What is the algoritm for determining board rankings? There are a couple of maps I felt I was pretty good on and I was curious to see how I was ranked and was surprised to see players with what appears to be worse records ranked higher. For example, here's the ranking for the map "Random Mazes" (http://www.wargear.net/boards/view/Random+Mazes/Rankings). I've won 5 out of 8 games but am ranked behind Pappy who is 1 for 9 and has a lower H-rating and number eliminated.
If the answer is, "yea, we know it's messed up and we're working on it", that's fine by me. If the answer is it's working fine and makes sense, I'd be interested to understand what is the rationale and algorithm?
I'm surprised to see five people ranker higher who only have one win. Wonder why there isn't a minimum number of games required to be ranked?
Also, why is the total number of players eliminated a significant indicator as opposed to the number of players beaten. Whether or not I eliminate someone is often just a matter of luck (particularly in fogged games) and sometimes something I'm purposely less likely to do if my strategy is to push opponents away from me and into each other as is the case on the "Crystal Caves" map where I'm seeing similarly confusing rankings (http://www.wargear.net/boards/view/Crystal+Caves/Rankings):
Barring that there is problems, peeps with less wins have higher rankings because they beat higher ranked players.
If pappy won that single game - and it was a game with the top 16 ranked players in it, it's totally concievable that his score could be higher than 9 wins against much lower ranked players.
http://www.wargear.net/wiki/doku.php?id=general:help:rankings
But it looks like its losers score\winners score * 20 = amount your score changes for a win
Eliminated is a better indicator because you gain more points for winning the more players there are in a game. Someone might have one an 8 player game and have a better score than someone who won several 2 player games. And as ratsy said, beating higher ranked players on that board helps you more than lower ranking players on that board.
Korrun wrote:Eliminated is a better indicator because you gain more points for winning the more players there are in a game. Someone might have one an 8 player game and have a better score than someone who won several 2 player games. And as ratsy said, beating higher ranked players on that board helps you more than lower ranking players on that board.
I think (?) you're missing the distinction between total opponents and the total that you personally eliminated in each game. You can win a 16 player game and only eliminate the last player and you can win a 2 player game and the number eliminated is the same for each-- 1. I would agree that total number of opponents is more significant than total eliminated, particularly since sometimes it's a valuable tactic to allow someone else to eliminate an opponent.
I don't believe total # eliminated is a "significant" indicator. I never pay any attention to it and I'm pretty sure it has no impact on any stat or standing.
The total eliminated has no bearing on the score you get for a win. Scoring is based solely on your score (if you win) and each loser's score.
Formula is (like Ratsy said):
(Loser's Score / Winner's Score) * 20 = Plus/Minus Score Change
and add that up for each person you beat.
The reason "eliminations" is listed is, I guess, that some people get off on that.
And so, the reason most of us highly ranked people are highly ranked is we won some massively large games. Lotsa points, like 200, can come out of one of those. .
ratsy wrote:Barring that there is problems, peeps with less wins have higher rankings because they beat higher ranked players.
If pappy won that single game - and it was a game with the top 16 ranked players in it, it's totally concievable that his score could be higher than 9 wins against much lower ranked players.
http://www.wargear.net/wiki/doku.php?id=general:help:rankings
But it looks like its losers score\winners score * 20 = amount your score changes for a win
2-Apr-14 | 15:00 | Random maze | Random Mazes | 16 | Loss | 9 | |
22-Feb-14 | 14:38 | 16P Random maze | Random Mazes | 16 | Win | 477 | |
24-Jan-14 | 20:25 | Random mazes | Random Mazes | 16 | Loss | 11 | |
11-Dec-13 | 7:19 | 16X16 | Random Mazes | 16 | Loss | 15 | |
12-Jan-13 | 0:35 | RM16C | Random Mazes | 16 | Loss | 7 | |
28-Nov-12 | 21:02 | Twisty Turny | Random Mazes | 16 | Loss | 9 | |
4-Oct-12 | 17:51 | RM16 | Random Mazes | 16 | Loss | 5 | |
30-Jul-12 | 13:09 | Random mazes | Random Mazes | 16 | Loss | 26 | |
3-May-12 | 15:38 | clear skies | Random Mazes | 16 | Loss | 7 |
Above is from Pappy's Global Ranking history - the last column is how much he won/lost in Global Ranking points from each game. Not the same as his board score, but those would be similar.
Since he lost his first 7 and then won a 16 player he got a huge jump off of that.
----
This is an example of why a lot of heated discussions have been had about determining rankings in different ways. I agree by looking at your record vs. his on this board it does seem you should be ranked higher, but who knows - who you win/lose against is also important -
Long time listener, first time caller here (on the ranking discussion). The current ranking system is outstanding. High global ranking is reflective of sustained excellence across all games....lots of CP is reflective of translating ability across many boards. Even if "ranks" (a vanity in my opinion) are only based on CPs, at a glance in a game I can see if a player is to be feared. If someone has a global ranking of 2000, but few CPs, they are damn well on my to be feared list in a given game. Although I don't think there are very many 2000s out there with very few CPs.
Thanks for all the input. Feel free to continue discussing, but I've gotten what I need. While it is somewhat satisfying to see myself ranked higher on those boards that I think I've done well on, it's not really that big of a deal. In fact, it might have been Ratsy that said in another topic that not drawing attention to yourself is a secret to success and I agree with that and think not being ranked too high on a board can work in your favor. I'm soon to win a 3 player game against two top ranked players and while I think some of that is due to me being a pretty competent player on the particular map in question, I suspect most of it is due to them focusing on each other instead of me because they were relatively much higher ranked than I at the beginning of the game.
You've got a killer moniker. GL!