I don;t know what everyone else thinks, but I personally don't think that PPSC games ought to be scored differently than WTA games in the Ghost rating. PPSC is designed to keep people who are not super into the game (yet) from quitting when they are losing, as far as I can tell. These people likely care little about their ghost rating.

And in PPSC games, one three-way and one five-way draw are better than a solo in the Ghost Rating, which I do not think is fair, personally. Ghost Rating is supposed to capture who plays diplomacy best the right way. PPSC is not the right way, so why support it?

I guess I'm still not buying the basic premise: that people's play style is influenced by Ghost Ratings.

I think Ghost Ratings are just fine but PPSC shouldn't count in those calculations.
I have almost exclusively won PPSC games not WTA. Draws are a common result in WTA because the players are more experienced and will usually not let anyone solo.
Some of the very big players can still win in wta though.

@uclabb
I don't think it makes much sense to have GR reward results in PPSC games in a different way to WTA does. It would create a lot of confusion, and wouldn't really be fair to someone who thought a strong second was a good result, as it is by the reward system of PPSC.

Maybe an argument could be made though to count PPSC games as less important and influential in GR than WTA games are, rather like gunboats are given less importance. This would reflect the general view the WTA is a better reflection on the game of diplomacy.

Personally, I agree with TGM on the issue MM raised- the system we have now already heavily rewards wins and changing it would bring about other complications as he described.

Interesting discussion.
@strat: maybe not influenced directly, but people do play to "do well". If drawing is considered to be close to winning (or in any case an acceptable outcome) then people will change their playing style accordingly.

For instance: if all games were WTA everyone would value drawing as a clearly less desireable outcome.
As people commented we need PPSC to stop people from going into CD. That is a shame, but i suppose there is other ways to achieve the same goal. I've heard things like a reliability-rating or even warning signs next to players who have gone into CD in their last few games.

If going for the win is much more true to the nature of the original game then i say it should rewarded more. But i haven't got much experience with the original game. I learned diplomacy mostly from here. I do know that WTA or PPSC makes for two totally different games, but not sure if its a bad thing. Every person enjoys another playing-style.

Would be insightful though to see a "win, survived or defeated"-only rating. That would more accuratly reflect solo-skill. Or a WTA-only rating. Something like that.

I wish I had something wise and thought provoking to say. I do not. I just wanted to know how to find out your GR? Is it listed on this site?

@MM - I don't think you can neglect the 007 vrs 160 vrs 250 question.
By GR, 007=160 (well, almost - I think it actually depends slightly on the order currently).
Now, if you feel that winning is more important, you might equate 007=250. Fine, but now you're saying a win and defeats are better than draws.
I know this is quite obvious, but my point is that until everyone agrees on the relative merits of each result the rating system shouldn't be changed.

Hmm, strange. I understand the problem, but have a feeling the logic is fundamentally flawed.

1. Making solos more valuable will only make them more rare. Look at the TMG tournament - last time I checked there were only 3 wins in 49 games (?). If you want more wins play PPSC.
2. I also had more than half wins at some point - and still have to respond to awkward "alliance, hmm, you have more solos than draws" openings - but this comes with the territory. It's the same for everyone else. Plus, it's not realistic - not if you're playing against top competition at least - people have to understand this is not some conspiracy and it is not related in any way to scoring systems. You can take the top seven, most vicious players, based on any and all scoring systems, put them against each other and see how many solos it will produce. We've actually had games like this. Massive draws mostly. People found out that first you need to survive, then think about winning - and it's not because they are playing 'alliance'
3. As with most things in life, leaving does not help. One can suggest a new rating system, organize a tournament to their liking, play private games, play anon WTA, etc. What we have here is actually very good and gives a lot of options.

Btw, I'm considering drafting a new tournament idea - all anon, PPSC, League-style. Should have more solos than anything before :)

Interesting.
How would you expect it to be played: ie would that be definitely saying take the 14sc second over the 3-way draw?

@rhino86, go here: http://tournaments.webdiplomacy.net/ and look for the ghost rating section.

"Making solos more valuable will only make them more rare. Look at the TMG tournament - last time I checked there were only 3 wins in 49 games"

That's the point. Wins are so rare that they should be rewarded more greatly. As for rarity change and such, it's actually reciprocated. Wins will be less apt to happen because others will expect it. In contrast, there will be more wins because more would go for it. All in all, the rating system doesn't change the game; it just determines the points/rating.

I think having a deposit equal to the pot buy-in that would be forfeited if you go CD could help a lot with CDs. But then there is the question of who gets the deposit? Goes into the pot? Then it might just go to the single person that got the win due to the CD. Better for it to be split between the eliminated/survived players to refund them some points due to the game being messed up by the CD.

The problem I find myself having is that I go for the draws to survive. In order to be safe I build safe alliances and then don't ever feel I'm in a position where I could go for the win. My stats put me pretty firmly in the 0-0-7 category. But I feel like I'm missing out on not ever getting wins. Setting up a draw is so safe. Of course I also almost only play league/masters games so I think the quality of players is pretty high.

What about making it so that no one profits from a draw. A draw means no one wins, so everyone gets their points back whether in the draw or eliminated. That would discourage going into a game just looking for a draw.

But then you aren't measuring the skill that it took to not be eliminated and get to the draw position keeping someone from almost soloing. So it just doesn't feel right either.

I like solos more in PPSC as they are not as dangerous as one in a WTA game. In PPSC others are reluctant to stab as often or change sides, whereas i find in WTA once u gain a lead of 2 or 3 sc, u lose ur allies... and then the game continues to shift like that.
Thoughts?

That's why I mostly only play WTA.

"What about making it so that no one profits from a draw. A draw means no one wins, so everyone gets their points back whether in the draw or eliminated. That would discourage going into a game just looking for a draw."

That would be a fun game option. No one would ever go for that every game. Most people would think of it as wasted time, but I think that's golden. Good luck trying to win that game.

Ghost, I didn't have time to respond fully, but one of the main points of this thread was to question your basic assumption, that every game should be zero sum. I'm suggesting that the rating sum is not required to be similar to energy, where it is neither created nor destroyed. In fact, I'm suggesting GR be created by means of destroying it! (OOoooohh)

So, as an example, let's say there are only three outcomes on this site with the following result:

1. WTA solo = 100% GR of everyone in the game.
2. WTA draw = 100% - (5% x number of players in draw) GR of everyone in the game.
3. PPSC game = 75% GR distributed how it's distributed now.

So, with a WTA solo, GR gets awarded as it does now. With a WTA three-way draw, three players evenly split 85% of the total GR in the game, while 15% of the total GR in the game gets destroyed. With a PPSC game, regardless of if there's a solo or draw, the GR is distributed as it is now, but only 75% of the total GR gets distributed and 25% gets destroyed. The specific numbers I'm using are arbitrary and would need to be fine tuned (see below), but by destroying GR in WTA draws and all PPSC games, then a WTA solo would be more highly rewarded than it is now. Yes, I understand that three three-way draws will not increase a person's GR by as much as a single solo, but this concept I'm presenting can be fine tuned a bit to increase the weight given to a WTA solo.

So, why destroy GR? Well, it seems to me that, just like with points, GR is inflating over time. I quickly looked back at GR from June 2008, June 2009 and June 2010 and compared the GR of the #10, #100 and #500 rated players to see how their GR compared:

June 2008 June 2009 June 2010
10th - 306 10th - 378 10th - 414
100th - 144 100th - 200 100th - 294
500th - 94 500th - 115 500th - 134

It looks to me like there is indeed GR inflation. I'd think you math/computer guys could figure out a way to assign values to my concept above that will result in GRs that have little to no inflation over time, even if just by a little trial and error.

Ughh, long day, I'm tired, will comment on other people's comments more tomorrow. Thanks for all the feedback so far, I think it's an interesting discussion.

“As people commented we need PPSC to stop people from going into CD. That is a shame, but i suppose there is other ways to achieve the same goal. I've heard things like a reliability-rating or even warning signs next to players who have gone into CD in their last few games.”

One alternative is to make each player put down a deposit for the game, so that going into CD looses the bet + deposit, as Alderian mentioned

“I know this is quite obvious, but my point is that until everyone agrees on the relative merits of each result the rating system shouldn't be changed.”
Also, I’ve never wanted to change the reward system from points to Ghost-Rating. If you play for points in a game, you also play for ghost-rating. A change would alter that.

“So, why destroy GR? Well, it seems to me that, just like with points, GR is inflating over time. I quickly looked back at GR from June 2008, June 2009 and June 2010 and compared the GR of the #10, #100 and #500 rated players to see how their GR compared:

June 2008 June 2009 June 2010
10th - 306 10th - 378 10th - 414
100th - 144 100th - 200 100th - 294
500th - 94 500th - 115 500th - 134"

But the percentiles of those ranks have gone down, Looking at percentiles:

June 2008 June 2009 June 2010
1% - 309 - 300 - 317
10% - 145 - 146 - 150
50% - 94 - 94 - 97
90% - 74 - 78 - 78

Which shows a slight inflation between 2009 and 2010 , but if anything a deflation between 2008 and 2009. That is consistent with the fact that the average game count went down from 2008 to 2009 very slightly (~half a game), but went up by about 4 games between 2009 and 2010. Ratings at the top tend to go up as game count goes up, and ratings at the bottom tend to go down (and vica-versa.

"But the percentiles of those ranks have gone down, Looking at percentiles:"

I have no idea what you're talking about, Ghost, but I trust you are not concerned with the top players' ratings continuing to climb for a good reason regardless of if I understand it or not. I still like my idea, though, even if it ended up being something closer to 2% (or even 1%) times the number of players in a WTA draw gets destroyed and 10% (or even 5%) of a PPSC game gets destroyed.

Anyone else besides Ghost have a comment on this concept of "creating" extra GR for a WTA solo by destroying GR for all non WTA solos?

MadMarx - What TGM means is that if you are the person in either 5th place our of 500 people or 15th place out of 1500 people (to use examples), then your GR would be the same over time (since you are in the top 1%). So as more people are added, then it actually makes sense that the GR of people at the top goes up. If there are only 100 people total, for instance, then being the 50th person does not mean much (you are average). If there are 2500 people, then being the 50th person means you are in the top 2%, so you should have a higher rating.

Also, I think you must have a zero-sum method of ratings assignment or else you will eventually suffer ratings compression (everyone has the same rating), which is even worse than inflation.

Thanks, nola, that makes sense!

Not to be all soft or nothing, but as an example, I think Rait is by far the best player this site has ever seen and I don't like the idea that his accomplishments get washed away over time (which isn't to say he won't come back and kick all our butts again very soon). I think it will be difficult for everyone on this site to ever all have the same rating even with a little bit of deflation, such as the 1%/5% version of my last post, but I do see your point, thanks again.

PS - If we do start to see serious deflation, then the modifiers can just be adjusted, nothing has to be set in stone.

But why shouldn't a player who has gone away have their rating lost over time? They are doing nothing to keep it up. I like that someone who I can't challenge can't remain on the top of the pile. King of the mountain shouldn't be a title for life. you should have to defend it regularly. Just like in boxing. You can go out the champ, but then someone else will get a champions belt and you will be the "former champ". Same thing here. You can't reach number one then quit as the all time winner because no one can beat you if you don't play.

Maybe what we really need is a "challenge" system where you can challeneg players above you to a game and, when you get seven, the game starts. But if they refuse too many challenges in a row, they lose points and GR and those who challenged them get those points and the GR boost.

Good point Draugnar, but I'm the type that likes baseball less because the steroid era has wiped out some very impressive records of the past when they shouldn't have, if you get the point I'm trying to make.

@MM

Rait looks to be a brilliant player - I haven't played much with him unfortunately - so clearly my opinion is based on limited information - but looking briefly at his games it is easy to note he's never won a game when you or me were present. Actually he always finished with a lesser or equal result to you or me. So I'll once again say that it really depends very much on the competition.

If someone does not want their ranking washed down the right thing to do is play and keep winning, maybe win more than before. Stopping time is never the solution :)

Plus, what difference does it make if someone kicked ass a year ago - one's form may change, there are new people coming in, some of them very good - if you can't beat them don't blame the system :)

There're no steroids here - and I'd actually argue it's much harder to win now than it was two years ago.

But there are no steroids to be had here. Not unless you have some good sapho to speed up the mind with.

Oh, and a record is different than a ranking. The win/loss record would still stand. It's the ranking that is diminished if you refuse to compete.

Drag +1