|
The usual: Don't flame casters or players. Don't discuss balance in this or any LR thread. Don't call your fellow posters idiots. Be excited, happy and entertained. |
On March 17 2013 21:34 Big J wrote:Show nested quote +On March 17 2013 21:25 revoN wrote:On March 17 2013 21:24 Big J wrote:On March 17 2013 21:18 revoN wrote:Sucks that all but one of the KeSPA players are on the same side of the bracket. I guess the finals will be another one-sided roflstomp with Flash/Rain/Bogus/Parting eliminating each other early  Yeah, because that happened so often. Number of Kespa final wins in the last year: 2. Amount of tournaments: 99999999999. It will surely not be "another" one of these. I was talking about this tournament. Have you seen their play? Do you really believe Life or MC stand a chance? All the best players are on the same side of the bracket. No, I missed the two ESF vs Kespa matches, Rain vs Jjakji Innovation vs Leenock. So the best Kespa Protoss can beat a Code B Terran and the best Kespa Terran beat Leenock. At least one of them was to be expected and the other one: yeah Leenock is quite good, so 3-0 is impressive. But you're going quite overboard if you word it like you did.
You forgot that Rain also beat a Code B Protoss as well 
Ok, I might have gone too far with the 'rofltstomp' but judging from what I've seen so far I still believe the finals will be kinda one-sided.
|
United Kingdom14464 Posts
just noticed the jjakji results, my sweet prince I'm sorry.
|
On March 17 2013 21:18 TheBB wrote:Show nested quote +On March 17 2013 21:10 bittman wrote:On March 17 2013 21:06 Zealously wrote:On March 17 2013 21:00 gamerdude12345 wrote:On March 17 2013 20:57 Grovbolle wrote:On March 17 2013 20:45 Shikyo wrote:On March 17 2013 20:12 Grovbolle wrote:On March 17 2013 20:07 Hider wrote:On March 17 2013 19:49 Grovbolle wrote:On March 17 2013 19:39 Hider wrote:
[quote]
Innovation 2%?
This is completely useless. Statistical methods are supposed to help when quantifying stuff is to complicated to do subjectively. If you can't improve the methodology, aligulac is just a waste of time. Based on what? I agree that innovation is underrated in this, but we can't just magically up his rating. It will rise as his results improves. He loses to too many randoms in SPL. Plus we are still waiting for points to "transfer" to kespa players, but since they are very isolated (most only play proleague) it takes time. Please untwist your panties and chill, nobody forces anyone to use it. And no that is not what statistical methods are used for, plus this is "raw" stats, the interpretation is up to you, plus I already stated that currently the system is very shaky (due to the change from WoL to HotS). Plus this Terran surge is based a lot on HotS, of which we have very few matches. In regards to Last, please tell me which historical data that indicates that he is a beast of another world: http://aligulac.com/players/137-Last/results/ Don't publish ratings untill you improve this methodology. LIke what is the point of publishing heavily flawed statistics? This just making aligulac looks foolish, which is unfortunate. We have discussed this, if we should just "shut down" for 6 months untill we have more data, but then again, HotS is just a big patch to the game. When is it okay to publish ratings then? How much data is enough? How should we account for players we "know" are really good but haven't proven it (lots of Kespa players falls in this category)? What if they actually aren't and peoples bias shines through. Say what you will about the method, it is really strong in predicting matches, but strength lies in numbers, not in single brackets. But that being said, we are always trying to improve. Your percentage calculation is really messed up. Last doesn't have a 89% chance to lose the next Bo5. Even if he is a huge underdog. Also, I don't really care who wins and who loses. I just would prefer for both the semis and the finals to be non-mirrors. Please remember this is based mostly on WoL results still. Would you favor last over life in WoL ? (2035) Life 0-0 Last (1466)
-------------------------------------
38.81% 3-0 0-3 1.98%
31.50% 3-1 1-3 4.33%
17.05% 3-2 2-3 6.32%
-------------------------------------
87.36% 12.64%
Median outcome: Life 3-1 Last why use wings of liberty results to judge a HotS tournament? I don't think jinro is the best foreigner just because he got to the ro4 a few years ago. I think this has been answered; a bunch of times each time someone posts an Aligulac prediction. Yeah basically everytime someone posts this people go: "Oh my god your stats suck. I cant believe <favouriteplayerxyz> isn't #1. Did your system not watch him play last round? LOL thank you for this. Yes we are a bit behind schedule implementing the automatic stream-watcher and twitch-chat-reader system that will allow us to give points to hyped players. Hopefully when it's done the ratings will tell everyone what they already know so that we have achieved total uselessness. I'm also excited about the upcoming reddit witchhunt analysis tool that will automatically dump the rating of a player who is caught doing anything slightly controversial.
My points was that the %s are too extreme. People regard Flash as a god in BW but he still had only around a 70% winrate. If you gave me 9-1 odds for a Bo5 between 2 established korean progamers, even if the other player was a large favorite, I'd probably take them every time. I'm not saying that the order of the players is wrong or that you're looking at the wrong stats or anything like that.
It's not like this really matters to me at all but it seemed like my point was misunderstood.
|
On March 17 2013 21:34 Big J wrote:Show nested quote +On March 17 2013 21:25 revoN wrote:On March 17 2013 21:24 Big J wrote:On March 17 2013 21:18 revoN wrote:Sucks that all but one of the KeSPA players are on the same side of the bracket. I guess the finals will be another one-sided roflstomp with Flash/Rain/Bogus/Parting eliminating each other early Yeah, because that happened so often. Number of Kespa final wins in the last year: 2. Amount of tournaments: 99999999999. It will surely not be "another" one of these. I was talking about this tournament. Have you seen their play? Do you really believe Life or MC stand a chance? All the best players are on the same side of the bracket. No, I missed the two ESF vs Kespa matches, Rain vs Jjakji Innovation vs Leenock. So the best Kespa Protoss can beat a Code B Terran and the best Kespa Terran beat Leenock. At least one of them was to be expected and the other one: yeah Leenock is quite good, so 3-0 is impressive. But you're going quite overboard if you word it like you did.
Dont know about best Kespa terran..
It's a toss up between Innovation, and Flash
It's too bad none of the Kespa zergs made it..
Would have been interesting to see how the likes of Effort/Soulkey/Zero/Hydra would have done against the Terrans.
It's interesting how the brackets are stacked up.
The left hand side is eSF + Last, the right is Kespa + Leenock/Jjajki/MKP
|
pff all koreans again. This game can only revive to its old glory when they start inviting much less koreans for mlg imo.
|
At least, instead of clicking on the Liquipedia tab, TL aka indirectly users contrubuting (for free) aka you, could provide the quality level TL use to have. Example, depending on what hotel/resort you go, you expect a certain level of service from some, that's called professionalism.
|
On March 17 2013 21:48 Visas wrote:
pff all koreans again. This game can only revive to its old glory when they start inviting much less koreans for mlg imo.
The tournament would suck without the koreans, gameplay skill would be low, games be stale and the foreigner win wouldnt feel like anything at all because of the prestige the tournament will not carry to that of tournaments with koreans.
|
On March 17 2013 21:48 Visas wrote:
pff all koreans again. This game can only revive to its old glory when they start inviting much less koreans for mlg imo.
Are you serious? Oh yeah let's watch bad players play each other, that would be great!!
|
opterown
Australia54784 Posts
On March 17 2013 21:49 ChiIIgetoutGG wrote:
At least, instead of clicking on the Liquipedia tab, TL aka indirectly users contrubuting (for free) aka you, could provide the quality level TL use to have. Example, depending on what hotel/resort you go, you expect a certain level of service from some, that's called professionalism.
i don't think you know what or who you are talking about; the matchups are available in the second post. the schedule is not known yet, you are one ungrateful poster lol
|
On March 17 2013 21:48 Visas wrote:
pff all koreans again. This game can only revive to its old glory when they start inviting much less koreans for mlg imo.
Trollers gonna troll
|
Pretty much all the matchups left could easily be potential GSL finals matchups
|
Give me games better than what we've had so far, a tall order.
|
I'm so pumped for this. Really just want to see good games and I hope for a Flash vs Life final.
Seriously this grid is so good. I am really looking forward how good Last will be against Life, his bo5 vs Flash and bo7 vs Rain/MC were sick in OSL. Now its bo5, so he's unbeatable hahaa :D!
MC/Bomber, Flash/PartinG, BoguS/Rain ... you cannot ask for more, seriously you can't. Hopefully we will see games that equal the quality of the names.
|
On March 17 2013 21:48 Visas wrote:
pff all koreans again. This game can only revive to its old glory when they start inviting much less koreans for mlg imo.
You can watch your second rate foreigner play SC2 on a live stream, most of us want to see the best of the best
|
On March 17 2013 21:45 Shikyo wrote:Show nested quote +On March 17 2013 21:18 TheBB wrote:On March 17 2013 21:10 bittman wrote:On March 17 2013 21:06 Zealously wrote:On March 17 2013 21:00 gamerdude12345 wrote:On March 17 2013 20:57 Grovbolle wrote:On March 17 2013 20:45 Shikyo wrote:On March 17 2013 20:12 Grovbolle wrote:On March 17 2013 20:07 Hider wrote:On March 17 2013 19:49 Grovbolle wrote:[quote] Based on what? I agree that innovation is underrated in this, but we can't just magically up his rating. It will rise as his results improves. He loses to too many randoms in SPL. Plus we are still waiting for points to "transfer" to kespa players, but since they are very isolated (most only play proleague) it takes time. Please untwist your panties and chill, nobody forces anyone to use it. And no that is not what statistical methods are used for, plus this is "raw" stats, the interpretation is up to you, plus I already stated that currently the system is very shaky (due to the change from WoL to HotS). Plus this Terran surge is based a lot on HotS, of which we have very few matches. In regards to Last, please tell me which historical data that indicates that he is a beast of another world: http://aligulac.com/players/137-Last/results/ Don't publish ratings untill you improve this methodology. LIke what is the point of publishing heavily flawed statistics? This just making aligulac looks foolish, which is unfortunate. We have discussed this, if we should just "shut down" for 6 months untill we have more data, but then again, HotS is just a big patch to the game. When is it okay to publish ratings then? How much data is enough? How should we account for players we "know" are really good but haven't proven it (lots of Kespa players falls in this category)? What if they actually aren't and peoples bias shines through. Say what you will about the method, it is really strong in predicting matches, but strength lies in numbers, not in single brackets. But that being said, we are always trying to improve. Your percentage calculation is really messed up. Last doesn't have a 89% chance to lose the next Bo5. Even if he is a huge underdog. Also, I don't really care who wins and who loses. I just would prefer for both the semis and the finals to be non-mirrors. Please remember this is based mostly on WoL results still. Would you favor last over life in WoL ? (2035) Life 0-0 Last (1466)
-------------------------------------
38.81% 3-0 0-3 1.98%
31.50% 3-1 1-3 4.33%
17.05% 3-2 2-3 6.32%
-------------------------------------
87.36% 12.64%
Median outcome: Life 3-1 Last why use wings of liberty results to judge a HotS tournament? I don't think jinro is the best foreigner just because he got to the ro4 a few years ago. I think this has been answered; a bunch of times each time someone posts an Aligulac prediction. Yeah basically everytime someone posts this people go: "Oh my god your stats suck. I cant believe <favouriteplayerxyz> isn't #1. Did your system not watch him play last round? LOL thank you for this. Yes we are a bit behind schedule implementing the automatic stream-watcher and twitch-chat-reader system that will allow us to give points to hyped players. Hopefully when it's done the ratings will tell everyone what they already know so that we have achieved total uselessness. I'm also excited about the upcoming reddit witchhunt analysis tool that will automatically dump the rating of a player who is caught doing anything slightly controversial. My points was that the %s are too extreme. People regard Flash as a god in BW but he still had only around a 70% winrate. If you gave me 9-1 odds for a Bo5 between 2 established korean progamers, even if the other player was a large favorite, I'd probably take them every time. I'm not saying that the order of the players is wrong or that you're looking at the wrong stats or anything like that. It's not like this really matters to me at all but it seemed like my point was misunderstood.
The reason for this is that because the database is predominantly foreign, the parameters are fitted to the foreign scene. People don't notice this because they are only interested in prediction for Korean matches. It's a demonstrable fact that about 80% of the time a player is given 80% chance to win he wins, but the average user is looking at a biased sample. It's explained in the FAQ.
I still haven't worked out exactly how to solve this problem.
|
On March 17 2013 21:53 Headnoob wrote:
Give me games better than what we've had so far, a tall order.
I dont think their will be any game that beats the bomber and gosuuser game, unless bomber pulls something out of his ass against MC
|
On March 17 2013 21:06 Otolia wrote:Show nested quote +On March 17 2013 21:00 gamerdude12345 wrote:
why use wings of liberty results to judge a HotS tournament? I don't think jinro is the best foreigner just because he got to the ro4 a few years ago.
You seem not to understand clearly. TheBB explained the system there : http://aligulac.com/faq/It's not that we use WoL results to predict HotS but rather that there hasn't been enough games played per player on HotS for the current ranking to be accurate enough. Furthermore, in the case of Last as a KeSPA player he doesn't have enough games to have reached his HotS rating especially considering he is doing very poorly in ProLeague which undermine his rating. Edit : And by accurate enough, I mean accurate enough to prevent criticism such as the one above.
I guess accurate enough is subjective. Sometimes I see it have somewhat decent predictions but then it gives huk a 47% chance to win his match and then last getting less than a percent to win the mlg.
Also, I heard that there was booing in the crowd last night. Did they ever get kicked out or did they just stop? Seems super BM to actually boo a player.
|
On March 17 2013 21:52 LighT. wrote:Show nested quote +On March 17 2013 21:48 Visas wrote:
pff all koreans again. This game can only revive to its old glory when they start inviting much less koreans for mlg imo. Trollers gonna troll
The glory days WERE Koreans, not Incontrol and Destiny waling away at each other.
|
On March 17 2013 21:52 LighT. wrote:Show nested quote +On March 17 2013 21:48 Visas wrote:
pff all koreans again. This game can only revive to its old glory when they start inviting much less koreans for mlg imo. Trollers gonna troll
Some people rather would have _insert any left korean_ 4-0 _insert any foreigner_ in the finals instead of having a good show. While I don't want to disregard foreigners with this statement, they are not close to Korea and I'd rather watch two "unknown" Koreans in any big final instead of a Korea vs non-Korea final which is over after half an our.
SaSe vs TaeJa were not so bad games and Bly made a good stand against Flash. SEn I only heard off wasn't too bad vs PartinG but yeah, we got those match ups left with only Koreans and I couldn't be more excited. All the match ups are actually so tough to call and could go either way, that its really thrilling. If people would stop hating on the Koreans and actually read into them and their scene they would probably enjoy a tournamnet a lot more.
|
does some1 know where to find the polt vs life vods?
|
|
|
|
|
|
EPT
