| Forum Index > SC2 General |
|
Elentos
55566 Posts
On April 06 2017 04:42 ZigguratOfUr wrote: What has Dear done in recent months except lose a lot of PvPs. Should without a doubt be in the bottom 3. He's beaten aLive, GuMiho, ByuN, ByuL and soO in the last month. Outside of PvP, he's actually a threat. Unless he goes full Dear. | ||
Shellshock
United States97276 Posts
On April 06 2017 04:51 Elentos wrote: He's beaten aLive, GuMiho, ByuN, ByuL and soO in the last month. Outside of PvP, he's actually a threat. Unless he goes full Dear. so basically he's not a threat | ||
stuchiu
Fiddler's Green42661 Posts
On April 06 2017 02:13 Lorning wrote: Joke list ByuN should be #1, Maru #2 Could have been but Gerrard threw them out. | ||
|
FrostedMiniWheats
United States30730 Posts
Pretty much... Top 10 power rank stats yo... Low sampling and sub 50% in games and matches xD | ||
|
Elentos
55566 Posts
On April 06 2017 05:04 FrostedMiniWheats wrote: Pretty much... Top 10 power rank stats yo... Low sampling and sub 50% in games and matches xD They wouldn't need to make an article if they were just gonna throw the players at you in order of their Aligulac winrates for 2017. Anyone can look that up for themselves. For the record, that would be: 1 - Inno 2 - ByuN 3 - Maru 4 - herO 5 - Solar 6 - GuMiho 7 - TY 8 - Stats 9 - aLive 10 - Dark 11 - ByuL 12 - Hush 13 - Bunny 14 - soO 15 - Ryung 16 - Dear I'm sure you'd find nearly everything about that list wrong. Point being, PRs are opinion pieces in the end so they don't have to align with statistics. And in this case the writer seems to be of the opinion that beating 5 of the players ranked ahead of Dear in the last month is enough to put Dear at 10 despite all his PvP losses. Although I wish the PRs were longer so there'd be more text invested into justifying every position. The texts range from 5 (Maru) to 22 (Stats) lines and the lower end seems really low. | ||
munch
Mute City2363 Posts
On April 06 2017 05:16 Elentos wrote: Although I wish the PRs were longer so there'd be more text invested into justifying every position. The texts range from 5 (Maru) to 22 (Stats) lines and the lower end seems really low. Not enough time, maybe for the next set after exams | ||
|
pvsnp
7676 Posts
Not enough time, maybe for the next set after exams Empathy. IRL always takes priority. As an idea, I remember when stuchiu wrote his "Greatest Players" list he put a short explanation at the end of each player bio on why Player X was ahead of Player X+1. I.e. "Stats is ahead of Inno because he won GSL Season 1, a stronger result than anything Inno has done recently." | ||
|
Hider
Denmark9433 Posts
On April 06 2017 04:22 The_Red_Viper wrote: These days power ranks are fairly hard to make tbh, what do you consider? Like there are online tournaments every week but the participation is fairly inconsistent in general. There is no weekly proleague anymore, etc I don't think the list is too bad, at a certain rank it just is fairly hard to be sure about anything. I think the best way to make a PR would be to have a point system you simply have to apply for the recent tournaments though, that way it's at leat 100% consistent Nah, use Aligulac as baseline. And whenever you want to deviate from their rankings, you make sure you have very convincing arguments. The problem is whenever you don't actually look at all of their results but just cherry pick. That's a consistent problem with most of TLs power ranks. | ||
|
The_Red_Viper
19533 Posts
At the end of the day this is just a nice little list done by a writer who had the passion to do it in the first place and not something which is used to place odds, etc. | ||
|
starkiller123
United States4030 Posts
| ||
|
FrostedMiniWheats
United States30730 Posts
On April 06 2017 05:16 Elentos wrote: They wouldn't need to make an article if they were just gonna throw the players at you in order of their Aligulac winrates for 2017. Anyone can look that up for themselves. It isn't just in order of the winrates on Aligulac. I just used that as a starting point as I feel those poor stats alone could keep Dear from being ranked as highly as he is. Going more in depth, there's also: - Sample size (pretty low compared to others on the list) - Level of competition (Decent enough) - Offline results should carry more weight (of which Dear has played very very few of this year) - Results with respect to balance (Dear has done fairly good in PvT) - Consistency (of which he hasn't demonstrated this year) - Peak performances (nothing extraordinary so far this year..) all in all, I can't see Dear beating anyone except Hush and Bunny here in ranking. So he beat a few people ranked ahead of him. You can say the same of Solar and ByuL who also blow him out of the water on pretty much every other front. On April 06 2017 07:39 starkiller123 wrote: Even when I really disagree with these articles I still really enjoy reading them and that's the take-home message. Thanks Munch :3 | ||
|
usopsama
6502 Posts
| ||
|
munch
Mute City2363 Posts
On April 06 2017 07:04 Hider wrote: Nah, use Aligulac as baseline. And whenever you want to deviate from their rankings, you make sure you have very convincing arguments. The problem is whenever you don't actually look at all of their results but just cherry pick. That's a consistent problem with most of TLs power ranks. Statistics are meaningless without context. TY went 20-14 at IEM Katowice @ 59% map winrate INnoVation went 13-7 @ 65% Stats went 25-12 @ 68% GuMiho went 17-9 @ 65% Ryung went 14-7 @ 67% aLive went 15-7 @ 68% Who had the best tournament? You can say "Filter by KR only"; but then are you saying that Hush is better than ShoWTimE? That Impact is better than Nerchio? That some random online game played at 3AM KST is equally important as a GSL match? All arguments are biased and cherry picked; just realise that and move on | ||
|
ClaudeSc2
United States73 Posts
| ||
|
StorrZerg
United States13919 Posts
| ||
|
Hider
Denmark9433 Posts
On April 06 2017 07:32 The_Red_Viper wrote: At that point they could link aligulac though, no? I am fine with having the PR be an opinion piece which is obviously based on stats but can weigh them somewhat subjectively (like putting more emphasis on recent offline events than online ones, etc) At the end of the day this is just a nice little list done by a writer who had the passion to do it in the first place and not something which is used to place odds, etc. Yes, but you still need to be aware of the data. If you only are aware of like 20% of all games played by the players, you end up with these flawed ratings. What you should do is to understand what the "numbers say", and if you think the numbers are misleading, present your case. However, if you only look at 20% of the data, Aligulac is always gonna outperform. So to sum up, use Aligulac (or another "data-based method") and then add subjective analysis on top of that. Statistics are meaningless without context. TY went 20-14 at IEM Katowice @ 59% map winrate INnoVation went 13-7 @ 65% Stats went 25-12 @ 68% GuMiho went 17-9 @ 65% Ryung went 14-7 @ 67% aLive went 15-7 @ 68% Who had the best tournament? What's your point? Aligulac takes into account the relevant context, as in the skill level of opponent. And if you think Aligulac is missing important factors, then present your case. Also, statistical analysis is always just a matter of how good it is. If it's good enough it always beat subjective analysis (not saying Aligulac does that though). You can say "Filter by KR only"; but then are you saying that Hush is better than ShoWTimE? That Impact is better than Nerchio? That some random online game played at 3AM KST is equally important as a GSL match? That's what a data-based analysis takes into account. You look at all relevant data as in, skill level of opponent and importance of match. It's quite naive if you just think numbers = win/rate. | ||
|
Shellshock
United States97276 Posts
| ||
|
Pandain
United States12989 Posts
| ||
|
ZigguratOfUr
Iraq16955 Posts
On April 06 2017 11:44 Pandain wrote: If soO wins this does that count as a GSL victory? No. But if he loses this in the finals I'm willing to consider it a GSL second place. | ||
|
Pandain
United States12989 Posts
On April 06 2017 12:44 ZigguratOfUr wrote: No. But if he loses this in the finals I'm willing to consider it a GSL second place. Seems fair | ||
| ||
|
|
PiG Sty Festival
Maru vs ShoWTimE
Clem vs Serral
PiGStarcraft803
3DClanTV
 24Sparkling Tuna Cup
uThermal 2v2 Circuit
Replay Cast
Wardi Open
Monday Night Weeklies
Replay Cast
Replay Cast
Replay Cast
The PondCast
[ Show More ] |
|
|
EPT
