Death Taco was a 1600 rating player who managed to beat HHH and not lose to a group of 1300-1500 players. He lost three matches in a row to the only other good players (good at that time) that he faced. If you want to say that he was a good player because he reached rank 9, then go ahead. I am more interested in who is the best player, not who can game the system the best. Your technicality argument reminds me somewhat of the people who defend a win after booting their opponent on territory picks. Yes, they did win the game, but it didn't really show what anyone was interested in seeing.

Was he a good player because he reached rank 9? Of course not. Don't you see that I am in fact arguing the opposite? I am arguing that highest achieved rating may actually be a better indicator of skill than the highest achieved rank. Your argument does nothing to respond to this.

Highest ranking is better than highest rank obviously, because the latter heavily depends on who is and who is not on the ladder. In Impaller's times ranking that'd give #1 easily would be 4-5 at the moment and it's still with the absence of many players.

I see your point now Huruey, sorry. Well, Death Taco also reached 1975 rating, whereas he would have had a 1600 rating without his game against HHH.

My bigger point was that gaming the system is not really a sign of good playing, and should not be considered as an important aspect. Either rating or rank, taken alone, represent potential gaming of the system numbers. Ultimately what I was against was caring about who hit the highest rating number. I can understand the desire to do so, since it's not really capped the way rank is. That said, the highest ratings will almost always be to the people who stall while losing. To me, caring more about rating than rank is a fools errand. That's the same reason that I would like to see turn 0 boots taken out of the 1v1 comparisons, since (even though they reflect on rating and ranking) those games don't show what we are interested in.

Well, Death Taco also reached 1975 rating, whereas he would have had a 1600 rating without his game against HHH.

I'm pretty sure that's not possible. That's an increase of 375 points, much more than half way between Taco and HHH's scores. Where did you get this figure? Have you tried using the ratings calculator?

My bigger point was that gaming the system is not really a sign of good playing, and should not be considered as an important aspect.

Please note that at no point have I tried to argue that either rank or rating are perfect indicators of ability. Some of the best players on warlight aren't even members. I was under the impression, though, that people wanted to make some kind of meaningful comparison of ability on the ladder.

That said, the highest ratings will almost always be to the people who stall while losing.

That is not so. Stalling has a bigger impact on rank than it does on rating, especially if you're talking about within the top 10. If you are not playing your games, your rank will fluctuate more on its own than your rating.

To me, caring more about rating than rank is a fools errand.

Why should caring about getting a higher rating be any more foolish than caring about getting a higher rank? When I played, my intention was to play warlight, not to play some strategic ladder waiting game.

That's the same reason that I would like to see turn 0 boots taken out of the 1v1 comparisons, since (even though they reflect on rating and ranking) those games don't show what we are interested in.

I'm not even sure where you are trying to fit this analogy in. It seems completely irrelevant to this argument. :P

I think you and I are discussing two seperate things at this point, so I'm just going to drop my side of it.

I was not arguing for rank being a descriptor any more than rating, though I can see how you might think I was arguing for that.

an equation could be made to rank players. an example rating system:

highest rating ever + (net wins vs #1s X 10) + (net wins vs #2s X 5) + total # of days as #1 + overall ladder winning % as a value XX.XX

games, wins and losses, decided by boots where winner is undecided don't count.

for example, my score (i know my boot games, wins and losses -- vs chris, vs teddy as denzy, vs zaeban as self and as denzy, vs chas, vs ixxxx) so it's quicker to use as an example:

2236 + (12-10 as gui 4-0 as denzy = 16-10 = 6... X 10 = 60) + (9-5 = 4... X 5 = 20) + 63 + (77-29 as self, 10-2 as denzy = 87-31 = 73.73% = 73.73) = Total Score = 2452.73

Huruey, is there a ratings calculator? I thought the way our ratings were calculated was a secret.

The ratings system is detailed on the wiki. You can download software to simulate the ratings system, and you can download the ladder history, see what would have happened if some game ended differently, and also simulate possible future scenarios to see how the ratings might be affected.

http://wiki.warlight.net/index.php/Ladder_Ranks_and_Ratings

What WL needs is a true rating and a true highest rating ever achieved.

To get your true rating you need to have 50+ active finished games.

There are a lot of problems with the current system. It has a bias towards playing higher ranks so playing fewer games will artificially increase your rating. Also 10 games isnt a large enough sample size, someone could have a 2400 rating over 10 games but be no where near a 2400 rated player.

These ladder settings would be ideal
180 day expiry instead of 90
20 games to be ranked and get a rating
50 games to get your true rating.