Is this a consequence of some Ladder Matching Rules Battle.net is using?
Why is Dayvie at the Top of the Ladder?
Blogs > ItsYoungLee |
ItsYoungLee
Korea (South)227 Posts
Is this a consequence of some Ladder Matching Rules Battle.net is using? | ||
seRapH
United States9706 Posts
| ||
blahman3344
United States2015 Posts
| ||
mesohawny
Canada193 Posts
"quality" being an arbitrary term in relation to their opponents ladder ranking. | ||
Chill
Calgary25951 Posts
On August 14 2010 01:40 ItsYoungLee wrote: Can someone explain to me how someone with what appears to be 91 more wins and only 31 more losses is rated lower? Can someone explain to me how you couldn't figure this out yourself? | ||
spinesheath
Germany8679 Posts
My guess is that it's a consequence of the bonus point system, high losses, low gains at the top of the ladder. Dayvie climbs in points because the bonus point system gives him more than he loses. TTOne loses more than/just as much as he gains from the bonus pool. It's fairly safe to assume that you are bound to lose points if you are at the top of the ladder and "only" have a 68% win ratio. | ||
EffectS
Belgium795 Posts
On August 14 2010 01:47 Chill wrote: Can someone explain to me how you couldn't figure this out yourself? For that Answer my friend, I'd like to advice you to Check out the following site: + Show Spoiler + 50 Things that Kill braincells | ||
ItsYoungLee
Korea (South)227 Posts
On August 14 2010 01:47 Chill wrote: Can someone explain to me how you couldn't figure this out yourself? It's not that trivial as an explanation of "Dayvie just played more favored people while TTOne didn't". If the system is good it should average out and you shouldn't have people with different win percentage of 4%+ with a combined sample size of over 400 games have similar rating. In this case it's even worse because the direction is reversed (lower win % --> higher score). The p-value for that has got to be miniscule. The explanation must lie in how the system matches opponents. I don't have that great of a knowledge about how it works, and would like to understand. | ||
Kralic
Canada2628 Posts
| ||
n.DieJokes
United States3443 Posts
On August 14 2010 01:55 EffectS wrote: For that Answer my friend, I'd like to advice you to Check out the following site: + Show Spoiler + 50 Things that Kill braincells Can someone please explain to me why EffectS tlpdized his post? | ||
ItsYoungLee
Korea (South)227 Posts
+Blizzard is reluctant to give all details of their ladder ranking system because then it's possible it could be abused. Just thought with the intelligent TL userbase here someone could venture some plausible guesses to explain such non-intuitive results. | ||
Biff The Understudy
France7749 Posts
On August 14 2010 01:55 EffectS wrote: For that Answer my friend, I'd like to advice you to Check out the following site: + Show Spoiler + 50 Things that Kill braincells Priceless | ||
Kralic
Canada2628 Posts
Maybe TTone lost a lot of games to Dayvie? If you beat someone you gain a lot more points then what you would have won. Say Dayvie beats TT1 Dayvie gains 20 points and TT1 loses 10 points. Dayvie just gained 30 points on TT1 in the ladder. | ||
AnodyneSea
Jamaica757 Posts
| ||
ItsYoungLee
Korea (South)227 Posts
As for the bonus pool, I don't think it should make much of a difference because they probably started laddering at the same time, at the release of Starcraft 2. Thus they both gained the same amount from the bonus. I found this website which helps to explain the system a lot, http://lorehound.com/starcraft/starcraft-2s-battle-net-leagues-ladders-and-rankings-explained/ but it still seems to me that the data I see for rankings seems discordant with a well-designed laddering system. Another explanation that seems possibly plausible to me is that these two users play as part of two different Battle.net 'cohorts'. If TT1 only plays at night and Dayvie only plays during the day, this could somewhat explain how they could have different %ages with similar ratings. Do they live on opposite sides of the Earth? lol | ||
Chill
Calgary25951 Posts
| ||
broz0rs
United States2294 Posts
http://www.sirlin.net/blog/2010/7/24/analyzing-starcraft-2s-ranking-system.html | ||
EffectS
Belgium795 Posts
On August 14 2010 01:59 n.DieJokes wrote: Can someone please explain to me why EffectS tlpdized his post? + Show Spoiler + I clicked it, and it is my geek nature not to more than is needed, so I diden't remove it. happy now? | ||
Navane
Netherlands2725 Posts
If best winrate -> best ranking, then ranking would be rather redundant. | ||
404.Delirium
United States1190 Posts
I really just responded half-heartedly to the thread because I got a mild kick out of the ad banner. + Show Spoiler + | ||
| ||