|
On December 07 2010 03:43 Shadowed wrote:
That would kind of ruin the point of having a comparable system if you still sectioned it thought, but I don't feel like getting PTR right now to check.
I may be wrong. I am also as lazy as you re: PTR. It wouldn't totally ruin the point because they could provide a consolidated list on their site (and you could on yours  ), especially since the list may be many hundreds of players that would be unwieldy in-game. Also I imagine it's easier to implement.
|
Actually nobody being in the masters league yet is a good point :<
|
By the way, here are the updated data (in bold) for EU server :
S-rank (+0) :
Agria Alpha
Akilae Lima
Aldaris Foxtrot
Feld Delta
Forge Eta
Hellion Echo
Mar Sara Sigma
Phoenix Mars
Ramsey X-Ray
Rimes Echo
Roach Lambda
Scimitar Psi
Shuttle November
Siege Tank Gamma
Stukov Foxtrot
Tal’darim Theta
Uraj Eta
Zealot Alamo
A-rank (+63) :
Khala Tau
Muadun Tau
Norad Phi
B-rank (+126):
Chau Sara Quest
Gorn Mars
Reaper Gravity
Routhe Victor
C-rank (+189):
Boros Tango
Brood Lord Mars(*)
Executor Alamo
Hauler Foxtrot(*)
Scourge Charlie
Shiloh Oscar(*)
Vermillion Nu
D-rank (+252):
Alzadar Sigma
Char xi
Gantrithor Bravo
Jorgensen Zeta
E-rank (+315):
Archon Beta
Broodling Mu
Corsaire Pepper
Duke Sigma
Khaydarin X-Ray
Liberty X-Ray
Marine psi
Meinhoff Pi
Sakai Lima
F-rank
Queen Iota
|
If I compile the data, will you do the analysis for SEA?
|
whoa egstrifecro gets a division named after himself... im jealous
|
On December 07 2010 03:17 Excalibur_Z wrote:Here's this week's Top 200 breakdown: Show nested quote +
Blizz Name Record Rating Division
1 TTOne 654-375 3083 Medic Mu
2 ROOTslush 683-409 3014 Medivac Alamo
3 EGStrifeCro 534-348 3007 EGStrifeCro
EGStrifeCro has his own division? He's even more of a badass than I thought!
(Akilae Omicron as of 29/11. Thanks for all your hard work Excalibur, you are ace.)
EDIT: hee hee, you win this round, Stoned!
|
I think it's neat how perfectly the math worked out. It's never easy to reverse-engineer these things.
Thanks to you, I've finally found why I keep getting stuck with good players no matter if I place diamond the hour after ladder reset or the week after. Climbing to #1 sucks when your division has twenty players in the top 200.
|
maybe if someone does this for SEA, Tawain and Russia, we can have a world-wide top 200? =)
|
There is little point in comparing people that don't play each other on the ladder.
|
United States12236 Posts
On December 08 2010 20:37 Mr Mauve wrote:Show nested quote +On December 07 2010 03:17 Excalibur_Z wrote:Here's this week's Top 200 breakdown:
Blizz Name Record Rating Division
1 TTOne 654-375 3083 Medic Mu
2 ROOTslush 683-409 3014 Medivac Alamo
3 EGStrifeCro 534-348 3007 EGStrifeCro
EGStrifeCro has his own division? He's even more of a badass than I thought! (Akilae Omicron as of 29/11. Thanks for all your hard work Excalibur, you are ace.) EDIT: hee hee, you win this round, Stoned!
Ugh, dammit. Good catch you guys. Fixed! =)
Also it took a very long time to parse Korea, so I'm quite hesitant to do SEA. A major issue is the time difference which makes things more difficult. All of the region data is collected at the same time, but it is posted by regional community teams at different times. In the case of Korea, the kr.battle.net blog post was posted about 14 hours after the snapshot, which is plenty of time for massgamers to play more than 25 games (which is how long the match history shows on the website). If they play more than 25 games then it's impossible to ascertain their rating at the time of the snapshot. It also meant that more players had records that did not match their snapshot records, which meant more working backwards along their match history. SEA will be no different, but if you'd like to cover that region Shadowed and I would greatly appreciate it.
|
On December 08 2010 21:23 hunter3 wrote:
I think it's neat how perfectly the math worked out. It's never easy to reverse-engineer these things.
Thanks to you, I've finally found why I keep getting stuck with good players no matter if I place diamond the hour after ladder reset or the week after. Climbing to #1 sucks when your division has twenty players in the top 200.
Which makes me think that this top 8 in every division does the blizzard tourny should be rethought, doesnt really make sense when rank 100 of one division can be rank 1 in another but top 8 of every division participate? Maybe im missing somehting
|
United States12236 Posts
On December 09 2010 01:33 CrayonKing wrote:Show nested quote +On December 08 2010 21:23 hunter3 wrote:
I think it's neat how perfectly the math worked out. It's never easy to reverse-engineer these things.
Thanks to you, I've finally found why I keep getting stuck with good players no matter if I place diamond the hour after ladder reset or the week after. Climbing to #1 sucks when your division has twenty players in the top 200. Which makes me think that this top 8 in every division does the blizzard tourny should be rethought, doesnt really make sense when rank 100 of one division can be rank 1 in another but top 8 of every division participate? Maybe im missing somehting
That top 8 in every division thing hasn't been true since early beta. The divisions used to look like this:
![[image loading]](http://i9.photobucket.com/albums/a96/Weroe/SC2%20PvT%20Strat/top8.jpg)
|
So do division tiers affect how many ladder points you carry over upon promotion?
I'd imagine you wouldn't carry over as many points if you move into a stronger division
|
On December 13 2010 09:44 RME wrote:
So do division tiers affect how many ladder points you carry over upon promotion?
I'd imagine you wouldn't carry over as many points if you move into a stronger division
If points carried over after promotion then yes division tiers would affect that. That's a big if, however. We're still not certain how your points post-promotion are determined exactly.
|
I will attempt to answer the unanswered questions
- How does the system determine which division tier a player should be placed? Obviously by MMR, but to what extent?
As you may remember in beta, the division name was natural numbers. My guess is that, at the beginning, blizzard creates divison 1, 2, 3, ...n. And place players randomly in them. Once those are full, blizzard creates division n+1, .... m, etc. This may explain the "adjustments" since if you are in divison 1 through n, you started playing rly early and u have 0 bonus points, if you are in division n+1 through m, you probably started playing a few weeks later than those mentioned previously, so u get +63 points etc....
- How many tiers of divisions exist within each league? Blizzard had mentioned that additional divisions would be added to Bronze to solve the Bronze Zero phenomenon, but have divisions (or, extra layers of granularity) been added to other leagues previously?
As in the previous question, tier is probably determined by time of creation of the division. More points are added to later-created divisions since those players have a late start competing against others
- The tiers defined in my post could only be determined due to the Top 200 listing, but how would an average player manage to determine his division's tier without a global reference point like the Top 200? Furthermore, if the Top 200 goes away once the Grandmaster league is added, will we be able to determine tiers in any capacity?
I would suggest the approach to attempt to determine the creation time of each division and try to find a precise relationship between creation date and bonus points
|
United States12236 Posts
That's not right, though. Division creation time has nothing to do with the tier of the division. Look at Ret for example. Phash Kappa is a relatively new division, yet it has an offset of 0. Scout Kilo was active from Day 1 and yet it has an offset of 315.
|
This week LA TOP200 is interesting.
Does anyone know the guy named HellForce? There seems to be 2 of them and both of them are in a S ranked division.
One should be around 31 on top 200 (as it stats in sc2ranks), but isn't. The other is ranked 160 on both sc2ranks and top200.
Does anyone know what happens with banned people? Do their profile stays? I would bet he was banned and that's why his older profile isn't on top 200 (nor has any play since nov 30th) for 2 weeks straight. But he probably bought a new account and has once again climbed there.
If anyone knows this guys so we could comprove the theory, this makes sense that banned players aren't on top 200, even though their profile stills exists.
So we have 1 less mystery in our hands :D
http://www.sc2ranks.com/masters/la
http://us.battle.net/sc2/pt/blog/1670515#blog
http://us.battle.net/sc2/pt/profile/204057/2/HellForce/
http://us.battle.net/sc2/pt/profile/6208/2/HellForce/
|
New NA top 200 is out, if you haven't noticed yet.
|
On December 14 2010 08:40 SDream wrote:This week LA TOP200 is interesting. Does anyone know the guy named HellForce? There seems to be 2 of them and both og them are in a S ranked division. One should be around 31 on top 200 (as it stats in sc2ranks), but isn't. The other is ranked 160 on both sc2ranks and top200. Does anyone know what happens with banned people? Do their profile stays? I would bet he was banned and that's why his older profile isn't on top 200 (nor has any play since nov 30th) for 2 weeks straight. But he probably bought a new account and has once again climbed there. If anyone knows this guys so we could comprove the theory, this makes sense that banned players aren't on top 200, even though their profile stills exists. So we have 1 less mystery in our hands :D http://www.sc2ranks.com/masters/lahttp://us.battle.net/sc2/pt/blog/1670515#bloghttp://us.battle.net/sc2/pt/profile/204057/2/HellForce/http://us.battle.net/sc2/pt/profile/6208/2/HellForce/
SDream... as you said, some of us are missing from the official ranks. I'm rankead 163 in SC2ranks master league but I didn't get to the top 200 on battle.net. Also, another user from my division is ranked 287 in SC2ranks and 199 in the official rank.
Edit: The people ranked close to me:
160 sTryDB 1,880 (2,006) 200 162 55.25% Argo Echo (#9) 3 hours (blizzard: 153)
161 Conquerz 1,878 (2,193) 267 216 55.28% Turaxis Omega (#6) 7 hours (blizzard: not ranked)
162 SNascimento 1,877 (2,192) 272 241 53.02% Preserver Dixie (#7) 8 hours (blizzard: 154)
163 Sembei 1,876 (2,191) 212 181 53.94% Raven Chi (#6) 7 hours (blizzard: not ranked)
164 Horseman 1,876 (2,002) 276 241 53.38% Argo Echo (#10) 3 hours (blizzrd: not ranked)
165 Strider 1,872 (2,061) 153 114 57.30% Void Ray Tau (#16) 3 hours (blizzard: 155)
I don't have a lot of bonus pool (played last game saturday and had 0 there), maybe they are getting better position than us because they are counting theirs bonus points (CARRAS, the user ranked 287 in my league has more than 200 points reserverd, and maybe because of that he gets to the top 200 and most of us ranked better than him in our league dont).
|
Sembei, if you do the math of the 200 people in the official top 200 (points - divisions offset) and look at everyone bonus pool, then you'll comprove that the bonus pool doesn't make any role on the math.
My theory is that they take the X% better players by MMR (not points), and just then make the Top 200 with only these players. They might actually take the top 200 players by MMR and just sort them by points + divisions.
I can't prove anything though, but one thing is for sure, bonus pool doesn't have a role in the math for top 200, but that makes sense that someone with 200 bonus points so close to your position is actually better (has a better mmr) than you.
|
|
|
|
|
|
EPT
