|
On October 08 2015 08:30 oo_Wonderful_oo wrote:
So you're basically stating that Tigers lucked out to Worlds and never were good.
I see.
combining a straw man with a slippery slope makes your argument super weak.
like if you want to draw that KOO was the best team in the world based on one bo3 vs skt and a bunch of teams who where rebuilding (including SKT playing tom sometimes). And then have them jump of a cliff a month later for iem then suddenly be back for summer then fine. I'm not arguing that KOO is or was not good. I am rejecting the statement that they are the best team in the world at that time. But like your argument here your seem to be only able to view things in a binary fashion.
So if we have SKT> KT > KOO/ cj/JAG currently at the end of summer.
now kt had score doing a role swap so back in spring then they would be on a lower tier and cj is has ambition jungle.
so in spring i'm saying it was more like
SKT=KOO>rest of lck
with SKT being a known strong team. We have 1 bo3 and a strong spring season to put KOO on a similar level at the time. If you want to put KOO is better then argue it but they didn't play anywhere near enough games to qualify that statement that they clearly best in the world. You start seeing the progression of this tier list to the first once after iem where SKT pulls ahead of KOO and the rest of lck in more in league with KOO during round robin of summer.
important things to note
KOO has remained a top team in korea throughout what i have said. Just because they are not the best in korea doesn't mean that are not good. Deserving of a worlds place is not equivalent to be the best team during a round robin in summer. Can you expand what you mean on this? If you can point out where i said KOO shouldn't be at worlds tell me. Or that KOO is a terrible team.
|
Idk why you backed down even an inch koo is lucky to be at worlds
|
See, I think a lot of people (thorin, the Chinese experts themselves, other analysts who predicted similarly) are arguing with Twitter or some imaginary critic. They are like, "well you didn't do any better" or "we used the right mental process to come to xxx predictions." They are refusing to engage with people who make the substantive criticisms:
1) There are critics who have consistently stated that they overrated the strength of Chinese (born) players, and thus likely the overall team strengths of those teams.
2) There is criticism of people like Montecristo who basically accepted fairly uncritically #1. His major concerns were with the infrastructure. Which leads us to #3...
3) The Chinese experts (who would lead us to believe they used the correct mental processes) essentially facilitated insider trading. This is revealed when they all wrote up the "excuses" pieces this week talking about iG's vacation and LGD's internal struggles.
What this means is that even if China goes 9-0 next week, they still aren't vindicated, because they all are backpedaling and making these excuses. If some guy like Drexxin stood up and said, "Fuck it, all this is bullshit, 9-0 China week 2, and Kid/Clearlove/GodV will be the team MVPs", and that happens, he would be vindicated, but the rest would still be clowns.
|
The land of freedom23126 Posts
On October 08 2015 09:26 MuddyJam wrote:Show nested quote +On October 08 2015 08:30 oo_Wonderful_oo wrote:
So you're basically stating that Tigers lucked out to Worlds and never were good.
I see.
combining a straw man with a slippery slope makes your argument super weak. like if you want to draw that KOO was the best team in the world based on one bo3 vs skt and a bunch of teams who where rebuilding (including SKT playing tom sometimes). And then have them jump of a cliff a month later for iem then suddenly be back for summer then fine. I'm not arguing that KOO is or was not good. I am rejecting the statement that they are the best team in the world at that time. But like your argument here your seem to be only able to view things in a binary fashion. So if we have SKT> KT > KOO/ cj/JAG currently at the end of summer. now kt had score doing a role swap so back in spring then they would be on a lower tier and cj is has ambition jungle. so in spring i'm saying it was more like SKT=KOO>rest of lck with SKT being a known strong team. We have 1 bo3 and a strong spring season to put KOO on a similar level at the time. If you want to put KOO is better then argue it but they didn't play anywhere near enough games to qualify that statement that they clearly best in the world. You start seeing the progression of this tier list to the first once after iem where SKT pulls ahead of KOO and the rest of lck in more in league with KOO during round robin of summer. important things to note KOO has remained a top team in korea throughout what i have said. Just because they are not the best in korea doesn't mean that are not good. Deserving of a worlds place is not equivalent to be the best team during a round robin in summer. Can you expand what you mean on this? If you can point out where i said KOO shouldn't be at worlds tell me. Or that KOO is a terrible team.
What is your problem, lol.
Initial argument was that LimpingGOAT always was saying that GE Tigers isn't even close to be best team in the world but it's blasphemy, because GE Tigers was coming undefeated in matches and dropping only 2 sets in first half of LCK Spring. while EDG was busy 38-6'ing LPL, where they were only good team initially (Vici Gaming came to power post-IEM Katowice, so did LGD, Snake were running 2 comps at a time and praising lord Ella).
You're trying to argue now that GE Tigers wasn't best team in Korea in WHOLE LCK Spring.
Who the fuck cares about whole LCK Spring, if span of time we were discussing is pre-IEM Katowice.
On October 08 2015 09:56 Slusher wrote:
Idk why you backed down even an inch koo is lucky to be at worlds
Should've let Ggoong and Watch to go.
|
On October 08 2015 04:53 LimpingGoat wrote:
I also thinks its retarded that people on Reddit, and Thoorin even made a video defending analysts looking stupid in retrospect after this Worlds' group stages.
If your analysis completely dismissed the possibility of the actual outcome, it should be criticized and the analyst should take the opportunity to improve themselves, not just say fuck you guys can't predict everything it was right at the time hurr durr.
They made the video to try to educate people like you. In fact, Thorin explicitly discussed that it is time-consuming and impractical to go through every possible outcome, but if you did, you'd comment on the small but relevant chances of an upset... But since reality is bounded by space-time, we lack the ability to discuss these things ad nauseum to make it painfully obvious to "critics" such as yourself that anyone with a grasp on reality might have given C9, for example, a 30-40% chance of upset each game. But that means, as Thorin described, if you never give them more than 50% chance of winning a game, the accurate prediction is 0-2.
And even as a C9 fan, I have 0 criticisms about the analysis. I bet against them too, though I hoped I was wrong. There was no evidence to support such improvement. LDG maybe we could have predicted a decline after the coaching staff debacle though the degree was very surprising. Group A, meh, not that far from expectation. Group C is performing as expected.
|
On October 08 2015 05:23 Numy wrote:Show nested quote +On October 08 2015 05:17 LimpingGoat wrote:
It's true that I am also overconfident in my own predictions sometimes. But I don't like people justifying wrong predictions with "they were right at the time". I would never get something wrong and then defend it, that's retarded. If that means I stop making predictions on things that are really close or have volatile formats such as bo1, then so be it (and so should actually popular analysts/prediction makers). It's better than calling it one way and being wrong. After MSI's failure you defended your assessment that TSM was a top tier team and merely that "Bo1s don't matter". Then before worlds you predicted TSM would easily make it out of groups, going so far as to say they could beat anyone in a bo5. This is the pot calling the kettle black.
No I didn't. I defended TSM's strength due to the format that they failed at after MSI, which is different from defending any prediction I made that ended up being wrong. To be honest I don't remember exactly what I said before MSI, but certainly if I said that TSM will definitely make it out of groups, that would be a stupid prediction and I wouldn't ever defend it. In the same way that I said TSM will 100 percent get out of groups which was stupid to say. Most of my shitty predictions happen because I get baited into saying bold stuff that I shouldn't say. But now that you mention it, bo1's really don't matter in the sense that they can be disgustingly misleading for fans and it's a real coin toss as to whether or not they are an actual representation of what an extended series matchup would look like between two teams.
|
Probabilistic predictions are a much more accurate way to determine how good a forecaster is. But even If C9 had a 40% chance in each game, then that means the favored team is only 36% to win both games in a row, thus the more likely outcome is that C9 wins one of the two games. It's lazy predictions using deterministic criteria that got exposed by this tournament the same way Nate Silver eviscerated the political "gut feeling" pundits, which are what today's so-called analysts are.
Where mathematical models replace eyeball test idiots like Thorin, Monte, Yamato, and all the rest, like in betting markets, the models simply do better. Your average expert analyst in league of legends is only slightly better than a coin flip to predict a game, despite claims of expertise.
|
Bearded Elder29903 Posts
Worlds back today baby ~~
|
On October 08 2015 12:59 Sleight wrote:Show nested quote +On October 08 2015 04:53 LimpingGoat wrote:
I also thinks its retarded that people on Reddit, and Thoorin even made a video defending analysts looking stupid in retrospect after this Worlds' group stages.
If your analysis completely dismissed the possibility of the actual outcome, it should be criticized and the analyst should take the opportunity to improve themselves, not just say fuck you guys can't predict everything it was right at the time hurr durr. They made the video to try to educate people like you. In fact, Thorin explicitly discussed that it is time-consuming and impractical to go through every possible outcome, but if you did, you'd comment on the small but relevant chances of an upset... But since reality is bounded by space-time, we lack the ability to discuss these things ad nauseum to make it painfully obvious to "critics" such as yourself that anyone with a grasp on reality might have given C9, for example, a 30-40% chance of upset each game. But that means, as Thorin described, if you never give them more than 50% chance of winning a game, the accurate prediction is 0-2. And even as a C9 fan, I have 0 criticisms about the analysis. I bet against them too, though I hoped I was wrong. There was no evidence to support such improvement. LDG maybe we could have predicted a decline after the coaching staff debacle though the degree was very surprising. Group A, meh, not that far from expectation. Group C is performing as expected.
It's not about going through every possible outcome, it's about not exaggerating shit in your predictions. "I'd be surprised if TSM will win even a single game", when TSM is obviously the caliber of team when it's fully possible, and quite likely that they will at least win one game. The biggest problem is just trying to predict bo1 games between top teams. I think it's much better if analysts just don't predict a bo1 between say C9 and AHQ than if they make an arbitrary prediction so they can gloat if they are right. You can say it's meaningless if they are wrong because it's a bo1 and they can't account for every possibility of course there is a chance it can go either way. But then why make it in the first place, because it will be just as meaningless if the prediction is right.
|
Predictions for today's games guys? I want to say...
KOO > CLG
FW > Pain
KOO > Pain
FW > CLG (not sure on this one but just not feeling CLG)
KOO > FW
CLG > Pain
Wish I had more knowledge of the foreign scenes though 
|
On October 08 2015 13:41 739 wrote:
Worlds back today baby ~~
So hype. Can't participate in the thread at work, but can get away with watching occasionally. Going to be a really fun week to watch
|
On October 08 2015 13:39 NovaTheFeared wrote:
Probabilistic predictions are a much more accurate way to determine how good a forecaster is. But even If C9 had a 40% chance in each game, then that means the favored team is only 36% to win both games in a row, thus the more likely outcome is that C9 wins one of the two games. It's lazy predictions using deterministic criteria that got exposed by this tournament the same way Nate Silver eviscerated the political "gut feeling" pundits, which are what today's so-called analysts are.
Where mathematical models replace eyeball test idiots like Thorin, Monte, Yamato, and all the rest, like in betting markets, the models simply do better. Your average expert analyst in league of legends is only slightly better than a coin flip to predict a game, despite claims of expertise.
This is basically all that needs to be said. If the "experts" had a basic understanding of statistics and checked the betting odds before they went on air they'd all be way better at predicting (eg C9/TSM going 0-6 was statistically unlikely).
|
Poland3748 Posts
On October 08 2015 07:03 Poopi wrote:Show nested quote +On October 08 2015 06:53 nimdil wrote:On October 08 2015 05:58 Gorsameth wrote:On October 08 2015 05:42 Poopi wrote:On October 08 2015 04:42 nimdil wrote:On October 08 2015 03:50 Poopi wrote:About the Hai saying "we are good" and stuff, I think I read in a LemonNation interview that C9 were indeed doing bad in scrims (which so far isn't very promising when you are the outsider already), so they probably didn't expect to 3-0 even if they thought it was possible. I don't think saying out loud that they are good will hinder them so why not  . Iirc the only way for them not to qualify would be losing all of their games, which is very unlikely to happen. Worth to point out that assuming every match is 50:50 (which is stupid but still), C9 is in the best position Are you sure? 87.5% chance to advance would say yes they are in the best position (assuming 50:50 for every match) It's 87.5% they win at least one match and they will advance no questions asked. But even if they go 0-3 it's still not 0. I.e. if they go 0-3, one of the teams go 3-0 and of course the other two goes 2-1 and 1-2. They are tied for 2nd place so still ~50% chance to win. Different scenario is that they go 0-3 and all other teams go 2-1. In such case it's 4 way tie and it's still 25%. Truth be told I didn't calculate the odds. Sorry I just found mildly amusing that you found worth to point out that C9 is in the best position, since C9 is 1st at 3-0 and the other teams are at 1-2. If we knew the exact odds of the future C9 matches, the only way for them not to be the favorites to qualify, being 3-0, would be if they had like only 5% chances to beat the other teams or something like that... And even paiN / BKT probably would win at least 1 games out of 20 against teams from their group, despite being assumed as the weakest teams of the tourmanent by far, vs some of the best.
Well the difference between C9 and SKT for coin-flip results are a lot closer than you apparently think (assuming my calculations are correct): 92,97% for C9 and 92,31% for SKT/Origen.
|
The main issue is their income is linked to fans viewing them well and reading their content which now leaves a period of "o shit" and looking like a farce during the biggest lol competition of the year. But lets backtrack.
So all year china talk people and thorin hype up china. They seem to of mostly by people like thorin and Froskurinn being condescending and just dismissive of anything lcs. Maybe they viewed it in part as a strategy to get people to care about lpl. So naturally you get a backlash when the results all go against them across multiple teams. Then is comes down to what you said about those games. If you pointed to certain areas of interest and talk about details of the macro play then your analysis holds up. Or did they infact just go no they have no chance and give no real insight.
The whole percent thing seems to miss the point to me. You play a bo1 which gives no indication on a percentage (you can play a bo5 and still have no indication). Its more important to talk about aspects of the game for example predict various smaller things. Say as cs lead at 10 mins on a tower laner after a laneswap. People are too focused on getting the winner right than the reasoning. There is a difference between being analytical and being talking about lol alot.
predictions
KOO > CLG >FW >PN
|
On October 08 2015 12:59 Sleight wrote:But that means, as Thorin described, if you never give them more than 50% chance of winning a game, the accurate prediction is 0-2. OK but can you explain to me why the "accurate" prediction shouldn't simply be what is most likely to happen? Like every other person in this thread with a basic undestanding of statistics points out that being favoured (or disfavoured) in each game usually doesn't result in sweeps.
|
On October 08 2015 14:32 Mondeezy wrote:Predictions for today's games guys? I want to say... KOO > CLG FW > Pain KOO > Pain FW > CLG (not sure on this one but just not feeling CLG) KOO > FW CLG > Pain Wish I had more knowledge of the foreign scenes though
FW vs CLG is hard to predict but I think CLG has it unless their morale was destroyed by KOO or they lose pick/ban horribly
|
Oh shit group stages start so early. Forgot we now 2 hours behind.
|
On October 08 2015 15:07 tarath wrote:Show nested quote +On October 08 2015 13:39 NovaTheFeared wrote:
Probabilistic predictions are a much more accurate way to determine how good a forecaster is. But even If C9 had a 40% chance in each game, then that means the favored team is only 36% to win both games in a row, thus the more likely outcome is that C9 wins one of the two games. It's lazy predictions using deterministic criteria that got exposed by this tournament the same way Nate Silver eviscerated the political "gut feeling" pundits, which are what today's so-called analysts are.
Where mathematical models replace eyeball test idiots like Thorin, Monte, Yamato, and all the rest, like in betting markets, the models simply do better. Your average expert analyst in league of legends is only slightly better than a coin flip to predict a game, despite claims of expertise. This is basically all that needs to be said. If the "experts" had a basic understanding of statistics and checked the betting odds before they went on air they'd all be way better at predicting (eg C9/TSM going 0-6 was statistically unlikely).
Why is Monte considered an "eyeball test idiot" (whatever the fuck that means)? Because he said he "would never count out Hais ability to pull out a win due to epic shot calling" when asked if c9 would go 0-6? Or was it when he said it was possible (just possible) that TSM wouldn't win a game? Neither statement is even remotely unreasonable given what we knew a fortnight ago.
The game isn't about statistics. The game isn't about flipping a coin 1000 times and having it come up heads more times than tails. It's absolutely possible to watch performances of players and teams, judge how well they react and execute and then make a call on their relative strength compared to opponents they're facing.
Not to mention I don't remember a single post by any of the analyst haters here bigging up c9 or predicting LGD would crash and burn. Feel free to put up or shut up next time.
|
Poland3748 Posts
It's funny that we complain that analysts are wrong.
I admit I love the idea of unpredictable results at Worlds and if analysts are correct, the results are not unexpected and so the tournament is predictable. So the fact that at this world week 1 analysts are frequently incorrect doesn't necessarily mean they are bad at what they are doing because 100% correct predictions are impossible.
|
|
|
|
|
|
|
|
EPT
