|
Here are the stats of number match played per race from aligulac from the last two report so it's the top players matches.
To calculate these I made
total_X= 2*num_XvX +num_X*Y+num_XvZ
I count twice mirror matchup because there are 2 players of the same race playing against each other.
http://www.aligulac.com/periods/100/
number of matches played :
P :1406
T : 927
Z :1445
http://www.aligulac.com/periods/101/
number of matches played :
P: 2320
T: 1343
Z: 2021
In both theses period you can obviously see that Terran is vastly underrepresented and that protoss and zerg are relatively even in representation.
Does that mean there is imbalance? I don't have a clue.
But at least, you people, have to acknowledge the underrepresentation of Terran race at top level.
|
Canada11509 Posts
Subtracting player skill seems a bit weird because how do quantify it besides their ladder results? However, I wonder if it has to do with their internal mmr. For instance, at the top level, GM, which is what they would be most concerned about, there is a smaller population base to play on. So they might be playing down as much as they are playing up or their own skill. But if Blizzard knows the mmr, they can drop the weight a bit on wins over opponents with less mmr as well as losses vs opponents with higher mmr. After all, auto-match making must find you an opponent at some point and if it a slow day/ night and has to expand search all the time, the wins and losses might be affected that way.
I'm probably wrong because I have only a foggy notion on how their mmr stuff works, but that's what came to mind when I read this thread.
|
On January 11 2014 00:38 Hider wrote:Show nested quote +On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system.
Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something.
And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now.
|
On January 11 2014 00:38 Hider wrote:Show nested quote +On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system.
You can actually count. Here's the number of people that have at least 1 game played this season:
T: 8
P: 22
Z: 13
When we look at players that the coaches trust, there are nearly triple the P compared to T. And nearly twice the Z.
|
They probably mean they are taking into account the perceived strength of players (like when Stardust got matched against that top bronze dude, they probably didn't give his win as much input) and also the perceived strength of their individual match-ups (like when I went 1-11 PvT this season, they didn't care that much because I'm historically 30% PvT so it makes sense that I keep losing).
Anyway considering how we're all going to agree in the end that these stats are worthless in themselves, can we move on from the stat-creating discussion?
|
On January 11 2014 00:45 The_Red_Viper wrote:Show nested quote +On January 11 2014 00:38 Hider wrote:On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system. Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something. And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now.
Look at the numbers above, if you have 8 T in total, this means that most likely you have only top 10 T in the world in proleague (minus Taeja and Innovation, Bbyong and Polt can be contentious if you're generous), but if there are 22 P in proleague, you cannot necessarily have only top 10 P in proleague.
|
On January 11 2014 00:48 Ghanburighan wrote:Show nested quote +On January 11 2014 00:45 The_Red_Viper wrote:On January 11 2014 00:38 Hider wrote:On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system. Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something. And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now. Look at the numbers above, if you have 8 T in total, that means is that most likely you have only top 10 T in the world in proleague (minus Taeja and Innovation, Bbyong and Polt can be contentious if you're generous), but if there are 22 P in proleague, you cannot necessarily have only top 10 P in proleague.
These statistic is completely useless cause you don't have all the top players for each race playing in proleague.
Protoss has most of the top players in there, terran and zerg not so much.
If you aren't happy with counting the "lesser tosses" into the winrates, then just take the top matches, shouldn't be too hard.
|
3 things:
1 - DK reminds me on the prime minister of my country - so denial, so awkward, and so concerned about nothing as long as the majority aren't "unsatisfied", or the statistics "stay or are shown' as relatively the same
2 - Pretty sure he waited for this "moment of reverse" to say that Protoss can lose.. And it's not my problem that Protosses win, it's the damn lack will to fix things, or at least experiment with that.. Almost as if he's afraid of experimenting to make things any better.. As if the "data" shows everything you know..
3 - BOTH SwarmHost and WM are released in it's primary form.. There were NO changes to both of the units almost NONE - in terms of how they work --> they could've experiment with SHs being less expensive but more mobile and Locusts last shorter or, try with WMs stunning the non-bio units they hit, but no - not a single damn thing to make those units any change.. Almost as if he's afraid to make an improvement at this point..
And now with the nerfs - things got even worse - WM is weak against Zerg while simultaneously still remaining useless against Protoss.. As long as the nerf of WM was in the same time an improvement vs Protoss - would take it any day of the weak, but no - almost as if purposely having the attitude - the unit wasn't meant/intended to fight vs Protoss..
As for the SH - nothing was - not even tested to see if they could make the unit still count but not creating the stalemates..
Seriously, i'm disgusted by the OP post.. I really hate people with that attitude (unfortunately they're around every day) - wait one moment of change and then go - see - told you - it's fine.. Really nihilistic and non-caring at all in almost every possible way
|
On January 11 2014 00:51 The_Red_Viper wrote:Show nested quote +On January 11 2014 00:48 Ghanburighan wrote:On January 11 2014 00:45 The_Red_Viper wrote:On January 11 2014 00:38 Hider wrote:On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system. Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something. And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now. Look at the numbers above, if you have 8 T in total, that means is that most likely you have only top 10 T in the world in proleague (minus Taeja and Innovation, Bbyong and Polt can be contentious if you're generous), but if there are 22 P in proleague, you cannot necessarily have only top 10 P in proleague. These statistic is completely useless cause you don't have all the top players for each race playing in proleague. Protoss has most of the top players in there, terran and zerg not so much. If you aren't happy with counting the "lesser tosses" into the winrates, then just take the top matches, shouldn't be too hard.
Let me explain why that's a bad idea:
Let's only take top 3 of both races, ok? We have easy access to that data:
Top 3 T winrate against P: 86%
Top 3 P winrate against T: 100%
Edit:
Just to set the record straight, for all races, we can find worlds best that aren't playing in season 1 of proleague:
P does not have Dear (you know, the best P of 2013), Liquid'Hero (multiple premier winner).
Z doesn't have Life and Leenock.
|
On January 10 2014 23:37 Hider wrote:Show nested quote +On January 10 2014 23:35 Grumbels wrote:On January 10 2014 23:33 Hider wrote:Especially in the late game against Protoss and a meching Terran, the Zerg has just nothing to deal with these kind of armies without the Swarmhost in its current form (Just watch all the VOD's of Zerg players who don't use Swarmhosts... they just lose at one point).
So: Yes, everyone hates Swarmhosts, nobody hates them more than the Zerg themself, but right now they need them. And as long as they don't change A LOT, the Swarmhost will be needed in the future as well. I actually think it isn't that hard to "fix". By fix I mean we make the following changes; 1) Nerf it into obviolion like Reaper was in WOL. This is simply accepting that the SH is extremely flawed and just bad for the game. In LOTV it can then be replaced by a different unit. 2) Make Ravens 3 supply instead of 2 --> Nerfs terran mech turtle a lot. Swarm Hosts less neccsary. 3) Nerf protoss air (I guess VR's and Tempests are the problem here) --> Makes Swarm hosts less necessary. Alternatively, we could also make HT's (and Ghosts) 3 supply, which will make protoss turtling less efificent. My favored suggestion is to make the locusts last only 20 seconds, but to give the upgrade by default. And maybe add a small health buff. I think the main problem is that swarm hosts are bad in mid-game, but are never vulnerable in late-game. I think regardless of what small change you make to it its bad because the opponent can't really attack into it. Lurkers worked in BW because the zerg had much less stuff + Lurkers covered a much much smaller area. Swarm Hosts cover a way too large, even if they lasts 20 seconds. If they cover a smaller area but you increase HP as a compensation it will likely make it even better when you turtle behind them (since they are more cost effective). I believe the fundament issue here is the problem of free units.
Well, you can't attack into siege tanks either outside of using certain expensive set-ups (vipers, flanks). Stronger locusts that last a shorter time give strong positional control, but only if you babysit them, and they are more vulnerable, rewarding skilled players. I think it would make them less bland, personally, but of course it isn't guaranteed to work.
On January 11 2014 00:38 Hider wrote:Show nested quote +On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system.
As an example of this, let's say that terran is favored vs zerg, then more comparatively weaker terran players will advance, meaning that protoss players will have an easier vs the terran players, meaning that zerg will have an easier time vs protoss players. This way the more obvious tvz imbalance is sort of massaged out of the game and is much less noticeable.
|
hey why not make the game flip a coin at the start so we get exactly 50%, doesnt mean the design is good and its fun to play.
|
On January 11 2014 00:58 Ghanburighan wrote:Show nested quote +On January 11 2014 00:51 The_Red_Viper wrote:On January 11 2014 00:48 Ghanburighan wrote:On January 11 2014 00:45 The_Red_Viper wrote:On January 11 2014 00:38 Hider wrote:On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system. Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something. And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now. Look at the numbers above, if you have 8 T in total, that means is that most likely you have only top 10 T in the world in proleague (minus Taeja and Innovation, Bbyong and Polt can be contentious if you're generous), but if there are 22 P in proleague, you cannot necessarily have only top 10 P in proleague. These statistic is completely useless cause you don't have all the top players for each race playing in proleague. Protoss has most of the top players in there, terran and zerg not so much. If you aren't happy with counting the "lesser tosses" into the winrates, then just take the top matches, shouldn't be too hard. Let me explain why that's a bad idea: Let's only take top 3 of both races, ok? We have easy access to that data: Top 3 T winrate against P: 86% Top 3 P winrate against T: 100%
I am saying if you aren't satisfied with the winrates cause of variance in skill, just look at top terran vs top toss.(and even that is not that usefull, but yeah)
What would be more important to look at is the rosters of the proleague teams and how many toss, terran and zegr players there are overall. And maybe try to look at the overall skilllevel of the races.
|
On January 10 2014 20:34 fx9 wrote:Show nested quote +On January 10 2014 20:17 Faruko wrote:"but Protoss players have lost at a noticeable clip in that tournament. As of this writing, they’ve recorded only 11 wins in 35 non-mirror matchups" Like i said, just look at the higher level protoss, they are doing fine on Proleague (except sOs  !!!), proleague are full of protoss players that dont even have 50% of Parting/Rain/Jilla skills, they are pretty much thinking that the race is far more important and powerful than the player itself But even Rain and Jilla lost last week. So your point is???
Oh come, they lost 1 match Jilla lost to Sniper which is a really good Zerg and Rain to Byul which, again, its a damn good Zerg (4-0 actually)
So my point is, Protoss players which are GOOD players are getting the results they deserve, the other ones... not so much.
So the problem comes from Proleague which are just abusing Protoss even though they have better players in other races
|
On January 11 2014 00:45 The_Red_Viper wrote:Show nested quote +On January 11 2014 00:38 Hider wrote:On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system. Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something. And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now.
Well, he is not a statistician by his CV qualification and I'm pretty sure he isn't the guy who makes those statistics or the MMR and all the ladder calculations.
He is probably that guy that gathers all the balance information from the intern and extern statistics, gathers the feedback from pros and the guys responsible for community feedback and plays the game a lot himself at a high level, and then has to make the decisions.
|
So i hear ToD is a player DK speaks to. Anyone know his email?
|
On January 11 2014 00:43 Falling wrote:
Subtracting player skill seems a bit weird because how do quantify it besides their ladder results? However, I wonder if it has to do with their internal mmr. For instance, at the top level, GM, which is what they would be most concerned about, there is a smaller population base to play on. So they might be playing down as much as they are playing up or their own skill. But if Blizzard knows the mmr, they can drop the weight a bit on wins over opponents with less mmr as well as losses vs opponents with higher mmr. After all, auto-match making must find you an opponent at some point and if it a slow day/ night and has to expand search all the time, the wins and losses might be affected that way.
I'm probably wrong because I have only a foggy notion on how their mmr stuff works, but that's what came to mind when I read this thread.
I think there are two ways: assume that the ladder results are indicative of player skill, (and therefore start out with an assumption that the game is balanced.) Which is then updated to reflect ladder data that shows otherwise. (and as a result you have your newly adjusted win percentages) I think the weakness of this approach is that it will tend to underestimate the imbalance, because the initial assumption might be flawed. The second approach is to assume a normal distribution for player skill, so that if you're in the 50th percentile of mmr for protoss players, that you're equal in skill to a terran player that's in the same percentile, even if the protoss is in diamond and the terran in gold. (of course this breaks down for GM, but then Blizzard's stats don't include GM) I don't know Blizzard's approach though, since they're not exactly forthcoming.
There is also the question of, I guess, variance of mmr per match-up: how consistent is mmr in predicting the outcome of the game. Is it lower for protoss match-ups? (because of all-ins and such) I think it would be interesting information and it's something that certainly should be included in the ladder statistics blizz publishes.
It's all a bit complicated, which is why I'm hoping for someone knowledgeable to inform me about it. I find it very interesting though. :o
|
On January 11 2014 01:07 Big J wrote:Show nested quote +On January 11 2014 00:45 The_Red_Viper wrote:On January 11 2014 00:38 Hider wrote:On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system. Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something. And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now. Well, he is not a statistician by his CV qualification and I'm pretty sure he isn't the guy who makes those statistics or the MMR and all the ladder calculations. He is probably that guy that gathers all the balance information from the intern and extern statistics, gathers the feedback from pros and the guys responsible for community feedback and plays the game a lot himself at a high level, and then has to make the decisions.
Well yeah obviously, but i am pretty sure he has better understanding of these things than 99% on teamliquid.
I just can't stand all these people who think they would do such a good job and preach that DK/Blizzard has no idea about anything, it is just so damn stupid to think.
|
On January 11 2014 01:01 The_Red_Viper wrote:Show nested quote +On January 11 2014 00:58 Ghanburighan wrote:On January 11 2014 00:51 The_Red_Viper wrote:On January 11 2014 00:48 Ghanburighan wrote:On January 11 2014 00:45 The_Red_Viper wrote:On January 11 2014 00:38 Hider wrote:On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system. Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something. And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now. Look at the numbers above, if you have 8 T in total, that means is that most likely you have only top 10 T in the world in proleague (minus Taeja and Innovation, Bbyong and Polt can be contentious if you're generous), but if there are 22 P in proleague, you cannot necessarily have only top 10 P in proleague. These statistic is completely useless cause you don't have all the top players for each race playing in proleague. Protoss has most of the top players in there, terran and zerg not so much. If you aren't happy with counting the "lesser tosses" into the winrates, then just take the top matches, shouldn't be too hard. Let me explain why that's a bad idea: Let's only take top 3 of both races, ok? We have easy access to that data: Top 3 T winrate against P: 86% Top 3 P winrate against T: 100% I am saying if you aren't satisfied with the winrates cause of variance in skill, just look at top terran vs top toss.(and even that is not that usefull, but yeah) What would be more important to look at is the rosters of the proleague teams and how many toss, terran and zegr players there are overall. And maybe try to look at the overall skilllevel of the races.
First, this game of threads is played such that you make a claim and show proof. What you're doing is just making claims, and then having me show the FALLACY of your claims. That's rude and lazy.
Still, let me do that again.
Top P versus top T only?
That's entirely arbitrary because we're basically asking for those with recent wins only. That takes out TY's win against sOs as sOs has lost 3 games in a row and has no wins. It also takes out Zest who I think is excellent, but who has a terrible winrate. That leaves the following games. And even those are pretty damn arbitrary.
Rain>Dream
Sora>Flash
As for rosters, they show NOTHING about balance because the rosters are decided a year in advance and are not flexible enough to respond to meta game shifts and balance patches. This should be obvious.
So, instead of throwing out ridiculous claims, try to figure out something sensible that survives some empirical evidence coming close.
|
On January 11 2014 00:45 The_Red_Viper wrote:Show nested quote +On January 11 2014 00:38 Hider wrote:On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system. Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something. And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now.
Eh... There are more protoss players in proleague than terrans. This means that they are of a relatively lower quality (like top 0.0001% of all terran players and top 0.0002% of all protoss players).
David Kim certainly didn't get his job for his understanding at statistics btw - That's really my main issue w/ him.
What I want for Blizzard is to explain us how they make these adjustments. Their statistic-guy needs to come out and write a post detailing how adjusted win-rates can be 50-50 despite terran being so hugely underpresented.
Well yeah obviously, but i am pretty sure he has better understanding of these things than 99% on teamliquid.
I just can't stand all these people who think they would do such a good job and preach that DK/Blizzard has no idea about anything, it is just so !@#$%^&* stupid to think.
99% ppl on TL doesn't complain about his lack of understanding of statistics. I do, and fair to say my level of education of statistics is at a higher level than David Kim's (giving his education background).
|
On January 11 2014 01:12 The_Red_Viper wrote:Show nested quote +On January 11 2014 01:07 Big J wrote:On January 11 2014 00:45 The_Red_Viper wrote:On January 11 2014 00:38 Hider wrote:On January 11 2014 00:28 Satiinifi wrote:
I guess it all depends on how you measure balance, if you take the best player of each race and make them duke it out the game is probly balanced, has always been, but if you take 100-1000 top players of each race, its extremely extremely fucking imbalanced. Its same as saying that the wealth in the world is fairly balanced, if you take 3 of the richest fuckers out there, but if you include the couple of billion people who are fucking starving, it doesnt look too fair anymore does it. David kim is just a fucking delusional guy living in his very own world, far from realism. But there is also another problem wiht using only the very top... Its not evenly matched. Proleague is still something like this: - Top 30 protoss - Top 10-15 terran - Top 20 zergs Eventually we will also expect that these win-rates go towards 50-50, because the terran players that play in proleague is of a lower decimil (they are the top-top-top while the protoss players are "just" top-top). David Kim simply fails to see that all of the metrics that he use to balance the game, eventually goes towards 50-50. That's just the mechanics of the competitive and the ladder system. Yeah right, he doesn't understand it, i am sure he got the job at blizzard for his looks or something. And yeah the protoss players in proleague are only "top-top", pls name me 5 protoss players that are better than herO, sOs, Rain, Parting that aren't in Proleague right now. Well, he is not a statistician by his CV qualification and I'm pretty sure he isn't the guy who makes those statistics or the MMR and all the ladder calculations. He is probably that guy that gathers all the balance information from the intern and extern statistics, gathers the feedback from pros and the guys responsible for community feedback and plays the game a lot himself at a high level, and then has to make the decisions. Well yeah obviously, but i am pretty sure he has better understanding of these things than 99% on teamliquid. I just can't stand all these people who think they would do such a good job and preach that DK/Blizzard has no idea about anything, it is just so damn stupid to think.
There is a problem with transparency i think. If people knew how some of those stats were calculated it would be a lot better. Showing a 6 day period doesn't help credibility either. It's easy to get the impression that they are only showing what makes them look good, especially with the WOL winfestor precedent and the denial of imbalance we had back then.
|
|
|
|
|
|
EPT
