"As I understand, Rinascimento isn't counted in the rankings on account of its unbalanced start."

Leif - If this is the case, I suggest we remove Imperial Dip II from the rankings as well as that in another case of a greatly skewed start position. I believe those are the only 2, but I could be missing some.

Imperial II is not that screwed as Rinascimento.
In Rinascimento there are very large countries like Napoli and Venezia that are much more likely to solo or even survive than Siena for example, even with the best diplomatic skill.
Also Rinascimento is a really liked variant here. It would be a shame if people afraid of their ranking would avoid these games.

As far as I can tell, it looks good.

Except that I am no longer in the #1 position... Oli, please work your magic and correct that for me. :-)

I beg to differ on your analysis of Imperial II Oli. The reason I say so is that it is nearly impossible for the tiny powers to win. Just like in Rinascimento where Ferrara or Siena are never going to solo barring some ridiculous NMRing and CDing, simply because one of the 3 large powers (Stato/Napoli/Venezia) is almost guaranteed to survive and be in a much stronger position. In the same way, how is Mexico or another 3 SC power expected to win in IMP II? Even if the large powers fight, one of the 4 (France, Russia, Britain, Holland) is going to come out on top in at least one area of the map that gives them an extreme advantage over the localized powers that start small.

In theory anyways. I realize there's a very small sample size, but the best result of the "smaller" countries so far has been me and GOD as Japan and Mexico in gameID=11397 where we were still almost 20 SCs behind France and Britain and hemmed in. I suppose one could make the argument that if the small powers banded together they could take down the larger powers, but the reality is, getting that to happen is nigh impossible.

Anyways, just my .02. Feel free to ignore me : p

As for Imperial II, I think I agree with Oli that it is much more balanced. Pirates is another variant with some odd weightings for start locations, should that be excluded too?

What about Youngstown Redux?

Or any non-symmetric map that doesn't have extensive balance play-testing like classic?

The argument can be extended as far as you wish, and where the line gets drawn is arbitrary for a site focused on variants. I would personally say include Rinascimento and don't exclude any variants.

Possible options are:
Let the community vote which *imbalanced* variants to include/exclude
Allow unranked games (option chosen at game creation)
Normalize based on starting SC's (or country performance if SC's are equal)

Those options aren't exclusive either, especially since we will be looking to have multiple ranking systems.

I think an 'unranked' option for game-creation would be a great idea, if possible. That would make SRGs a lot more appealing, especially unbalanced ones like Shift.

"Allow unranked games (option chosen at game creation)"

That might be a good feature to add for several reasons. Play testing... Imbalanced variants... Highly random variants... Games between highly varying skill level players... etc.

I will add a caveat...

Oli,

I know you already mentioned you weren't interested in the idea of allowing unranked games because that option could be abused. I'm not trying to generate a groundswell of support to try to overrule you, but I don't understand the abuse reason you cited. If you can provide more elaboration, that'd help inform my thoughts on why I (and apparently others) think we should have that option.

I updated the Wiki with the actual calculation:

@Leif:
The idea behind the rating is that a rating that is too good for your performance (because of a lucky Survive in a 34-player game) will likely come down very quick, because the penalty in the next games is very high. If you could create unrated games players with unexpected high-ratings can choose not to join rated games anymore and sit at the top of the table forever.

Also each game will just be one single data-point in your rating-history.
If you fail below your "real" rating, you will get a bonus next game, as your expected performance is lower than it should and a good result will score even more.

"players with unexpected high-ratings can choose not to join rated games anymore and sit at the top of the table forever"

All the more reason to record both max rating achieved, and have a monthly leader board in addition to the current rankings.

The reality is that
a. while this rating system works to correct a person's rating toward their expected rating, this ideal expected rating can never actually be reached (it will always be fluctuating), thus recording the max rating a player has achieved means that players can still point to their best performance without having to worry about risking that performance in future games (while others can point to current rankings for their ego trip).
b. a person's rating is expected to change as a player improves his play style, so having a monthly leader board allows a person to track their improvement through time without having to worry about losing an 'all time best' status.
c. The majority of the community is here to play and won't be looking to 'pad' their ratings with an uncharacteristic performance and then stop playing, but will be looking to play additional games, so this is likely a corner case for a few individuals anyways.
d. any 'sitting on top of a rating rewarded by an uncharacteristic victory' behavior will be naturally discouraged, as such players will suffer the 'you are a coward glares' from the rest of the community if they refuse to play any more unranked games after a certain point.
e. additionally such behavior is easily remedied by removing non-active players from the listed rankings (had no ranked games in the past 3 months, don't list them), and/or having the monthly leader board where the non-active player would drop off the rankings.

The benefits of having an unranked games option have already been mentioned, and it could even be a flag that the mods could set for a SRG's if you don't wish to give the power to the people, and then coded in as a user's choice game creation option only for certain variants (like Rinascimento, Pirates, Youngstown, Imperial II, and others that have imbalanced starts).

I might change my point of view, but if we allow this we need to start over from 1000 for everyone, as many will try to get unfavorable games excluded...

A mods-only for some SRG's option sounds more reasonable...
But in the end we need a working rating-system first...

Agreed, I am simply spouting features.

I don't like the idea of starting over, and would still be comfortable with the decree that no previous games can be excluded.

I think we are close if not there on the first rating system though, hence my having begun to list/discuss further options. I will pipe down if more needs to be discussed for the current rating system.

I'm still confused as to how a game that I lost to another player's solo grants me almost twice the number of points that I won from soloing another game? How is this possible. Or as Gunnery Sgt. Hartman would say, "Why are you not stomping Private Pile's gut's out?"

Game ID references would help..

Can you give me the GameIDs?

What does "SRG" mean please ?

Special Rule Games:
http://www.vdiplomacy.com/wiki/index.php?title=Tournaments_and_Special_Rule_Games

Oli, there’s no need for the gameIDs:

Ruffhaus,
let me use my narcissism...
you are comparing a strong survival position in a Point Per Supply Center vs a victory.
And I am ready to bet 1000 that the first one was achieved in a large size variant, the latter was achieved in a smaller-size variant.

Thank you kaner406 :)

"And I am ready to bet 1000 that the first one was achieved in a large size variant, the latter was achieved in a smaller-size variant. "

And you would be wrong, and not for the first time.

Here are the games in question for those who asked:

Whatever Happened to WWIII? (WW4)
http://vdiplomacy.com/board.php?gameID=9310
Result = Defeat (as Indonesia 42 SCs) to solo victory by Dr. Recommended (Nigeria), Points Awarded: 19

Colonial Empires (Imperial II)
http://vdiplomacy.com/board.php?gameID=11411
Result = Solo Victory by RUFFHAUS 8 (Britain 72 SCs), Awarded: 15

Africa (Africa)
http://vdiplomacy.com/board.php?gameID=10994
Result = Solo Victory by RUFFHAUS 8 (Ethiopia 36 SCs), Awarded: 10

So I get the WW4 is a large map variant, and the it's possible that the map size and player rankings might tilt the scales some. But Imprerial II is not a small map. It have 172 SCs on it. The Africa map is not small either. It has 63 SCs in it. The point that everyone seems to be missing is that a game lost to another players solo awarded 19 , which is more points than two games game won by my solos.

How can you lose a game, and score more points than any victory? Losing to a solo should be worth less points than a 1v1 win or a Sail Ho game. At some point victory has been cheapened, and losing is being championed.

@ Decima Legio: There is no such thing as a "strong survival" when survival is used in the vernacular that it is here, meaning surrendered a solo victory to another player . A solo victory means just that: solo. As in one. One winner, and everyone else loses. The reason that games set victory conditions at 50%+1 SC (as in the number of 18 in standard Diplomacy) is because once a player reached 18 centers he has more than 50% of the supply centers and is tactically unstoppable. Therefore the game is galled because it's pointless to play on. There is not 250 army hand of cards to turn in with Diplomacy. It's over, and the soloist wins.

Why do I harp on this? Because by setting the scoring vehichles this way, we are teaching players to play the game incorrectly. We are rewarding failure by tricking players into thinking that a 16 supply center finish to an 18 SC solo is a good result when it's terrible. It's worst possible result imaginable. The previous Dipoints scoring system already encourages the notion of "strong survivals". Why would we compound the problem by further encourage players to throw solos so that they can earn more points?

@Ruffhaus:
We had this discussion between PPSC and WTA already. I won't change this. My system will work like to the current DPoint system to avoid confusion.
We all know that a PPSC has not a "Winner" like the WTA-games, ist just a race to the most SC as fast as you can and once one player reaches the target SCs it does end. And the player that triggered the end is called winner, but that does not mean everybody else lost.
This is not in the spirit of the original game (That's a WTA-game), but it's thge way many people here like to play the game. And as I like to keep the DPoints around I don't like the idea that in the same game people have a different reasoning for their play. Some player playing for maximum DPoints and some for Maximum VPoints would really screwthe game much more.

A valid point is the value of the large games. I'm still experimenting how to sdjust scoring to large games. The formula I have now usually makes the defeated loose about 7-10 and the winner/survivors winning the more, the more they defeated.
That's still a lot of points for the 20+ variants.

Ruffhaus,
for the N-th time, if you only knew what you’re writing about, you should be aware that when we talk about the “size” of the variant, we are referring to the number of players taking part in the variant.

So, no matter if you like it or not, Classic-Chaos and WWIV are LARGE size variants, while Imperial II and Africa are MEDIUM size variants. For what it’s worth, my bet would be definitely won.

The absolute SC number has never been a significant parameter for rating calculations. Not for WTA of course, but for PPSC the only significant measures regarding SC are the final SC relative ratios.

Decima, for someone that runs around accusing people of not knowing what they are talking about you run your mouth pretty reckless. And neither you nor Oliver has bothered to address why we are awarding a similar number of points for failure as we do for victory.

If you want to get bogged down in the semantics of SMALL, MEDIUM, and LARGE, go buy a value meal at McDonalds. Thes size of the game as defined by the number of players has nothing to do with measuring the achieveament of the perfomance of a player in that game. But if you had the balls to play against the real competition on this site you might know that.

Oliver, I do appreciate all your work. Unfortunately I think that it's fatally flawed, and encourages players to allow solos to occur. The VDip Points system already does this, and now you admit that it is your specific intention to have the new system do this as well. Clearly there's an absolute disconnect between us on the definitions of victory, loss, success, and failure as it relates to Diplomacy. This is your sites, and yours to run as you please, but his thread solicited opinions, and I have given mine. My opinions are not designed to attack you or anyone else, but to provide insight based on experiences and perpsectives of how Diplomacy games are scored elsewhere. Perhaps no one here cares that players are not learning the game properly here. But it will speak volumes if one of our players show up at tournaments around the world thinking that a "strong second place" means something. And when the other fice players at the table look at him in contempt and hear that he plays his games at VDiplomacy, it's a black eye for us. Maybe no one cares about that. I do. I like this place , even if there are douchebags who cross-game target and cross-forum agruments people that they dislike. I like it even if the scoring systems reward failure, and are akin to the 'everyone gets a trophy' logic in youth sports, which is coincidentally the same logic that destroyed the game of dodgeball (the greatest game ever played) in schools. Thank you for hearing my opinions, even if you've completely rejected them.

Ok, clearly you don't like PPSC idea. Legitimate.

So please, may you explain to this narcissist douchebag (Decima Legio) why do you play A LOT of PPSC games?

Which one of the 4 words in the acronym is not clear to you?

PPSC is a bastardized but popular version of the original diplomacy WTA game style and unfortunately often does emphasize supply center count even in a loss.

If the community's desire is that a solo victory trump all survives for PPSC (which is not how games have been played on this) then I agree the current algorithm is flawed.

The way to fix this (and you would likely have to re-start the ranking system over with everyone at 1000), is to scale the winner's SC count up to 75% (instead of 50%) and have the survivors' shares of the SC's scaled to fit in the remaining 25% of the SC count but with the ratios that they achieved in their survive. This way, a strong survive, while gaining points (to satisfy the PPSC crowd) doesn't gain near as many as in a draw where they would have a larger proportion of the score adjustment, that is if they did their due diligence and stopped the solo and reduced the draw count.

Can this sort of scaling be implemented alongside the current ranking system for comparison?

That makes sense Leif. I don't think you should get (-) points for survive(you did indeed make it through the game) however, as suggested it should be drastically low as a percentage. This kind of scaling may encourage the incentive to make more of an effort to stop solos and not just give them away.