EPTSC:R might do something similar.
BW Remastered Profiles - Page 11
| Forum Index > BW General |
|
eviltomahawk
United States11135 Posts
SC:R might do something similar. | ||
|
L_Master
United States8017 Posts
On June 02 2017 11:35 eviltomahawk wrote: Here's a breakdown of SC2's league distribution goals. SC:R might do something similar. Bottom 4%? Bronze is literally like the 8 yr olds division then. Either way, based on that LOTV target it looks like you would have: Little Kids + People on a touch-pad = Bronze CPU Rank/D-/Low D = Silver D/D+ = Gold C-/C = Platinum C/C+ = Diamond B-/B/B+/maybe A- = Master A/A+/Oly = Grandmaster This going by percentage breakdown's from what I remember from ICCup back in the days of high activity with a decent number of koreans playing. If Korea was seperate and/or there was an NA server that might move up ever so slightly. Was wishing I could go check old ladder archives to see the breakdown for sure, but those don't seem to exist  | ||
|
eviltomahawk
United States11135 Posts
http://www.teamliquid.net/forum/brood-war/99016-iccup-rank-median | ||
|
oOOoOphidian
United States1402 Posts
On June 02 2017 12:52 eviltomahawk wrote: I remember an iccup ranking breakdown from this thread http://www.teamliquid.net/forum/brood-war/99016-iccup-rank-median Based on that and on current actual SC2 distributions (source http://sc2unmasked.com/Stats ): A to A+ = Grandmaster (0.5%) top of B- to A- = Master (5.2%) top of C- to B- = Diamond (24.6%) C- = Platinum (21.5%) Top half of D+ = Gold (20.3%) Bottom half of D+ = Silver (20.9%) D- to D = Bronze/Unranked (7%) | ||
|
L_Master
United States8017 Posts
On June 02 2017 13:22 oOOoOphidian wrote: Based on that and on current actual SC2 distributions (source http://sc2unmasked.com/Stats ): A to A+ = Grandmaster (0.5%) top of B- to A- = Master (5.2%) top of C- to B- = Diamond (24.6%) C- = Platinum (21.5%) Top half of D+ = Gold (20.3%) Bottom half of D+ = Silver (20.9%) D- to D = Bronze/Unranked (7%) That I think is using the initial assumptions, which is basically looking only at D+ and above, but that's already a pretty selected group. D rank players certainly make up a sizeable portion of the playerbase. From what he looked at about 50% of the D ranks had a 1/3 winning rate, suggestive of true D. About 50% did not. If we assume (and I suspect this would over-inflate skill) that of those 50% with a winning record they were distributed similar to the ladder; then you would add about 20k to D, and distrubte the rest to other groups. So of the ignored 40k: 22,008 to D 11,004 to D+ 5,502 to C- 2,751 to C 1,376 to C+ 688 to B- 344 to B 178 to B+ approx 200 to A- and up Which leaves: 22,008 D-/D 20,485 to D+ 10,438 to C- 4,849 to C 2,580 to C+ 1,562 to B- 652 to B 486 to B+ 336 A- or better These yield an estimated breakdown of 34.7% D-/D 32.3% D+ 16.5% C- 7.7% C 4.1% C+ 2.5% B- 1% B .7% B+ .5% A- or better From a top down perspective: A- is top .5% B+ is top 1.3% B is top 2.3% B- is top 4.8% C+ is top 8.9% C is top 16% C- is top 32% D+ is top 65% D is top unknown percent (amount D vs D- is unknown) Then from a percentile breakdown: A- is top 99th percentile B+ is 98th percentile B is 97th percentile B- is 95th percentile C+ is 91st percentile C is 83rd percentile C- is 67th percentile D+ is 34th percentile D-/D spawn 1st to 33rd percentiles | ||
|
CrymeaTerran
149 Posts
| ||
|
DSK
England1110 Posts
| ||
|
ShuriKn
Canada20 Posts
| ||
|
404AlphaSquad
839 Posts
300-100 = 83% winrate its like they arent able to display accurate stats in any of their games ^^ (SC2, OW and now SC:R) | ||
|
konadora
Singapore66357 Posts
| ||
|
Qikz
United Kingdom12023 Posts
On June 02 2017 21:58 404AlphaSquad wrote: oh god. It is blizzard math time again 300-100 = 83% winrate its like they arent able to display accurate stats in any of their games ^^ (SC2, OW and now SC:R) Ranks are never accurate imo. The only ranking system that comes close near to being accurate I'd argue is ELO from Chess. MMR is literally just a number and goes up when you win and down if you lose, it's entirely pointless because depending on who you play (see people B rank and above on iCCUP by just grinding games vs D players) then it's never going to be accurate. | ||
|
Drake
Germany6146 Posts
On May 27 2017 15:19 TT1 wrote: wait, you gave feedback.. did i read that right? feedback on what, exactly? you being paranoid about the dev team adding SC2 gameplay features like MBS and smartcast (by giving an inch!) has nothing to do with giving feedback on the new profiles, it's you being paranoid. also by gameplay changes i meant changing unit stats/the game's mechanics (they stated that they didn't add rebinds for f keys and control groups because they didn't want to touch the games mechanics). keep wearing that tin foil hat of yours tho, wouldn't want to ruin your fun. im hoping people like you will be able to relax once the launch happens and nothing ends up changing, god knows you need it. edit: and just to be clear i don't really have an issue with you wanting F - S ranks. what i have an issue with is your "give an inch, take a mile" statement, that's a terrible mindset to have. 120% agree, some people just cant stop crying | ||
|
ProMeTheus112
France2027 Posts
On June 22 2017 15:03 Qikz wrote: Ranks are never accurate imo. The only ranking system that comes close near to being accurate I'd argue is ELO from Chess. MMR is literally just a number and goes up when you win and down if you lose, it's entirely pointless because depending on who you play (see people B rank and above on iCCUP by just grinding games vs D players) then it's never going to be accurate. Exactly and I agree ELO from Chess seems good something similar or identical was used on old blizz ladder which seemed disfunctional for other reasons like hacks/bots and slow speed. | ||
|
thezanursic
5497 Posts
| ||
|
Chris_Havoc
United States600 Posts
On June 23 2017 02:39 thezanursic wrote: Is there any news on when SC:R gets released? Nope. No date. No date announcement. Nothing. Blizzard themselves may not know the date yet with all the issues coming from 1.18 development. | ||
|
CecilSunkure
United States2829 Posts
| ||
|
aLt)nirvana
Singapore846 Posts
Also is there a way to view bnet profiles via your browser? I am using starlog.gg now but i am sure they are getting their data from somewhere... | ||
|
Pippah
Denmark356 Posts
| ||
|
NeVeR
1352 Posts
| ||
|
sM.Zik
Canada2550 Posts
 | ||
| ||
