More on ELO rating:

Before a game starts the kdice server calculates everyones expected score based on thier rating. So suppose you have players ratings as follows:

R1=20
R2=0
R3=100
R4=200
R5=75
R6=100
R7=0

To calculate the expected position of player 1 the following equation is used:

EP1 = 0
for each player rating Rn
EP1 += 1/(1+pow(10,(Rn-R1)/400))

EP1 is now a value between 0 and 7 representing the estimated finishing place.

When the player finishes a new rating is cacluated based on the actual finishing position A.

R1 = R1+(32*(A1-EP1)),0);

Thank you for the information. I had been trying to deduce if number of turns, attacks, dice eliminated, spaces gained etc counted.

This ranking system makes game play very different than the first. Most games, once someone has a chance at getting first, everyone else gives up trying for it and start attacking each other, trying to get second or third.

This is a totally different gaming experience, and not as fun. In one-player, it was never "good enough" to get second, or third. Getting beat last is still getting beat. Think of it as it were actual countries fighting. Getting taken over last is still just as bad as getting taken over first.

I'm at 420 rank right now, top of the list. I got there only because people don't attack me once I'm well of, but instead fight each other. It's easy wins, and not competitive like it should be.

I've thought of other solutions (most involve totally giving up ranking for some other type of rewards), but here's one that might be best: For each specific individual, make 2nd, 3rd, and 4th place all worth the same, their ranking is helped just as much at 4th place as it is at 2nd. This changes endgame strategy (when there are 4 player or less). The only thing that those four players will have to go for is 1st place. Fighting for second will only harm them, they'll have to work on a strategy that gives them first place (this doesn't neccessarily mean having a truce. Players doing badly will still need to get more states. They just would get more selective about the states they took from others, and take states that don't give sabotage the game against that player or themselves)

This system also keeps players motivated to keep from coming in last, by keeping 7th worse than 6th, and 6th worse than 5th. But it should change endgame to where everyone is trying to win, and should make the multiplayer experience more rewarding. (i suggest making 2-4th place all worth either 3rd place or 4th place, because making them all worth 2nd place doesn't give enough difference in value for them to try hard as they may otherwise)

um, I'm not Withnail, I'm Mr. K. I don't know why it thinks I'm withnail. (no one else uses this computer)

In response to Withnail: I have found that most players go after the low-lying fruit and I do so myself when it benefits me. However, I have also found that when a monster is obviously being created, the other players are sometimes willing to hold a temporary truce and ally until the player in first place is knocked down to 2nd. They are usually only willing to do this when it becomes apparent that they will lose if they don't, which sometimes is too late. I like the human element of negotiating strategy during the game. I don't understand how the scoring system affects this (okay, I don't understand the scoring system). Would someone break down that mathematical explanation into non-math terms?

My biggest complaint about the rating system is that it is too harsh for a game that relies greatly on chance. For example I had a rating of 320ish and was playing a game with some lower rated people. I ended up being one of the last people to get a chance to move for the game. After losing a territory or two it finally became my turn, but when I tried to do an almost guaranteed victory move (5 dice vs. 2 dice in this case if I remember correctly) I simply rolled poorly and lost most of my dice and was forced to end my turn. I was eliminated before it became my turn again and lost around 150 points, give or take a few. Also, I've even see people lose before they even had a chance to move and had their points obliterated.

This might become less extreme when enough people are available for the 200+ tables, but as it stands losing out of pure bad luck makes this rating system impractical. In my opinion the ELO rating system might be a good choice for a purely strategic game like chess, but in a game like kdice, which is so dependent on chance, it won't work.

What BoneMan said. I've seen -- and it's happened to me -- where the last person to go is left with one territory and two dice. To lose 150+ points because of a bad draw seems a little harsh to me.

I'm all for rankings and ratings, but it seems a little unintuitive -- is that even a word? -- to use a rating system designed for a purely skill-based game (i.e. same pieces, same setup every time) in a game where SO much luck is involved.

Maybe something simpler, like assigning points based on where one finishes?

I dunno. Maybe I'm drunk. Again.

Yeah, I have to second Boneman as well (third him?). It's not like I've got a better idea, but it seems to me that the chances of an experienced high ranked player getting waxed by a lucky newbie are a _lot_ greater than my chances of walking into a major league chess tournament and taking down the champ.

Let me add something to my previous post:
My suggestion for a better system would be simpler than the current system. Maybe a flat point system in which, depending on where you place, you gain or lose points regardless of your opponents' points.

1st place could get 3 points and every position below below 1st could get 1 less than the one higher than it. e.g. 2nd would get 2 points, 4th would get 0, and 7th place would lose 3 points.

Feel free to play with the numbers all you want, but if you get a set number of points depending on your final position in the game it would be much easier to gain points and to keep them too.

The only problems I can see with this system is that people might give up on first halfway through and jockey for 2nd and 3rd like Withnail (Mr. K) said, and that people would be able to take advantage of newer players just to get higher points.

Anyway I still love the game, and thanks Ryan for making it!

Yeah, this rating system doesn't work.

I lost 100 points from one game, when I was left with a single territory and 3 dice. I attacked once trying to grab a piece and lost, game over.

I think your rating should not count if you're eliminated in round 1 or two.