later changes that.
It's more like adding a "reveal one card" to the end of Holdem, with an extra betting step. You wouldn't wait until after that to bluff, if you wanted to bluff.
So, yeah, there's going to be a boxing-in after that point, but on occasion, really interesting situations can come out of that reveal move.
I feel like the entire game is a step-by-step process of revealing more information about what your opponent has or could get. Very much like flop, turn, river in Holdem. Without the reveal step, and extra betting round, it felt too compact.
An extra coin penalty for leaving makes a lot of sense. What do other people think about that? One concern is that it would be hard to make it visible to the player. It's like, "pay one coin to leave."
See this other stuck topic, "Getting the Game":
http://cordialminuet.com/originensemble … c.php?id=2
In your invite email, I warned you that this was a back door, not a front door ![]()
Well, the client NEVER sends out the account key. It assumes that someone is listening on the wire at all times (so it's safe to play over coffee shop wifi).
It does send the HMAC_SHA1 of the account key, and the account key itself is 100 bits.
So, you're really talking about a kind of pass-through replay attack here. I'm not worried about someone reversing the HMAC to get the account key (40 billion years if you test a trillion account keys per second).
Hmm... what if the server URL itself was included in the HMAC? That would ensure that messages sent by a deceived client would be rejected by the server. I'll add this to the list.
Hmm... I think that some things can be left outside the game. Tracking pick order is pretty advanced, and you would have time to quickly jot down the column/row symbols on a piece of paper. I like the idea of a piece of paper on your desk accumulating an incomprehensible mass of Hebrew as you play.
Yeah, I think THAT Denied error (where there's no red [!]) has been fixed server-side.
I think the web server did go down one time, briefly, in the past two weeks now that I think about it.
Some of these things that you're describing are "non-choices, where you're still asked to chose."
I'm not happy that these moments exist in the game, but I don't know how else to structure it. Sometimes there IS a choice to be made at these points, but other times, not. Usually, for the reveal, you want to pick the square that creates the largest spread in the score graph---the most uncertainty. This could be done automatically for you by the computer, but it would feel weird if that's how it worked. So I left the choice to the player. There are some rare situations where there's actually a choice to make, but usually you want the largest spread.
Post-reveal, there are only 5 possibilities from your perspective. Both their scores above, both below, both between, interleaved below, interleaved above.
Two of these are force-folds for one player. But still, should the computer auto-fold?
The other three are juicier, with important betting decisions for both players. So, the game just lets you make the decision no matter what.
Note that very advanced players can control the game so that at the end, there is a juicy decision for the other player and a sure win for themselves. The best game state for you is actually not the one with the highest possible score, but instead the one where you're winning with the most uncertainty in your opponent's mind.
This mirrors poker, where a 22 with a 2KA on the flop is a really good game state in terms of winning money from your opponent, but a 22 is not a strong hand compared to KQ suited.
The "knowing you've won sometimes" mirrors poker as well (having the nuts). The "knowing you've lost" has no mirror in poker, so that's not so great.
Even if there is no reveal, knowing you've lost can happen when all 6 of their possible scores are above your true score.
But, in this game, it goes further, where you know you've lost and you know they know too (no chance of bluffing). That can happen even without the reveal.
The original game had no reveal step, but I felt like the final bets were placed with too much uncertainty.
I don't like the idea of breaking the math (because an advanced player would overcome that).
There could be some kind of deception in the reveal (maybe reveal a possible square for yourself that you didn't actually win).
Yes, I do have uptime monitoring. There were some major outages over the past few months, but many weeks ago at this point.
In this case, though, the other player was actively waiting, sending "Is the move ready yet?" messages, which is what finally made the game time out (the server is driven by the opponent who is waiting).
Thus, the server wasn't down. For some reason, MZO's messages didn't come through until it was too late.
MZO, I've never seen a red [!] in all of my testing of the live server (and I'm not on-site with the server or anything). Wonder if other people can chime in here about their experience.
Sorry for the trouble! This should never happen...
Looking up your account in the server logs, it looks like the most recent time that this happened was at 2014-12-09 17:42:48 GMT (adjust for your timezone), about 10 minutes ago. From the log, it looks like maybe your client lost connection, because two of your messages never made it to the server. By the time your move (in this case, your second column choice) came through, on your client's third retry, the game had already timed out. Yes, the client enforces a timer on you (red number that shows up), but if you complete your move on time but your move doesn't make it through to the server, your client will retry after 16 seconds or so. By that time, if you cut it close, the server has likely timed you out and the game has ended. Your move is thus denied.
Do you recall seeing a red [!] box show up? That means the client cannot connect to the server and is retrying.
Sadly, there's no way to prevent this from happening during times of lost connection. I'll tweak the server a bit to loosen up the grace period though. Still, it WILL happen sometimes.
I'm not sure how this should be messaged client-side. DENIED BY SERVER is probably confusing here. Maybe GAME ENDED instead? Or CONNECTION LOST?
You can also just create a game and leave it WAITING in the background while you do other work on your computer.
When an opponent arrives in your game, you'll hear the chime on your speakers and know that the game is starting.
Well, you have to muck around in the settings folder to switch servers (by changing the reflector), so I figure that anyone doing that will clear their account details, too.
There are no official instructions about any of this yet, of course. I simply haven't written the "how to run your own server" document yet. There were some server instructions that shipped with The Castle Doctrine that pretty much apply (but you don't need to do the ticketServer part for this game). You can look in the mercurial repository for The Castle Doctrine on SourceForge, or you can see a mirror of this text in the wiki:
http://thecastledoctrine.gamepedia.com/ … p_a_Server
I'm planning on hosting tournaments on a separate server someday, and for that I will simply ship out alternative client bundles that point to that server (to make it easy for people).
I think I've sent email invites out to close to 1000 people at this point. Who knows how many are reading the email (these addresses come from a "sign up for release updates" mailing list that has been running for many years).
115 people have created accounts, so that's roughly 10%
Well, it was going to be NEXT weekend... but it may not be happening after all.
Yeah, I have family visiting here right now, so I neglected to invite another batch of people yesterday. I just sent invites out to 300 more.
These are important questions to be asking.
One answer is "raise the stakes." If you're making X% an hour, and that results in a cashflow that is not worth your time, bigger stakes at the same skill level would raise that cash flow. It obviously requires more capital and more risk.
In one online poker room, the cheapest buy-in in the whole place is $5. It goes way up from there. The biggest game in CM history, so far, was $5.
I'm also curious about the "on its own merits" part. It seems like Holdem poker isn't so interesting on its own merits, though there is the thrill of the killer hand that shows up every hundred hands or so. That's just the gambling thrill though, I think. The thrill of the heavens aligning and you "getting lucky." This game doesn't have that rat-pushing-lever thing going on, so that "merit" is missing. That might also be why "games like this are legal." You're not hooked.
There is a thrill waiting here when your opponent has the second best score at the table and you know it---they're falling into your trap so perfectly, betting high trying to scare you.
Anyway, I intentionally designed a very simple game where betting strategies would be the focus---a game that wouldn't function on its own merits. I wanted to make a betting game, not a game with betting slapped on top. But that doesn't mean that what I did worked, of course.
No, the video was fine!
Can you play a short session that demonstrates this and then send me the recordedGame file? I can watch the recording on my end and maybe figure something out.
jasonrohrer@fastmail.fm
Also, did Yosemite produce any kind of crash report or log that you can send me? I know that MacOS is sometimes pretty good about that.
If you're willing to delve a bit deeper, email me, and I can walk you through the steps of running the game inside a debugger that would catch the crash once and for all and give solid info for fixing it. It would be REALLY helpful! I don't have access to a machine running Yosemite.