not necessarily, the survival with 8 SCs was told to me by another player. Apparently, it encourages players to continue playing despite knowing that they have really no chance of winning.

I think winning percentage needs to be the most important criteria.

re why 8 sc's is not a good statistic--- i had 12 in a game but then those of us who had banded together to stop the "apparent winner-to-be" decided the winner by a lottery (after the "apparent..." was pulverized) . It was easier for the lottery winner to take my sc's and i ended up with far less than 2. The "8 sc" rule certainly doesn't reflect my abilities (actually luck but we all have egos) in that game

correction - I ended up with 2 sc's at the end

so, unless we should also factor in tide tables, phase of the moon, what one had for lunch, etc, the simplest assessment of ability is:
1. winning average
2. total number of wins
3. taking over a civil disorder doesn't go into the calcualtion of the "winning average" - either as a game played or as a win (I originally thought civil disorder wins should count but some takeovers are right at the start of the game)

I do support the idea (I think it was put up by kestas) that at the end of the game, every remaining SC-s would give little to the final score - it would encourage also players to participate even if You would have for example only 2 SC left at the end (ordering them to support hold each other & not letting them to be taken that easily).

Otherwise I support the idea that statistics should basically show Your total wins & winning % out of total games, taking also into account in some form if You took over CD country or You were playing the country from the beginning (how exactly would it be calculated is up to discussion - if those games would be taken out from Your winning % score, CD winnings would be given extra points or something else)

Actually on the second thought I would also have opinion to see the winning % also for the last 10-15 games or so, to see the recent performance (people are advancing!), which I think both figlesquidge & fastspawn have brought up...

I do support the idea of giving points for being second, third...
This is an incentive for everyone to play in its best interest to the very end of the game.
Presently, there is no distinction for losing with 9 sc in 1914 and being eliminated in 1902 ; alliances are doomed, for there cannot be a "common victory". Thus, every game looks the same : 3 to 4 years to emerge as one of the big bosses, then stab your ally to be the only big boss, 3 years to finish with an already known result. Too repetitive imho.
We can try a lot of scoring systems. Example :
1 point for every year of survival.
20 points for winning with 18 centers or more
10 for being second
5 for being 3rd etc...
It would bring a lot of different strategies.
With the present system, it's all or nothing.
By the way, when I look at the record of some players, I take in account early eliminations, number of games, number of sc at the end of every game.
What we see too often today : some players, as they realize they won't win, simply go into civil disorder. Sad.

It'll be difficult to come up with something that isn't overly complicated, but still does the job.

It has to:
- Reward people who get as many supply centers as possible
- Encourage players to win and play with different people instead of constantly using the same alliances and playing in the same groups
- Penalize civil disorder, reward people who take over civil disorders


I was thinking what might be interesting is a kind of currency thing, where a certain amount of each player's points go into a game as a stake, and at the end the pile (of points) is paid out to each player depending on how well they did in the game.

This would prevent players from playing in the same groups; they wouldn't be able to keep winning amongst themselves and getting more points, because they would only have so many points to play with amongst themselves.
Also playing in more games wouldn't accomplish anything unless you did well in the games, because you have something to lose. This would make joining games and going into civil disorder when they fail less attractive.
It would also make it easy to implement the part which would prevent new users joining experienced games; they wouldn't have enough points to join a game where the stakes are higher.
It could also be adjusted so that if you frequently win with another player you and the player you win with won't get as much of the pile as either of you would have got if you had done well with someone who you hadn't done well with before.

It's also not *too* complicated.. But I'm just thinking out loud

Rank between first and second place should be a significant difference, but not between third and fourth place. That would encourage someone with four SCs to attack someone else with 4 SCs, instead of joining together against the winning player.

Reward 1st, 2nd, and 3rd on such a grading scale only if they have over 8 SCs -- but give a lesser amount of reward to anyone still in the game regardless of how many SCs they have. That way, so long as countries have less than 8 SCs, they will have absolutely no benefit to attacking other countries with less than 8 SCs.

when i mentioned 8 SCs in the first post, i was thinking along the lines of a 3 way, which is how usually most games that last past 1910 end up. In a 3 way, what usually happens is that there is a seesaw as once someone sees a person gain a slight advantage the other 2 gang up against him and force him down, before the cycle repeats. This is why a lot of games once they reach 1910, go on and on. the 8 SCs is actually a brib, a cop-out so to say. Once a player is tired of this detente, he can choose instead to support one player to get the prize at the cost of several lost points. In fact both players that are losing can get 8 SCs if they worked together.

I like the idea of penalising players for civil disorder forfeits. It should be worth as many negative points as a win gains positive points, so you can see serial forfeiters at a glance. The only exclusion would be when you use the Figlesquigle button to quit all games simultaneously to go on holiday.

I actually rather like kestas' idea. And on the idea of a score proportional to the amount of Supply Centres you finish a game with.

@EricHerboso, you WANT everyone to gang up on the big guy. As some people have complained, once you're the biggest, you seem unstoppable, especially since many players don't spend as much time on a game where it appears to be a lost cause.

Re: 8 SCs
Sorry, don't like this, 8 SC's seems to be an arbitrary number and doesn't correspond at all with Official Diplomacy. It would be better to be proportional to the amount of SCs you have at the end of the game.

@Epaminondas, I agree that certain games have a repetitive feel to them, but it is up to you to make them interesting. The big guy can always be stopped if the little guys work together. Because unless he has already one, you still have the numerical advantage. I don't like the idea of points for years stayed alive, because this will lead to farming of games, and people forcing stalemates to rack up the points, to the point that it overshadows the win bonus.

The one thing I would add to kestas' list is to reward players who play a higher "quality" of game. Of course, this would be a direct effect of a currency type scoring system, since players who play a lot of games would be limited by their currency and the price range of the games they join. And your profit would be determined by how many games you won vs. how many you lost, with varying price ranges.

The problem with a currency system is that, if new accounts start off with some currency, this might encourage the farming of multiple accounts, although I'm sure we can think of something to stop this.

I hesitate to put forward any numbers, because it is really quite preliminary at this stage, but here's what I'm thinking.

The creator of a game sets the buy-in price anywhere from 0 credits (Unrated) to the amount of credits they own, for this example we'll say 34 credits (convenient because of factors). That means that anyone who wants to join, has to pay 34 credits, and their credits go into the game's pot. So the game's pot at the start would be 34 x 7 = 238

At the end of the game, each player recieves credits based on the number of SC's they control.

Credits won = SC's controlled x (Game Pot/34)

So each SC controlled is worth about 3% of the game's pot, and in this case 7 credits. To make a profit, the player must survive with at least 5 SCs. The winner of the game, will recieve at least 126 credits in this game, approximately 52.9% of the pot, if he has 18 SC's.

Regarding Civil Disorders, I think this could be resolved with this. A player who falls into civil disorder, loses the credits they used to buy into the game. This will make even owning 1 SC preferable to falling to civil disorder, but does not put penalize THAT much for players who absolutely cannot access a computer. For players who take over a country in civil disorder, they get their buy-in free, since it was payed by the player that quit, so thus a free shot at credits.

Of course, everyone would start with a certain amount of credits, and for "pros" who wan't a challenge, simply setting the buy-in to slightly above the number of credits new players start with is an easy way to avoid newbies.

Currently, this system is zero sum, which is harsh, and encourages multiple account farming, so I propose some way to increase the total amount of credits in the system, other than through the creation of new accounts. As well, the system I am proposing does not yet have a system for discouraging repeated play with the same players.

For a non-zero sum game, adding a donation for credits option could net you some money to support the site as well as provide a flow of credits into the system. Or, game pots could have a percentage of credits simply added in. If this is the case, it would be easy to penalize players who play with the same people in rated games too often, by simply taking away proportionally from their prize.

If a currency system is implemented, I would like to see two different ratings. One for the wealth of the player, and the other for their win : total games played ratio, or other relevant statistics.

Tell me what you think, it's open to tweaking, or even scrapping completely, but please back up criticism with logic.

Re: 8 SCs

it is not arbitary. 8 is 34-18/2, or the minimum share one person gets if he works with the eventual winner of the pot.

C - there is certainly good logic to your system but for those who just like to play for fun, it seems excessive. If enacted, an opt-out choice would be nice - that is, players could choose not to be part of the rating system.

As for civil disorder, I believe that banning from the site is the best remedy - the question is how many cd's can be tolerated and there should be some mechanism for those with legitimate excuses.

Finally, it is apparent that the majority of forum contributors disagree with my "KISS" approach - too bad ---- Occam's razor was getting dull anyway

i've got it! new players start with one unit of the currency, and then they join a beginner game with more rewards. this wouldn't stop new players, but it would take so long to farm points that it wouldn't work.

I basically agree with Kesta and Chris, but I'm suggesting some tweaks that would (1) prevent newbies from having to deal with a complicated rating system, per Writhdar, (2) keep the incentive for third place to stab fourth place in the endgame down to a managable level, per EricHerboso, (3) reward players who seek out new opponents, per Kresta, and (4) keep the incentive to play in thirty games at once down to a manageable level, per half the forum,

(1) Newbies shouldn't have to fiddle with buy-ins and stakes just to play their first game. So, let's let newbies into newbie games for free. A new account starts with no credits, making the creation of multiple accounts less tempting. A default new game would cost nothing to enter. A default game would pay, say, five bucks worth of Diplomacy Dollars to everyone who finishes the game (whether by being eliminated, drawing, or winning), plus EITHER one Diplomacy Dollar for each supply center held by those who cooperate to force a draw, OR a big prize of twenty to thirty Diplomacy Dollars for players who outright win. Since outright winning is theoretically the object of the game, it deserves more than just a reward for 18 supply centers.

Meanwhile, those who have won a few games against their fellow newbies will have the credits to prove that they're advanced players. Most advanced players will presumably be able and willing to fiddle with the game-start options, so they can set up non-default new games that require paying a stake of Diplomacy Dollars up front.

(2) The problem with dividing the entire pot according to Chris's formula [proportional by supply centers held at end of game] is that, as Eric points out, it will pretty much stop smaller players from banding together to stalemate a major power, since all supply centers are equally valuable. However, we can use Chris's formula to distribute half the pot, as long as we distribute the other half of the pot equally among all those who force a draw! If there's an outright winner, then the 'draw' half of the pot goes outright to the winner. That way, if you're in third place in the endgame, instead of trying to decide whether you can make exactly 8 supply centers by stabbing fourth place, you can weigh the risks of stabbing fourth (first place might win outright, depriving you of your share in the 'draw' half of the pot) against the rewards of stabbing fourth (you get some extra supply centers, and thus a larger share of the 'supply center' half of the pot).

Fastspawn, I agree with you that 8 SC is an arbitrary number--there is definitely some sense in it, maybe even more sense than any other number. In my opinion, though, there's no need to pick *any* number. Providing flexible incentives that reward both drawing and gaining centers will lead to a more interesting variety of strategies and of endgame scenarios. People will have more fun.

(3) It's easy to reward people who seek out new opponents. No game should be zero-sum, so in advanced games, ie, stake-games, instead of paying out $5 for finishing plus $1 per supply center or $25 for winning, pay people who finish the game (by being eliminated, drawing, or winning) according to this formula:

(Buy-in Stake) / (Total previous games played with all opponents + 1)

For example, if you played an intermediate-level game, wagering, say, 68 Diplomacy Dollars, and there were two opponents you'd seen before, one of them once and one of them twice, you would collect (68) / (2 + 1 + 1) = 17 Diplomacy Dollars just for finishing, a quarter of your buy-in.

If, on the other hand, you'd played just about everybody in the game at least four times, you'd only collect about (68) / (6 * 4 + 1) = 3 Diplomacy Dollars, rounded up. Of course, you can still win stakes off your opponents by beating them in game after game...if you play with the same people all the time, and you keep beating them, your rating should go up and theirs should go down. There's a nice limit, though, since after they lose all their points to you (since they lose them to you faster than they get them back from the increasingly-tiny finishing bonus), they won't be able to play stake-games with you, and you'll have to go find new opponents or be reduced to preying on newbies, which doesn't pay well enough to make you one of the website's Top 10 players.

(4) How do you keep people from joining 30 newbie games at a time, ordering a hold every turn, and collecting 5 Diplomacy Dollars a pop when they're eliminated? Simple. The community agrees on a limit (2 games a week, 3 games a week) for collecting newbie finishing bonuses. You can still play as many games as you like, and you can collect Diplomacy Dollars for winning or drawing as many games as you can manage to win or draw. You just can't collect points for simply being in the game beyond the level of simply being in the game that we're trying to encourage.

How do you keep advanced players from joining 30 games at a time and getting rich off the 'finding new opponents' bonus? Easy. If you're eliminated from an advanced game, even if you've never ever played with anyone in it before, you still only break even. The stack of bonus credits added to advanced games just isn't big enough to reward people for playing a zillion games. Of course, people can play as many games as they like...but if they lose almost all of them, they'll eventually lose credits, and if they go into civil disorder on a regular basis, then they won't even collect the new opponents bonus, so they'll lose credits so fast that they'll be back to newbie status in no time at all. And the newbies won't mind so much eating up countries that go into civil disorder--it means they can play with the big boys that much faster.

Oh, and I'm sympathetic to figlesquidge's holiday scheme. Don't know how to work it out, but it sounds like a good idea.

Finally, before you jump on my post as insanely long and complicated, keep in mind that (a) I've been engaging with everyone else's suggestions and concerns, (b) I've proposed solutions to more than one important problem, and (c) everything I've suggested can fit into 100 lines of code for the programmers and 1 menu screen for the users--it's only complicated to think about; it's not necessarily complicated to make or use.

Thanks for the input all, some interesting ideas floating around. One thing I'd like to address is Writhdar's comment about simplicity.
figlesquidge will tell you I'm pedantic about making things as simple as possible, so we are definitely on the same page there.

The points system will be a black box to players, they'll only have to know that they put some of their points in as a stake, and they get points back depending on how well they do in the game. It's pretty intuitive, in my opinion.
I think that problems with abuse and civil disorder all need to be taken care of somehow, and if a points system can help take care of: civil disorders; allied victories; keeping inexperienced players out of games for experienced people; multiple accounts; and people joining too many games at once; all in one feature, I think the simplicity of having one thing take care of all that is better than having several simpler things try and take care of each.

Chrispminis, you misunderstood me. If we reward by the number of SCs you control at the end of the game, then there will battles for third place, or even fourth place, which DISCOURAGES joining together against the winning player.

Kestasjk, please take what I am saying into account. I don't mind a credit system like you describe for first or second place, or maybe even third place. But if you continue rewarding according to number of SCs even down to those players that hold only five or six SCs at the end of the game, then you will be encouraging them to fight amongst themselves.

If a credit system is implemented, you MUST give an equal reward to all lower tier players. If you give a lesser reward to even those with two units left, then that player would do best credit wise to work together with the winning player in order to reduce another player to one SC. Using a credit system may solve many problems, but one new problem it will create is that the system itself will start _encouraging_ players to prey upon the weak rather than to band together against the top player.

It occurs to me: what if we used a non-zero sum credit system, but instead of adding the additional credits to the top end, we apply it to the lower end, by equalizing the lower tier players' return?

Say we used the system Chrispminis described, but whenever a player _would_ be getting a net loss, instead we add credits to cause him to break even? Then players could only lose credits by actually losing all their territories, and a player with four SCs would gain zero benefit from destroying a one SC player--although the top player would indeed gain a small amount from killing off the one SC player. This actively encourages the smaller countries to band together against the larger enemy.

OK guys I'm just giving you a suggestion here. Diplomacy has been played for quite a long time now, and there are many player groups around. One of them is www.dip2000.org. They have this Eagle Point System, that:
- rewards both solo and shared victories, the former more than the latter, because it should be clear that solos count more than draws
- punishes whoever lets the country go into civil disorder (in fact they ban the player from the site altogether, which I think is a good thing, but I've also been accused of being too strict)
- gives some points to whoever ends the game on the losing side, but always less than those given for a draw (it is better to draw than to stay alone)
- establishes an overall ranking for all players, and (wink @ Chrispminis ) sets the value at stake for each game based on the ranking of the players that are involved (basically, a game played by experts is more valuable than a game played by rookies)
If you IT wizards check the www.dip2000.org site, you will find an EXTENSIVE description of how the system works. I think it would be easy to get the needed logical statements out of that description. And maybe we could even adopt/adapt/improve that scheme...

Hmm, Jason has a good idea with the splitting of the pot like that. It would encourage more banding up against the big guy than mine would. I don't really like the idea of limits on games, I think the scoring system should be designed such that limits do not have to be in place. And that there would be natural limits.

and @Writhdar
I did say that you could play Unrated games if you choose. And that there would be two ratings. One which is your wealth, and the other which is your win percentage. If you don't care for the credit system you could simply play only Unrated games.

@AntoniusRex
Sounds interesting. I'll look at that later and comment on it. It would be helpful if you posted some example numbers.

While I don't think I've ever fallen into Civil Disorder, I can see why people might. Sometimes games last longer than expected and a holiday could force you to forfeit the game. Players shouldn't be banned for that. A limit on it wouldn't help either, so I thought just a straight penalty, which wouldn't be crippling, but wouldn't be laughable either would be good.

The good thing about a currency system is that its highly proportional. So people who leave high stakes games are more penalized, and should be, because with so much at stake, a player leaving could ruin the game.

As well, it's an easy way to separate to a degree "Experts" from "Newbies". And with limited currency it's an easy way to balance out a higher rating for those who play more games at a time.