??????
You are not logged in.
Pages: 1
A lot of people have been asking about this.
After thinking about it, I realized that it is unfair that you can skip the next ante by leaving at the right moment. I don't think that online Poker tables allow you to do this, because the next ante flies in right away.
And, if we're going to say that the next ante is due at the end of this round, then what if you know you've lost this round and leave before that? That seems like yet another way to avoid it.
SO, I just changed the server so that whoever leaves both loses the pot AND pays the next ante to whoever stays, no matter when you leave during this round.
In v12, this shows up a little strange sometimes, with some coins flying into the center from the remaining player before flying back, but that's fixed in v13. Also, the display of the house tribute is messed up in v12 when one of these penalties is displayed (client behaves like penalty was one less and there was no rake as it displays coins flying for the remaining player).
Offline
So if the ante is 12 and you leave/disconnect, you lose 25 coins? Seems like there is a very strong incentive with this to play out games after you've gotten past the first couple rounds, which as far as I'm concerned is good. But still, ouch!
Canto Delirium: a Twitter bot for CM. Also check out my strategy guide!
Offline
Yes, if you leave in the middle of the current round, you'll lose what's in the pot (12) plus the ante for the next round (13).
I thought about having it only trigger at the very end of the round, but that would just train people to leave right before the round ended (like, after the reveal move, when they could tell they had won or lost, they would leave before the final reveal to avoid the penalty).
I'm not entirely sure that a rising penalty makes sense here. What we're really trying to prevent is someone who wins a coin or two and then leaves. Especially at higher buy-ins, this could become a viable strategy (join table, scare opponent with all-in on first move, then leave with 1 chip, which could be a $10 chip in a $1000 game---almost no one would match the first all-in, so this would earn the perpetrator $10 reliably).
I think I'm going to set a fixed leave penalty instead.
It is now set to 6 coins flat, no matter what round you leave.
Offline
Perfect!
Canto Delirium: a Twitter bot for CM. Also check out my strategy guide!
Offline
As much as I know it annoys people, I like the idea of playing around with strategies of leaving early and minimising how much information your opponent has about you. But I agree that things like bullying people to fold early and then leaving is not satisfying for anyone. The flat leaving penalty seems like a fair compromise.
Offline
Nice!
Offline
I enjoyed much more the games tonight
Offline
I like the change but perhaps if you have the option to fold and leave at the same time you shouldn't have to pay a penalty. This prevents them from seeing a reveal for free as well. In poker you could quit at that time and not pay an ante (then again, you can leave after any hand and not pay in poker).
Offline
Poker fixes this problem by having players ante upon entry. A joining player mid-round in Holdem must pay the big blind, giving every other player a "free" round.
Paying on exit is fine, but currently the penalty for leaving early is not communicated to the leaving player very well, if at all.
Offline
Paying on exit is fine, but currently the penalty for leaving early is not communicated to the leaving player very well, if at all.
agree with this. perhaps when a player clicks "leave" there should be a box that comes up asking the player if he is sure he wants to leave and stating the penalty amount they will be paying for leaving.
Last edited by railroad animal (2015-01-30 06:47:38)
Offline
Good adjustment................I like it.
Offline
It is nice to take a little extra off Jeopardy Alcohol every time they leaves early. It still doesn't stop them from doing it though.
Offline
It is nice to take a little extra off Jeopardy Alcohol every time they leaves early. It still doesn't stop them from doing it though.
That's why I proposed and still like the leaving penalty to be 2x next ante, as the players would be really incentivized to look at just one more card. But, if the baby is crying or the boss is calling it's still not the end of the world if you have to get up and leave. That seems like a reasonable balance to me.
Offline
But, if the baby is crying or the boss is calling it's still not the end of the world if you have to get up and leave. That seems like a reasonable balance to me.
I often play under "brb, cat's on fire" situations. Regardless, I think the penalty's good for the game.
Try Linux, get free. #!++ (CrunchbangPlusPlus) is a stable distribution based on Debian 8. Keep it fast, keep it pretty.
Offline
Well, leaving with half your chips remaining, once you're down, is an important skill. Yes, it's really annoying when someone leaves before the end, or leaves when YOU are down before you can recover, but it's just as annoying in Poker.
What I REALLY don't want is people leaving after bullying a chip or two out of someone. That is not interesting at all.
In Poker, a player who is doing that will eventually get burned, because someone is going to have AA eventually and call them. Here, there is so much skill involved in your remaining picks that even getting a 36 on turn 1 is not enough to call an all-in bully, especially since you haven't learned to read them yet. In poker, if you have AA and call the bully, no further skill is required to win (though you can still lose, but your chances of winning are known at that point). Here, you could call with a 36 and still screw it up. Plus, calling gives your opponent lots of info about what you have, which they can use against you while picking in subsequent rounds.
So, I'm thinking that bullying a chip or two out of someone is a more viable strategy here than it is in Poker.
I was even thinking about that for Nate's $6000 game. Someone could just come in for one round, bully one chip out of him, and walk away with $60. I doubt he would call a big bet on turn 1.
Now, if someone does that, Nate would win $300 from that person.
Finally, with a fixed penalty and rising ante, the cost of leaving shrinks relative to the ante over time. This makes sense, because it's more feasible to leave after 6 or so rounds, which at least gives your opponent a decent game with you.
I hear you about it needing to be communicated better to the leaver. v13 will fix that.
Offline
Oops... there's also a bug where if you win all the chips, you still get a penalty if you leave first.
Just fixed this.
Offline
What I REALLY don't want is people leaving after bullying a chip or two out of someone. That is not interesting at all.
...
So, I'm thinking that bullying a chip or two out of someone is a more viable strategy here than it is in Poker.
Thanks! I think this is a good solution. These leavers were annoying.
Offline
I think someone might have written about this in a previous thread, but I can't find it...
So I was almost out, and I went all-in with my remaining 20 coins. I picked well and won, so I was up to 39 coins (after tribute). Right after this round, my opponent left. But, he didn't leave while the coins where being allocated to me. Rather, he seemed to leave during the brief space when the next board is being delivered to the client.
So this was very confusing: coins were flying everywhere. His ante coins were counted into the centre, then returned to him, then six coins went to me, and then something like four coins flew away, as if they were tribute.
I don't know what happened, but I ended with 45, so I got the leaving penalty, but not the next ante. Makes sense if he indeed left before the next board appeared.
I don't know if this is a small bug that can be fixed, because it might be very disorienting to new players. I know I found it strange!
Offline
Dan:
Can you email me with more info about this?
Like, do you remember the approximate date/time with timezone? Do you remember the stakes? Also email me with your account email address. Then I can look the game up in the logs and figure out what happened.
Oh, seeing that recorded game would be great too.
Yeah, client sometimes shows weird coin movements if some situation happens on the server that leaves the client hanging without a complete picture of what happened.
Offline
This is totally off topic (hey, Dan_Dan started it!)...
In hindsight, would you have designed the protocols differently? It seems like a pain to work backwards from the game state to the coin movements and actions that took place, and there have been quite a number of bugs resulting in crazy animations. And even the way that the reveal step is transmitted to the client is really weird.
Offline
Yeah, it was agile! :-)
Reveal step wasn't planned, so I had to add it later in a way that wouldn't break the rather-complex client logic.
Coin movement happens instantly on the server, so the client is left to infer what must have happened based on where the coins ended up.
This is a pretty clean way to handle it, except in a few tricky cases.
I guess the server could track and present a list of recent coin deltas and directions. But then we're in a situation where we're not sure what deltas the client has already seen and is done with. I guess there could be an ever-growing, numbered list of deltas that the server keeps appending to and sending the whole thing to the client each time.
Offline
Also, if you notice, I do ship games.
One of my secrets is a "get er done" attitude about protocols and code structure. I almost never do major refactors.
Someone recently was shocked that the protocol for this game wasn't json. Are you kidding me? Like, even though NOTHING here is written in javascript, and we're not actually serializing javascript objects, I should implement a javascript object serializer for my protocol?
"But what if the protocol changes? What if you need to add new elements to a message? Everything will break!"
Umm... if the protocol changed, would I really want old clients blindly ignoring some new element that they don't understand? When the protocol changes, everyone gets a new client.
This isn't an open internet standard that I'm developing here...
Offline
Oh, I certainly wasn't suggesting you change anything. Cost vs benefit looks bad. I was just curious about your opinion based on your experience with designing protocols. So, thanks.
Offline
I'd always go with text over binary, because debugging is so much easier, and binary doesn't save that much space for small messages like this.
That said, some structured format (like XML, json, or whatever) is overkill. The parser is often more complicated than the game itself.
Ordered, whitespace-delimited strings work great. Easy to parse with a simple split, easy to format with a simple join.
Offline
Pages: 1