Bryan, how about playing today at 21 gmt?

Quote from: Gabriel on May 15, 2012, 11:22 PM
hmm
Yeah, Im up everyday from 19 gmt
Can you do it any saturday? Id prefer a saturday... i usually have tests on monday

Hmm yes, but if it has to be Saturday 26th, then I'll be available since about 21 gmt.

lol, never letting shy play hysteria in my host again
statik played btw, not me Xd

Sunday 27th at 18:00 gmt, gab?

thriller

barman parachuterace battlerace bungeerace / dennis bigrr

Forgot about these?

HHC, today 20 GMT maybe? If not, then how about tomorrow 18 GMT?

Maybe eS has only one member who is pulling all the strings? Some kind of super villain of worms?

undoubtedly the greatest game ever played.

Quote from: Impossible on May 09, 2012, 01:47 PM
XDDDddddd NepTuNe joining mm? ekwjikhekehkeh

Yes, he ended up getting positive votes from everyone.

I'm fine with Sunday 20th, but have in mind that 16 GMT might be a little too early for the American part of mm even if the date fits them.

Deadline set. Reminder sent.

Quote from: Dub-c on May 07, 2012, 02:57 PM
I don't see how rating change holds any significance of luck factor. Could you explain this further.

When you face someone who's better that you (thus higher rated), it's natural that you will need some luck in order to win (by luck I mean any unintentional actions that bring you closer to winning, my definition is not limited to random crate drops). On the contrary, you will probably need some bad luck in order to lose against a weak (lower rated) opponent. Therefore we can see a simple relationship between the potential number of points you can win/lose and the amount of (bad) luck deciding about the outcome of the match.
I think that such relationship is less evident in the win/lose ratio statistics, because they provide no distinction between winning against strong and weak opponent. Take MonkeyIsland as an example, he's a top wormer without doubt but his overall winning percentage is rather low (57.78%), because the considerable number of his opponents are other very strong players. On the other side, someone who prefers playing against weak opponents will probably have a good winning ratio coupled with a lower rating.

The winning percentages: go to any overall scheme standings page (like this), sum up the number of won games from the top 10 players, divide by the sum of all games played.
The rating gain averages: go to the recent games from any scheme (like this), sum up the overall rating changes from the first 100 games excluding the playoff ones, divide by 100.

Quote from: Dub-c on May 07, 2012, 06:43 AM
Ok so roper is the luckiest scheme there is eh? Lets look at the numbers.

TTRR is widely considered the least luckiest scheme. So the top 10 overall winning percentage should be much higher then in roper right? [...]
The average winning percentage of TTRR is 78.603%

Roper should obviously be much much lower since its so lucky and you will lose so often because of luck. Lets look at the numbers. [...]
Ropers overall winning percentage is 75.205%. That's only 3 percent lower then ttrr? How is this possible for such a lucky scheme? How could people have such a high winning percentage when its so lucky? WOW?
      

Note: you have simply calculated the average of winning percentages without taking into account the number of games played by every player.
The weighted averages are 77.47% for TTRR and 75.21% for roper.

Let's look at the weighted averages from the other classic league schemes:
BnG: 84.68%
Elite: 72.94%
Hysteria: 73.82%
Shopper: 74.09%
Team17: 76.4%
WxW: 77.22%

These numbers make me think that comparing the top 10 players' winning percentages is not a good idea to measure how does luck affect the scheme. Players who dominate their favorite schemes are a bad test sample.

Instead, I think it's better to look at the average of the overall points gained by winners of a set of random games from every scheme. A higher average would suggest a more lucky scheme, because it would show that low-rated players occur to beat high-rated players more often.

Here are the results for 100 most recent singles non-playoff games in every scheme. (A total of 800 games analyzed)
BnG: 37.17
Elite: 30.58
Hysteria: 32.77
Roper: 32.95
Shopper: 30.9
Team17: 31.21
TTRR: 28.92
WxW: 29.03

BnG is a large peculiarity here, I think that the unusually high average rating gain is caused by the fact that it's the least played scheme and it's not an uncommon situation when a player's rating doesn't really reflect his/her skill level.
If my methodology is correct, then aside BnG, roper appears to be the most luck-based scheme, followed by hysteria and T17.

Edit: forgot to include WxW