I agree completely.

after reading that, and reviewing a recently completed game.. I think dangermouse may be onto something, if you go through the maps at http://phpdiplomacy.net/board.php?gid=1484 you can see that the moment MarekP and I agreed to work together (based mostly on the new points system) we dominated, we wern't bigger when we decided this, but everyone else seemed to betray eachother. It became more of an assembly line process. At the end turkey was offering support repeatedly (i believe to both of us) for the win, but there was no reason to take it since i was 99% sure he wasn't going to stab me and he was evidently pretty sure I wouldn't either.. The game ended with only two players with unit counts above 0. So while turkey stayed in until the end... he still got nothing out of the point system because we essentially made the decision to turn the game into a shutout.. and his ability to diplomatically influence the two major powers with the offer of the killing stroke was neutralized..

I like the points system overall, but I agree it should probably be a component of the game not the goal.

mabye we should reset the stats and start over using this proposed system.

however if a player plays one game and wins he jumps into first place and rait would drop far out of first

reset the stats? heck no.

oh, and berlinerkindl is right. turkey's ability to influence the major powers was completely neutralized. also, i think this is a good example for the "bonus to small survivors" argument.

I strongly agree that winning should be given a bonus.

Ok, dangermouse, that's a very good case I hadn't thought about.

I would now strongly recommend a higher percentage go to the winner. I would NOT recommend that the rankings be based on win percentage. This is because not every win is equal. It is MUCH easier for an experienced player to play amongst new players and win, and MUCH harder to play with other players on equal or better terms. With the point system as a ranking, as you grow higher in rank, you will have to play higher stakes game to see any meaningful growth, which means you will likely be playing people who are close to your level. If you choose to play less skilled players, you will likely gain less points. The point system automagically scales the win quality by skill.

But I do now agree that winning should be made more attractive. But I still think that the point system is a very good indicator of skill, and definitely beats out the old system, or a system based around a flat ratio.

mmm, I can see where you're coming from but I'm not convinced.

You could argue for the opposite effect; that if either of the players wanted to maximize their points they would work with player 3 and try to win the game, unless player 2 thinks he has no chance of winning in which case he would probably just leave without an incentive to continue.

You could also argue that player 2 and 3 are more likely to join up, because if they work together they can gain lots of territories from player 1. If player 1 and 2 join up, though, they can only gain the limited number of territories player 3 has.

Also if winning is all that matters why would player 3 even be sticking around? If he's such an underdog that player 2 has given up on winning and is just grabbing player 3's land there can't be much chance of player 3 winning anyway, so why would player 3 still be playing if winning is all that matters and he doesn't have a good chance of winning.

I do not think this issue should be taken to extremes. Give 10% of the pot to the winner and the rest divided up as it is now. That would be incentive enough for pushing for the win.

10% and this image emailed to you http://herringsalad.org/pix/view/13

If Player 2 is any good he will be offering Player 3 the chance to get more powerful at Player 1's expense so that the endgame will be between a more closely matched Player 2 and Player 1.
With the incentive lying merely in survival, Player 3 will be less inclined to take risks (such as trusting Player 2) and Player 2's leverage has gone. With the greater incentive in winning, Player 3 will have more to gain by stopping Player 1 winning, which is good for everyone (obviously, except Player 1).

Players can come back from very weak positions between two more powerful countries, particularly if one of them is helping. It saddens me when smaller countries spoil a game by giving up and allowing someone to win by attacking their potential benefactors for a short term (and what should be a pointless) gain; even supposedly experienced players have done this at my expense before the points system started!. The points system as stands just encourages this behaviour.

I agree with dangermouse here - with no real penalty for coming second its much more likely that one of the two classic alliances - Turkey/Russia or England/France - will form at the beginning of the game and will jsut roll East or West without worrying about a likely stab. After all ending with 16 centres is almost as good as 18 while the risk of stabbing your partner and going for the win is just too high.

I also agree diplomacy - basically once the first alliance has formed - there is no real need to talk to anyone else as - on margin - they cannot offer anything that beats being part of a juggernaut.

I would also suggest that the points system makes Germany/Italy/Austria an even worse initial draw as the advantage of being over the sytalemate line is now negated as you don't have to win to score big points - just be part of a winning alliance. The Central powers then face the almost certainty of a two front war against two alliances that crush them in the middle. And because of the need to maximise your SC number the alliances will tend to completely eliminate all the other playes so as to maximise their point score.

My hypothesis would be that this stem will cut the win percentage of the Central Powers markedly and will result in more eliminiations in each game.



I think the points ystem has a lot of good points in its favour - but the lack of incentive of going for a win has changed my playing style a lot - and I expect will change that of a lot of other players as well.

I have no doubts that the new rating system makes from PHP Diplomacy rather different game. Probably the game that may seem less engaging and funny to someone, but also the one that requires more (and better) diplomacy and reading other players' moves and intentions more ahead. Opening and middle game became more important than endgame. New kind of diplomatic offers appeared (I'm happy to come second if you join me).

I'm with Kestas in his thinking that it is too early to predict more now. This new Diplomacy may be better as well as worse than the old one. Let's wait a month or two before the final conclusion.

BTW, I'm the one who came second in Berlinerkindel's example. He's true that the new rating system influenced our decision in the end game, but it wasn't the only reason (at least mine). The major reason for me was that Berlinerkindel was extremely good and responsive ally, while the Turkey's diplomatic communication was far from perfect. I'd almost certainly stick with this alliance and came second even without the new system.

(Excuse my English, please, it's not my mother tongue.)

You did quite well with it anyway Marek

Hmmm, not sure what to make of this..

Giving an extra 10% to the winner seems like a stopgap solution that doesn't seem like it'd make much difference. If you get something substantial for coming in second you'll be happy to come in second if your other option is defeat. But if coming in second gets you nothing or very little compared to winning players will be inclined to leave once they know they're going to lose.

If wins are used instead of points for ranking it means the highest ranking players might not have the most points and so might be unable to access the high stakes games, which seems weird and complex. Also there needs to be a way to rate bigger wins in higher stakes games more than little wins. And if someone is only playing to win the problem of people leaving a game when they know they will lose returns.

It seems like a damned if you do, damned if you don't kind of situation :-(
Give a bonus to runners up and get fewer end-of-game comebacks, or don't give a bonus to runners up and get more end-of-game civil disorders; or pick a mid point/compromise (eg a survival bonus paid equally to all non-winning survivors) and get a bit of both (as well as making the system more complex which I desperately want to avoid)

I could go with that last bit. Winner gets a certain cut and all remaining players get an equal "survival bonus".

in the original game of diplomacy, they just counted how many times you won, for example in zines. this point system was only created to stop players from quitting and be more accurate, right? (see Chrispminis' argument against win ratio scoring) so winning is much more important than coming in second.we shouldn't make the difference too big, though, as that would encourage giving up when you're in second (10ish) to someone else in first (16ish).

any thoughts?

The free market idea would be to let people decide how they distribute their bet between the "common pot" and the "winner bonus". Then the market would decide the proportion, no need to argue for weeks ;) But it is rather complicated.

Now actually thats not a bad idea at all, and actually not too hard to do, although Kestas (or other dev) would have to edit some of the display functions as well as the awarding features.
What would happen to money contributed when someone joins the game?

wait wait wait wait... you guys are talking about a WINNING bonus.. damn, i thought dangermouse was requesting a WHINING bonus.... i have to rethink my position now!

I think the pot system is fiendishly clever but already pretty complicated and difficult to understand. If I had to come up with a scoring system I'd be tempted to keep it really simple, like 100 points for a win, 10 points for finishing a game with a supply centre (be it 1 or 17), and -20 for going into civil disorder. The best way to stop people giving up is to penalise them heavily.
However, there are many advantages to the pot system, now we have it I probably would not be so hasty to throw it away.

Incidentally Marek gave a good example of the only occasion when I think it is acceptable to help another player win - when one honestly deserves it more for the skill and the spirit with which he has played the game. But this only applies if there really is no chance of a comeback. It is always worthwhile to treat everyone, enemies, allies, and neutrals with courtesy and be careful who you backstab, because you never know who will end up holding the balance of power.
Notwithstanding the above, any alliance should ALWAYS end the moment the stronger party is on the verge of winning the game. Anyone on the verge of winning the game must be, by definition, enemy number 1.

of course. again, winning is much more important than coming in second.

Think the system now works pretty well - discouraging CD very important - but would like to see a mark-up for the winner - maybe 50% or so.

I think that bihary has got a good enough idea there. It's been mentioned before. Simply allow the creators of a game to allot a certain percentage of the pot to the winner, on top of the splitting of the rest of the pot. Then, people could choose what type of game they would like to play, and I think that over time, just as the points are expected to level off to truly reflect skill, I think that the win bonus will eventually level off to a percentage acceptable by most experienced players, and will become the standard of play.

It's interesting to see that many players here are willing to give up a win if they feel another player has truly earned it. It now remains to see which players are better at convincing other players that they have truly earned it. =)

I'm also for the good old winner takes almost eveything, all survivors gets the same system. And in case of (unimplemented) draws, everybody should get the same

"Notwithstanding the above, any alliance should ALWAYS end the moment the stronger party is on the verge of winning the game. Anyone on the verge of winning the game must be, by definition, enemy number 1."

If an alliance is played right there will be two even powers, it would essentially become a sudden death match, which ever scored first would win.

an alliance played right would mean 2 players advancing equally and the better player having a 3rd wheel to ruin all of the second players plan. Either by making the inferior partner anger the 3rd wheel to irrational behaviour or just plain tactical mistakes.

another argument:

if a player is on the verge of winning, but can be stopped if the other players predict which territories he will attack, before the points system all players would do everything they could to stop him (because... why not?)

Now, if they are worrying about finishing with fewer territories and they have an opportunity of grabbing some land from other players INSTEAD of possibly stopping player 1 winning, that seems a more attractive and prudent move. As someone posted to me (pretty much), "its unlikely that you can stop me, you might as well grab as much land from the others as you can" (I did but only because in this particular case it was a mathematical certainty the guy would win!).

So people won't fight to the death any more, they have too much to lose to bother actually trying to WIN the game.

Hoooo boy.. A lot of hate for the new points system.

I'd have thought that most people would want to win despite the points because the game isn't about getting points. :-/
The points provide features like letting players play with players of roughly their own skill, an improved (but of course not absolute) way of judging a player's skill, a deterrent to the problem of civil disorders, and a way of preventing players from joining too many games. But points aren't supposed to be the reason for playing phpDip, they just try and improve it by cutting out some of the unpleasantness of playing with bad players and having players leave mid-game.

That having been said, here are a list of the various scenarios that a player might find himself in, and a look at how the points system might affect their decision from the perspective that points are a matter of life & death:

- You have a chance of winning
---- You want to win
-------- New system: You'll try to win to get as many points as possible
-------- Old system: You'll try to win (don't ask me why anyone would want to do this without the precious points! ;-) )
---- You want to draw because of a strong alliance
-------- Draws aren't implemented yet. At the moment large alliances will pick on smaller players until one of the people in the alliance wins. This isn't ideal as it means smaller players lose points when they don't deserve to, but this is because draws aren't implemented, not because of a problem with the points system.

- You don't have a chance of winning
---- You don't have a chance of winning, but you have an alliance with a larger power who does stand a chance of winning.
-------- New system: You help out the larger power. You're not going to win but by helping him out you'll be able to get some land back from the larger player.
-------- Old system: You leave, the larger power that would have stood a chance of winning no longer does, because he doesn't have your support
---- You don't have a chance of winning, but there are a significant number of small powers still around.
-------- New system: You band up with the other small powers. You're not the biggest or the most well-positioned of the small powers so you're not going to win, but by banding up you could take a large slice of the lead player and get more points in the process.
-------- Old system: Unless there's a tangible chance of leading the pack of small countries, and taking victory after taking the lead from the leading player, you leave along with the other players. And you have to hope that no-one decides he's not going to win and leaves while you're leading him, and the other smaller players, against the lead player.
---- You don't have a chance of winning, and there's no way you can expect to get territories back from the leading player
-------- New system: You'll stick around and try to protect your territories against the lead player, or try and grab territories from the other smaller players.
-------- Old system: You leave the game and go into civil disorder. The leading player only needs to reach a threshold where the smaller players think they can't stop him, and they all give up.

Just to clarify the example berlinerkindl used, I was Turkey. I had promised something to France, but went back on that deal and as such there was never any chance of me reconciling with him. That was a lost cause. I then turned to England, but he seemed like a great guy who wouldn't want to stab someone in the back - I totally accepted that. So, whether or not they chose to simply split the winnings, I would have lost anyway because I dug my own grave.

I think I am starting to favor the eBay stars idea... (j/k)

I've turned on people in some games.. I'm not a saint.. never done it without (at least in my mind) cause.
I agree with the holes being dug point though... that can mess you up right quick in a game...

Kestasjk, don't sweat the points thing... it's change, people are resistant to change.. in a month it'll be the colours used on the maps... human nature is to be drawn to conflict, and if none is around... create one!

my last comment on this thread (i hope) and this is for all those bastards that crap on your work and bitch about how bad it sucks or are militantly against the points system.....

FORK IT IF YOU CAN DO BETTER!!!!!! It's not like you don't have the source available to you to do it... give the guy a break Open Source development is very demanding at times and when people do nothing but pick apart your work and not offer contributions (fixes/enhancements not cash, though I'm sure it'd be accepted) it can be overwhelming and drive the developer to drop the project because his love for it has been killed.

So to review.. don't tell him how lame you think phpDip is, just don't use/play it.And don't bash him over the direction he's taking it... if you have another idea.. just fork it and you can have it your way.. if you are willing to do the work.

Kestas, I think the majority opinion is supportive of the points systems. For me, I just don't think it's the obvious way to solve the cd issue.

Berliner: ummm, no. And maybe I'm out of line, but I thought the Forum was a place you could give constructive feedback on the site in order to help make it even better. I don't think a single person who has raised issues with the points system has been unconstructive. Not me, but several of them have been here a long time and helped with code.

Kestas, I'm sorry if the criticisms of the the points system have made you feel like your work is unappreciated. Far from it.

Why not stick to the point system, but change it so it doesn't matter how many units you have. Instead, winner whould get almost everything from the pot, and all survivors should get the same, as many of us suggested here. I think this is how diplomacy meant to be played. My english is not as good to compare this system thoroughly to the old / new as Kestas did. But you can argue that for this proposed system to work you need draws implemented. For example, in the case of 2 vs 1 stalemate position, in the proposed system two allied players would never stab each other, as it is not in their interest. So the draw is needed for game to finnish. In the current system, two allied players would stab each other just to get more SCs and will let the third player win.

sorry to interfer; i stopped playing before holiday season, but I am still watching the progress of this interesting project

my opinion is:
if there is a hall of fame, everyone adapts to the point system; just hide everyones points - so each player can only see his points and remove the hall of fame or change it to use an ELO based system (win percentage is inferior to ELO but might be sufficient)

I like the idea of the point system that filters out how can join a game; but I don't like it being used for hall of fame, as this kills the spirit of the game

... filters out who can join ...

I think the point system solves more problems than it creates, and for that reason it must stay. But it is an innovation, it isn't written in stone (or the official Diplomacy rules!), it does affect the way people play the game (deviating from face to face/postal), and there is nothing wrong with suggesting a few tweaks here and there. However, it's kestasjk's work, and it's his prerogative to ignore any suggestions if he thinks things are fine the way they are... but if anyone knows how to use the power of persuasion on him, it's a bunch of hardened Diplomacy players..!

lol, yeah.

I love the points system, kestas, don't get me wrong. But you were only comparing the points system to no system. Let's try comparing the current points system to one where the winner gets a bonus of some sort.

if you have a good chance to win, you still try to win, as before.

if you have a small chance to win, things change. For example, say that France has 12 units, I'm Russia with 8 units, and everyone else is small change. Under the current system, I attack my weak neighbors, and end up in second place to France. If the winner got a substantial bonus, instead I would work with my neighbors to try to stop France. Notice that this is also what would happen if I were actually playing Diplomacy in real life, just trying to win.

I don't want to diminish how awesome the points system is; I love it. But I think this change would make the incentives a lot more true to the spirit of the game.

Mahatma.. my previous comments weren't meant to offend.. everyone in this thread has been constructive in the criticisms, my comments were flavored by comments left in the recent past (other threads) bashing the work he's done... and even if you like the work, but feel you can contribute (code) you should, and if it's not adopted and you feel it should be.. you can fork.. one trunk, many branches, big beautiful forest of variants.

You're probably right anyway Berliner. I overreacted because this system really doesn't sit right with me and it's touchy given Kestas and others do this for free.

For all of us who think our goal when we play Diplomacy is, or should be, well-defined, let me reference a very nice article:
http://www.diplom.org/Zine/S1999M/Windsor/point.html
I agree with the article's point - Let all the opponents have their own motivation, if I am to succeed, my job is getting to know what they are.

You can read this article by the inventor of diplomacy when devising new point system http://www.diplomacy-archive.com/resources/calhamer/objectives.htm

He says "the draw, of course, is the only objective other than victory that is recognised by the rulebook", and argues why number of units should not matter, whether it's solo victory or draw

he's got some very good reasons, too.

He also gives some good reasons for why the number of units should matter. Thanks for the interesting and relevant articles but I think if they show anything it's that this is a hot topic for debate and that if you ask 5 people exactly how the points system should work you'll get 5 answers.

From your answer I guess we'll not get new point system on this site :-)

And yes, there are many systems, but I don't think there's one, besides this, that awards second placing as much as a victory

Coming second is rewarded by a 16:18 ratio to winning only if one player has 16 units and the other player has 18, in the majority of cases there'll be 2-3 smaller players so the ratio between first and second will be more like 6:18 or 8:18.

The only compelling argument against the system are the ones about draws, where alliances pick on smaller players to maximize the number of units they get. This is a good point that has been raised in this thread, and that'll be solved by separate code.

But yes, I'll wait and see if the gameplay deteriorates before changing anything ;-) Based on what I've seen so far most people play out of enjoyment and not to maximize the number of points they get, so I haven't seen any noticeable change in gameplay (other than the benefit of being able to play exclusively with more skilled players)

I can say from personal experience that this purported benefit of being able to play exclusively with the more skilled players is overrated.