A place to discuss topics/games with other webDiplomacy players.
Start a new discussion in the public forum
Post a new thread
If your post relates to a particular game please include the URL or ID#
of the game.
If you are posting a feature request please check that it isn't mentioned in the todo list.
If you are posting a question please check the FAQ before posting.
If your message is long you may need to write a summary message, and add the full message as a reply.
I noticed that Germany is considerably weaker in Diplomacy than it's historical counterpart. In game, Germany can very easily be knocked out by England and France. However, historically Germany was able to fight competently on three separate fronts. Does anybody feel that Germany is a little misrepresented in game? More importantly, are there any variants that address this?
In an effort to compromise the pro-ads versus anti-ads for games: Post here for your non-live games to cut down on the number of ads but still advertise games. Post game link, WTA or PPSC, and the bet. Note: this doesn't count for special rules games.
Hi all, when Amby and I were recording our recent podcast (diplomacygames.com) the topic of Bourse came up, and I got really excited and thought that I would try to get another game of this variant running...
Awhile back, kaner proposed a Known World gunboat tournament in which 15 participants would play 15 games, one with each nation. I searched back for the thread, then just decided to start a new one. I want to see if there would be sufficient interest in this to try to get it off the ground.
Hi folks, some of you may have heard me talk on the podcast about bringing the WWIV map to a Cold War circa 1984 1v1 variant. Interested in your thoughts about whether I use the standard WWIV map, the v6.2 version (is there any actual difference in the map itself?) or whether the sealane version would be better. Thoughts?
As many of you know, the Calhamer estate is being liquidated and the very first self-published Diplomacy board sold for just over $5000 last week. Well, something else interesting from the same sale - a bunch of prototype maps, these likely being from several years before the game was published.
#1 My username shows that i have 1690 while the Hof shows i have 1693 . If i look at the last game it shows i got from 1693 -> 1728. Doesn't really add up all together. Is there a reason for the differences?
Also the list of my games doesn't show the last games: http://vdiplomacy.net/hof.php?userID=4661
#2 When i look at the result i really wonder why Phil1986 only lost 2 ? Unfortunately i didn't find an explanation on how the points are calculated. I'd be quite interested to know more about the mechanics.
You're welcome, Devonian. Now, like then I think that the pairwise method of splitting a multiplayer game was a bad choice. Before that this system was put in place I raised numerous concerns about players behaviour that by a matter of fact revealed to be well-founded. Now, like then, I think that your ancient proposal similar to webdip Ghost Rating should've been a better way to approach a rating system for this particular website.
Unfortunately Oli persisted with his formula. Regardless of this, if we’re meant to have somehow a measure of skill, the current system is way better than the old one… and I don’t get why the D-point system is still around.
Something weird is definitely happening - at some point in the last day or so (apparently before the 1st FoW 2 game completed this morning) every game I'd completed since mid-Jan stopped showing on my hof page and are not counted to the displayed v-rank.
After reading the 25 pages i can summarize the following #1 Oli left the old point system in place to keep rewarding people when they win/draw a game. In the ELO system it could be you actually don't get any points as although you won the game you performed below expectation
#2 @decimalegio, what kind of negative behavior do you see as a result of the ELO?
#3 My biggest riddle still is the mV value. I don't know how we get to it.
In this game http://vdiplomacy.net/hof.php?gameID=27297 phil1986 has an mV of 1% the best explanation i can come up with after reading the 25 pages is that he it is because he joined the game later?
I noticed a recalculation of all the historical points took place (possibly because I'd mentioned an error in a game from a while back, possibly for some other reason). My guess is a game that finished early this year crashed the process, and so then a bunch of games from the subsequent months weren't counted. Once my score got recalculated with the next game ending, I went down almost 300 , but it is still excluding those missing months (when I had a hot streak (now ended)). My guess is someone will be able to figure out why it crashed and a recalculation can occur again. I also imagine it won't happen all that quickly.
#1 Oli left the D-point system because it is inherent within the Webdiplomacy platform code. As far as I remember, the Elo (not capital letters or the Chess player Retillion will have a stroke) system was designed by Oli in such a way that win and draws do not result in a V-points drop. If this didn't happen, it is due to non-standard conditions: altered victory conditions, civil disorders, bugs...
#2 "top" players tend to be afraid of losing V-points / position in the HoF. Because of this they tend to play only in elite clubs more than before. Either way, they opt for Unrated as a "conditio sine qua non" for their participation to standard games/tournaments. "bottom/average" players tend to tear down "top" players in order to get more V-points in reward when the game is over. This regarding to players that care about ratings. I note this as a difference of behavior among V-points and the old D-points. Regarding to players that do not care about ratings, there is obviously no difference of behavior.
#3 I can't recall how the match Value coefficients (mV) are calculated. It's a number between 0 and 1 depending on the relative SC count of the couple of players. I think that also the WTA/PPSC setting also factors on mV. mV is reduced when one of the players did take over.
I wonder if a ELO points decay has ever been considered. This is used in other online competitive communities, partly to combat effects like Decima's point #2, and to encourage play. A separate "all-time high" list could be maintained to look on past glory, while primarily maintaining a current ELO list.
I don;t think that you consider a points decay feature until you address the issues that arise in #2 in the first place. No one wants to play every game they enter with a bounty on their head. And the reality is that players have been targeting the top ranked players. This is why cypeg stopped playing. Yes, there's a vanity aspect to it, but the core issue is the problems with the ELO formula that create institutional headhunting. Once that is fixed, it would be okay to decay points from inactive players. Another idea would be to base the ELO on games within in a certain period say 12-months or 48 months. At the end of the day though any scoring system needs to be focused on rewarding player performance rather than tearing down others. Strength of opposition is a valuable factor to consider, but we need a way the doesn't penalize as much when top players get pigpiled for points.
In non-anon games, I don't think targeting the players with better records is ever going to completely go away, though. You can always look at the player stats, and even if you hid that, you can still potentially learn the reputations of the players. Realities of a multi-player game where cooperation is everything.
Don't know much about other Diplomacy rating methods, although I did read this, http://www.playdiplomacy.com/forum/viewtopic.php?f=565&t=34913 , the PlayDip Elo FAQ. (I haven't played there much so I have no feel for how good or bad the ELO ratings seem.)
One thing they mention that I didn't like, which is the possibility of draws lowering your rating. I think if you draw or win, you should at minimum get 0 . Although potentially losing points might force high-rated players to play for a better result, it probably also dissuades them from joining a game in the first place. It's not a perfect solution though, since the possibility of being eliminated by a group of much lower-rated players also weighs heavily if you really care about your high rating.
Not sure how it works in vDip, I should probably have a look at the code... or does anybody know the vDip rating equation?
Why shouldn't a draw between two low rated players and a high rated player cost the big guy points? Clearly, he didn't play up to his potential. It shouldn't cost as much as a loss, but it should cost something.
Because then he won't want to bother playing at all, is the main reason, if even a good result doesn't reward him.
Some more food for thought: http://www.stabbeurfou.org/docs/articles/en/DP_S1998R_Diplomacys_New_Rating_System.html
I have a hunch that you shouldn't simply linearly increase the rewards as the number of players goes up, due to the "chaos" of having more and more players. Not sure if anyone *does* increase it linearly though, but it would be the simplest thing.
You can see (with very little explanation, though most of it is self explanatory) the v point calculations for a game through the Hall of Fame. Easiest way I've found to navigate to your own listing is go to your profile then edit the url, replacing 'profile' with 'hof'.
By the looks of it, it's impossible for you to lose points in a draw. It's basically 1-(expected win prob. vs defeated foes) x (players in draw and leaver consideration) x (dependent on number of players). You're not compared to the people you drew with, just given a proportional amount of what you'd've gotten for beating everyone else.
Nice trick mouse, that shows a lot. So from some quick digging:
Winning Known World 901 (15 players): ~ +500 5-way Draw Knowing World: ~ +50 3-way Draw Knowing World: ~ +100 Losing Known World: ~ -40 Winning Atlantic Colonies (4 players): ~ +20 Winning First Crusade (7 players): ~ +100
Obviously these could vary a lot depend on other player skill, and I only glanced at a few games. But that win number for Known World is huge. Yes it's a map of 15 players, but it's not like the winner eliminated 14 players. He beat his neighbours, then a few more neighbours, then got half the board. Meanwhile, other players were being beaten by yet other players throughout the game. Is it worth 5 times as many points as a win in a 7 player game? Seem like there's a factor that blows up with big games.
Is anybody able to make minor cosmetic map changes to the variants, for readability? There's two small things I've noticed:
* Imperial Diplomacy II: there's a connection between Morocco and W. Med, but the map really doesn't show it. * First Crusade: The Sardinia supply centre in the large map is placed in an odd, almost invisible spot.