Hello.

The vdip points are a rating system that is more accurate than the dippoints system, as it includes how strong your opponents were and how it ended.

You can find the vdip HoF in Mapu's link.

Thank you :)

More accurate for what? This is silly.

It's a more accurate measurement of playing skill. Your v-points increase more when you beat higher rated players and less when playing low rated players. Also, its harder to manipulate, since you never run out of points and get a reset.

It's more like the ghostrating from webdip. We have these for some years now, but as there is just a HallOfFame, and nothing more they did not get noticed.

I just started to migrate the current dPoints system to the vDip system. I will add a much better integration with an in depth explanation in the next few weeks.

Basically they are distributed like an elo-based system, so you gain more points for beating better opponents than you. Everybody starts with 1000 , and you win or loose points based on how good your opponents are.
You constantly need to outperform yourself to gain more points. So based on your points and your opponent points it will calculate how likely it is that you win against this player and will only award points if you manage to play better than expected.

The complete calculation is here:
http://www.vdiplomacy.com/wiki/index.php?title=Rating

and the current HoF is here:
http://www.vdiplomacy.com/hof.php

I'm just a bit experimenting with this, so there might be some changes the next few weeks.

The hof page only lists the top 100. This let's everyone see their rating easier.

I will add a link in everyone's profile.

Can you also post the complete calculation on the Vdip Hall of Fame page, or at least a link to the calculations from that page? I was looking for the calculations a while ago, gave up because I couldn't find it.

Sure thing.
I do not have much time for programming, so this will improve over the next few days/weeks.

Waking up to see a giant number next to my username was confusing at first, but I am in favor of this ranking system, so long as the calculations can be easily and readily explained, especially to new players.

If you are not listed in the HoF and therefor do not have a direct link to your personal 'vPoints game overview', you can type "http://www.vdiplomacy.com/hof.php?userID=[your userID]" in your browser to get an overview on how much your finished games affected your rating.
You can found [your userID] on your personal profile-page.

This information will probably become irrelevant when Oli adds the links to the profiles.

so i have to go to my profile page now to see how many points i have for joining games?

@y2kjbk: Yes, at the moment that's your only option.
But it's very likely that I remove the dpoints for joining games.

How do unrated games affect rankings?

Not at all.

Can someone help me with the math?
Other than the mV calculation, is the math correct?
Also, how is mV calculated?


Based on:
1v1 WTA game
Both players equal rating of 1000
No replacements
Player 1 wins 18-16



-----FORMULA-----
Expected result player1 (Re1) and player2:
- Re1 = 1 / ( 1 + ( 10^( (Rating2 - Rating1) / 400) ) )
- Re2 = 1 - Re1

-----ACTUAL-----
Re1= 1/(1+(10^(0/400)=1/2
Re2=1-1/2=/2



-----FORMULA-----
Adjustments of the results for calculation:
- Set Resigned/CDed players SCs to 0, their rating to 0 too.

- WTA: Set the SC-count for all but the winner to "0".
- Draw: Adjust the SCs for calculations so that every drawer has the same SC-count.
- PPSC: Do not change anything with the data.

Calculation:
Rr1 = Pl1SCs/(Pl1SCs + Pl2SCs) (if both have 0 SCs set Rr1 = 0)
Rr2 = Pl2SCs/(Pl1SCs + Pl2SCs) (if both have 0 SCs set Rr2 = 0)

-----ACTUAL-----
Rr1=18/(18+16)=18/34=~ 52.9 --- For WTA, set to 1
R2= 16/(18+16)=16/34=~ 47.1 --- For WTA, set to 0



-----FORMULA-----
The score is based on the difference (D1) from both results:
D1 = Rr1 - Re1
D2 = Rr2 - Re2

-----ACTUAL-----
D1=1-1/2=1/2
D2=0-1/2= -1/2



-----FORMULA-----
Adjustments for each match (matchValue=mV):
- Value the importance of take-overs. (If a player bet only the half the whole match is worth only half.)
- Normalize the mV by the SC% difference of both players

-----ACTUAL-----
MV=???? What happens here?



-----FORMULA-----
Adjustments for the game (gameValue=gV):
- The game-value (gV) is set to: ((numberCountries -1 ) / numberCountries )^3 * (1 - (numberCountries / 100)
This gives small games a very low gV and the gV gets more, the more players are in the game.
- gV = 0 for Rinascimento games
- If the winner does not reached the supplyCenterTarget because of custom game-options adjust the importance of the game too

-----ACTUAL-----
gV=((2-1)/2^3 * (1-2/100)) = (1/2)^3 * .98 = 1/8*.98 = .1225



-----FORMULA-----
K-Factor (K) to get some real point-changes from the 0.xx results is set to 30.

-----ACTUAL-----
K=30



-----FORMULA-----
Final calculation for each match:
scoreMatchPlayer1 = K * gV * mV * D1
scoreMatchPlayer2 = K * gV * mV * D2

-----ACTUAL-----
scoreMatchPlayer1=30*.1225*???*1/2
scoreMatchPlayer2=30*.1225*???*-1/2



-----FORMULA-----
Final calculation for each game:
ScorePlayer1 = sum ( score for each matches)

-----ACTUAL-----
ScorePlayer1=30*.1225*???*1/2

I believe 1v1 is excluded from vRating. All 1v1 are unrated games So no vPoints or dipPoints are ever at stake.

And thank you, Oli, for getting these on our profile and all. It's much more representative of a players skill than available and total dipPoints. One big loss and your points can drop even if it was against the top players on the site.

1v1 do count

It does count, but I think the gV factor reduces the value of a 1v1 game. It does not eliminate it though.

I guess the number of points allowed to be risked in 1v1 is only 1 or something like that.

I think the K factor increases it, but I don't know how much. I have gained up to 7 from a 1v1 win, and lost up to 11 from a survive. Most of my wins are only 2 though. But, that's why I am trying to figure out the math.

Sorry, the points should be v-points.

Love the idea and switch to vpoints. Two questions though:
(1) Are gunboat and regular communication games weighted the same? I could see strong arguments either way on this, and was wondering how this was taken into consideration.

(2) Why aren't games ruined by metagaming/cheating removed from the stats and formula? Our vpoints change based on each actual result vs expected result, yes? Except, one's "expected result" is not accurate if others in the game cheated. To me, any game found to have involved a now-banned player should just be removed as having never happened for both the winners and losers. (Same could go for games drawn after CD'd players left, but that might be too much work.)

Good questions raapers2, but you could add some more:

(3) Why games with high RR requirements and a strict NMR-Policy are weighted the same as games that any unreliable player can join and in which tons of phases get missed due to default or disabled NMR-policy?
In the latter kind of games, the Luck has a very big part in the final result. Well, it's not luck if you are a reliable player who purposely join such shitty games to take advantage on unreliable neighbours. It's premeditated tactic.
Instead, the games in which almost nobody NMRs and, even if so, the phase gets repeated are much harder to win. They should weigh much more.

(4) Why unbalanced maps are kept in the count? If you win 2 Africa against the same 7 people, you should get the same V-Points (in theory, assuming these games ended at the same exact time). But the importance of your win is abslolutely different if you were Morrocco in one game and Mali in the other.
There are dozens unbalanced maps here. Try and play China in a Colonial, or Huron in an Indians, or Nomads in a Haven, or...

(5) Why different Victory conditions are not kept in the count? Why huge maps produce huge results? I still think that it's much harder to solo in a Classic (VC=50%+1) than in a WWIV (~20%). Also, this system overvalues such a victory: say you could get to 50 SCs with Sichuan growing faster by grabbing your 6-7 weak or NMRing neighbors in Asia, while the strong and reliable guys were stuck fighting each other in South America. It's just luck! They could do nothing to stop you, but you got a lot of V-points for having "beaten" them, while you haven't even interacted with them.
All great players at the top of the vHoF, I must say. But the top-top are mainly those who played more WWIVs than others, it can't be a coincidence.

(6) Does to rate games that lasted only a few turns make any sense?
http://www.vdiplomacy.com/hof.php?gameID=6764
Finished: Autumn 1902 - open the whole link to see how it was rated.

(7) Should you have a minimum number of games to be rated? I mean, in the top 100 we have players that play less than 15 games per year. To not mention what games.

(8)...

The issue with a rating system on this site is that it's asked to compare uncomparable things. The games put in relation to each other are extremely uneven and often they're not even about the same sport. Because the difference between a RR99 Migraine in which no turn was missed and an "open to anyone" KW901 is not only the number players. Likewise it's not only the number of players the difference between a IvG and a Classic, or between a Classic and a WWIV. Any the settings.
The Elo keeps in count your current rating to produce an expected result to be used in as many 1v1 matches as your opponents in the game are.
It's like taking the Top 100 tennis players and, basing on such ranking, produce expected results for other sports. Then you make them play 1v1 chess games and update the rankings. Still basing on the expected results of such ranking, some of them play several 4-players Texas Hold'em games and you update the rankings. Then you make a 100-runners marathon at the end of which the winner will be probably the new #1 of world tennis, because he'd have beaten all the other 99 in one time, scoring a disasterful number of points.
(And this marathon would still make more sense than our WWIV games, because the winner actually interacted with all his opponents, who *all* had an actual chance to beat him.)

You can't compare different sports and expect a meaningful result, raapers, any the system you choose.
But it's fun trying and it's fun talking about it.

...Were you agreeing with me, or pointing out that I shouldn't have asked any questions? I was trying to be productive with my questions.

I absolutely agree that it is near-impossible to compare all the uncomparable types of games and don't expect anyone to put in all the effort to try and come up with a formula to fit every last thing in. But you could add another 'x' amount for gunboats or subtract 'y' for CDs.

I'd also suggest using your own 'global benchmark' system to factor in with someone's expected result for a specific country on a specific map.

There are so many problems with the v points it's useless to use them. I would be pretty disappointed if regular points were removed entirely.

Also, can we please try to keep guaroz out of the discussion. He is impossible to understand, has limited ability to apply common sense, is completely and only interested in himself and is generally the most morally detestable person on the site. There are plenty of reasonable moderators let's let them speak on behalf of the site rather than the immature weasel that is guaroz.

Rating systems are silly. Why even bother with them? Who cares who has the most points? While the one in place attempts to account for strength of opponents, it ultimately just leads to head hunting, and people playing the rating system instead of the games.

Oliver, you've gone to a great deal of work to find a rating system that provides some fun for people to compare and discuss, but it's my opinion that these systems always lead to endless controversy and a deterioration of the quality of game play.

With all due respect to the previous two posters, I'd like to cast my vote that I like the rating system and am glad it's implemented. Thank you.