|
United Kingdom14103 Posts
On January 11 2013 22:02 Ryuhou)aS( wrote:Show nested quote +On January 11 2013 21:21 Qwerty85 wrote:
Even in a perfectly balanced game, if there is like 14 terrans in a tournament and 9 zergs in a tournament, statistically, terran as a race should have higher chances of winning 1st place.
Now we see exact opposite. Even though terrans are still most represented (they get eliminated in code S then get back through code A and Up&down), they have lowest chances of winning 1st place in gsl.
and yet, sourceThere have been 8 Terran first place, 8 Zerg first place, and 3 protoss first place in 1st and 2nd places there've been 17 terrans, 13 zergs, and 8 protoss. seems to me they had the highest chances of winning 1st place in GSL, since in the finals there's been 4 more terrans than zergs, and 9 more terrans than protoss
Yet there has been one Terran win since oct 2011...
Seems to me like they don't have the highest chances of winning...
|
I like how this post ignores everything past the RO32. Simply because there are a vast number of top Terran's in Korea the race will always have strong representation at the start of tournaments, however, the RO16 onward results for Terran are appalling bad in many recent seasons.
|
On December 02 2012 07:19 ChillPhiju wrote:
It sounds alot like: "Terrans had success and were stronger than other races so now another race is allowed to be imbalanced if so be"
And assuming you wanted to say this please don't argue like that.
Or would you like to be enslaved because we enslaved black people just because as a "revenge" they are allowed to. (I am sorry for making such a drastic comparison, but I hope everyone gets it for once and for all)
And on the other hand if Terrans were able to sustain such a amount of them in the GSL till now you shouldn't say they change the Meta because this indicates the complete opposite: They have been changing the Meta etc and it is not working anymore (Which means a buff is needed)
All in all I don't want to say any of this is true(Just wanted to point out some flaws in your argumentation which you got IMO) but Blizzard should let the possibility of buffing open.
It really doesn't sound like that's what he is saying at all, and also you are insane.
|
On January 12 2013 00:09 AlexanderDebois wrote:
I like how this post ignores everything past the RO32. Simply because there are a vast number of top Terran's in Korea the race will always have strong representation at the start of tournaments, however, the RO16 onward results for Terran are appalling bad in many recent seasons.
I didn't think this was true so I went back to check how often Terran had the fewest players throughout 2012;
Season 1
Final was PvZ, up to the final there were more T than Z
Season 2
Always more T than Z
Season 3
Ro8 was 3P - 2T - 3Z (as even as can be with 8 players and three races)
Ro4 was 2P - 1T - 1Z (again as even as possible)
Final was PvP so no T or Z
Season 4
More T than P
Season 5
More T than P until final which was ZvZ so no P or T
It's not as if all the Terran lose early. The only time Terran were the least represented race was season 3 and then not by much, and one of the races has to have the fewest (or tied fewest). The advancement rate of Terran was never far away from half.
|
On January 11 2013 23:09 exog wrote:
Haha so fun to se the terrans squirming trying to undermine the objective numbers to support their view... Cheers with your tears.
yes because the objective numbers would suggest that terran should be the most succesfull race as it is the most played race.
|
I don't think it's appreciated enough, but if balance was even then the racial distribution wouldn't change THAT much from the previous years except for the Code A prelims. It takes a lot to clean out the months of Terran domination if everything was perfectly balanced. I mean if all the terrans just randomly died in Code S ro32 and ro16 then that's A LOT of terrans in Code A 1-3 rounds away from Code S.
If you however had a complete wash and held another Korea WCS or GSL Open with qualifiers then I'm sure the racial distribution would be quite different.
Rather than looking at Code S with absolute terms, I believe balance there should be measured more as a derivative if looking at the GSL only. Though even measuring it as a derivative is a faulty measure.
|
|
|
I hope it stays that way. If it were not for the korean terrans there would be exactly none :/
|
On December 02 2012 07:44 Lunares wrote:
Just give the game time, enjoy what you watch, and wait for HotS if you are unhappy since there just isn't enough time to change the game by then (unless you just want to imbalance it completely, that's easy).
This is the problem with this post, you're trying to imply that the game doesn't need to be changed at all in a drastic way. "give the game time", it's been like over 6 months of broodlord/infestor killing everything. I think that's enough time. "enjoy what you watch", I don't know about you but infestors in every single matchup is not enjoyable no matter how much I "try" to enjoy it.
Koreans have always had the most/best Terrans, so it's no surprise that they're still doing well. Maybe the game is still balanced at the very very top but that doesn't mean we don't need to change the game. The whining is justified.
|
On January 11 2013 22:35 bbm wrote:
If balance is 50/50/50 ie perfectly balanced, and players are evenly skilled, in that, every match is decided with a coin flip, the rest of the world is 33%33%33% races (in that when someone drops out of tourney their replacement is random race), and you start off with 80% terrans, regardless of how many seasons you simulate you'll end up with ~80% terrans left.
I thinjk.
if someone could correct me if i'm wrong i'd be happy to retract this but i'm pretty sure that's the case based on my simple head-maths.
nah you would end up with the distribution the replenishment players have which would be 1:1:1 ratio eventually. It's like you have a jar with 32 balls, 26 blue 3 red 3 green. Each round you take a random half of the jar throw it out and replace it with random balls which have equal chance of being blue, red and green. Eventually all balls will have been replaced and each color is equally likely so you float around even distribution.
Of course this is not very relevant to GSL. I just think the reason terran has always been overrepresented in GSL is that the combination of format, maps and balance have favored terran almost all the time. For a long time terran was simply strongest at korean levels of play especially in a format like GSL where you can prepare extensively and a mappool that hasn't been bad for terran. Last half year or so zerg has been best but the GSL format always lags behind in distribution quite a bit because the rotation is quite slow and dropping from code S you have a good chance to requalify: you get placed into up & down etc. without having to fight through code B etc.
The makeup of the pro teams is also a bit tilted towards terran because they used to be best and Kespa had a higher amount of terran players I think. Finally those terrans also have some benefit from their stage experience etc. already.
All this basically leads to the GSL distribution lagging severely behind the actual distribution by the balance. Foreign tournaments however tend to have much shorter qualification which is probably why those are so zerg infested lately. Foreign tournaments don't feature the same amount of preparation time either which makes it better for zerg and protoss imo. Terran seems to have most options early on with lots of builds so knowing your opponent gives you more oppurtunities to play a style that abuses your opponents weaknesses, zerg however opens pretty much the same every time.
If they kept playing GSL for a long time with no changes it would probably gravitate towards 50% Z 50% T/P like many other tournaments.
Edit: all in all the format of GSL makes it terrible to judge balance upon. GSL numbers in a post related to balance always make me cringe because they are absolutely useless.
Most usefull statistic for balance is just korean GM ladder imo, a good combination between high sample size and only good players. From what I've read most top players use that serious practice anyway and not goofing around too much.
|
I don't get how this kind of statistic proves anything. First there is more top koreans playing terran than other races, then terran dominated the game for nearly one year, either because of opness, maps, or people not figuring out that fungal/Storm are good against bio.
Just watch TvZs, and either you deny that the current state of the matchup is stupid, or you'll just see the fact that zerg production system + cost efficient units can't be beaten in a standard vs standard games by equally skilled players, without major zerg mistake and/or super good play by terran.
See how Symbol rolled Byun this morning with only ling banes into ultras. He didn't even need infestors, he just won because he had TONS of units, because even 3 fast CCs can't beat an economic zerg opening into mass units. Not to take anything out of Symbol who is obviously an amazing player, the thing is that the game wasn't decided by either player's skill in my opinion. Both opened equally economically, and terran got crushed because zerg army supply went up way faster because of how its production works. Could Byun have done something to avoid that? He had tanks and a good position, he shouldn't have lost to just ling/bane. Yet he did, because Symbol was able to remax over and over, was on 4 bases, and teched to hive, while byun couldn't ever leave his base (the only attemps he made, he needed the units to defend).
This kind of game bothers me way more than any statistic. I have no problems with zerg winning a lot, if I see something in game that impresses me. I have problem with zerg winning that easily in the kind of game I described above.
TL;DR : stop posting random stats to cover whine about balance, just watch pro games and try to be objective. Stats in this game will never be a good balance argument in my opinion, because there are too many factors that can ruin them.
|
On December 02 2012 07:18 SKDN wrote:Show nested quote +On December 02 2012 07:15 EU.Pink wrote:On December 02 2012 07:12 peia wrote:this is because earlier, terran WAS the strongest race, ´but until mid 2012 terran has got the weakest (imao) and because earlier there were so many terrans because it was the strongest race, there still is this big MASS of korean terrans, who are still pretty good at the game, they belong into code S, and if you look at EU ladder 10 % terran, 30 % toss 60 % zerg, and also at IPL the race distribution is a big big joke  nice facts about EU ladder bro, can i see source? I believe his race distribution is imaginary if you want source look at http://www.sc2ranks.com/
Not on topic but what the fuck Aphrodite is 11 world wild (forgive the spelling. Don't know how to) ! How come we never see her play ?
On topic,
I believe like everyone that there is a lot of terran players. What you would like to show is ROW8 maybe ? Is it even ?
|
Just for fun I went over the winnings of all GSL Ro32's and added up all the prize money.
Terran: 1.051.000.000 KRW (243 players)
Zerg: 758.000.000 KRW (134 players)
Protoss: 547.000.000 KRW (135 players)
Average winnings:
Terran: 4.325.000 KRW
Zerg: 5.615.000 KRW
Protoss: 4.082.000 KRW
On average, Zerg seems to be the most successful in winning. Protoss and Terran are about the same, but there are just way more Terran players so that's why Terran players have won the most prize money in total.
This has nothing to do with balance! Just some fun facts that show that even though Terran players are more common in GSL Ro32, individually they aren't doing any better than P or Z.
|
On January 11 2013 22:02 Ryuhou)aS( wrote:Show nested quote +On January 11 2013 21:21 Qwerty85 wrote:
Even in a perfectly balanced game, if there is like 14 terrans in a tournament and 9 zergs in a tournament, statistically, terran as a race should have higher chances of winning 1st place.
Now we see exact opposite. Even though terrans are still most represented (they get eliminated in code S then get back through code A and Up&down), they have lowest chances of winning 1st place in gsl.
and yet, sourceThere have been 8 Terran first place, 8 Zerg first place, and 3 protoss first place in 1st and 2nd places there've been 17 terrans, 13 zergs, and 8 protoss. seems to me they had the highest chances of winning 1st place in GSL, since in the finals there's been 4 more terrans than zergs, and 9 more terrans than protoss
it's quite obvious that you know nothing about statistics, i don't even know what you want to show with those numbers. you trying to imply that the % number of medals is in direct connection with racial balance? lol
a good statistical sample is huge, not 19 medals, and the individual samples have to represent the whole perfectly. in what fucking world does MVP- who by the way is nicknamed Game Genie Terran and is probably more skilled than the others, who has a significant % of the medals himself alone- represent the rest of the terrans? with this nonsense you're implying the fact that MVP won 5 medals means that terran is overpowered overall? lol
to enlighten you, if the distribution of z/p/t wins would be absolutely and totally random,(which it isn't even slightly, since it's directly connected to INDIVIUAL PLAYER SKILL, which makes any kind of statistics of sc2 borderline useless in a mathematical sense btw), even then the racial representation wouldn't ever be 33/33/33% with such a small number as the number of sc2 medals. you would need thousands and thousands of medals to make it start to be perfectly even.
in truth the overall winrates are calculated after analyzing thousands of game results, so they are somewhat correlated to racial balance. somewhat. the correlation can never be perfect with individual player skill in the equation. there are people who are better at mathematics than the rest, there are people who are better at learning piano than the rest, and there are players who are better than the rest even if the race they play is at a bad spot. to make the statistics mean even less, the metagame is constantly shifting and there are sometimes bugs/map flaws etc that can be abused that have nothing to do with racial balance. statistics like that can't be used with MVPs and MCs and Lifes who are obviously more skilled than the rest and grab much more championships even if their race is not in a good spot at the moment.
so please don't ever try to use numbers like that to prove anything because you're making educated people's brain hurt. thanks
|
On January 12 2013 02:00 totauksz wrote:Show nested quote +On January 11 2013 22:02 Ryuhou)aS( wrote:On January 11 2013 21:21 Qwerty85 wrote:
Even in a perfectly balanced game, if there is like 14 terrans in a tournament and 9 zergs in a tournament, statistically, terran as a race should have higher chances of winning 1st place.
Now we see exact opposite. Even though terrans are still most represented (they get eliminated in code S then get back through code A and Up&down), they have lowest chances of winning 1st place in gsl.
and yet, sourceThere have been 8 Terran first place, 8 Zerg first place, and 3 protoss first place in 1st and 2nd places there've been 17 terrans, 13 zergs, and 8 protoss. seems to me they had the highest chances of winning 1st place in GSL, since in the finals there's been 4 more terrans than zergs, and 9 more terrans than protoss it's quite obvious that you know nothing about statistics, i don't even know what you want to show with those numbers. you trying to imply that the % number of medals is in direct connection with racial balance? lol a good statistical sample is huge, not 19 medals, and the individual samples have to represent the whole perfectly. in what fucking world does MVP- who by the way is nicknamed Game Genie Terran and is probably more skilled than the others, who has a significant % of the medals himself alone- represent the rest of the terrans? with this nonsense you're implying the fact that MVP won 5 medals means that terran is overpowered overall? lol to enlighten you, if the distribution of z/p/t wins would be absolutely and totally random,(which it isn't even slightly, since it's directly connected to INDIVIUAL PLAYER SKILL, which makes any kind of statistics of sc2 borderline useless in a mathematical sense btw), even then the racial representation wouldn't ever be 33/33/33% with such a small number as the number of sc2 medals. you would need thousands and thousands of medals to make it start to be perfectly even. in truth the overall winrates are calculated after analyzing thousands of game results, so they are somewhat correlated to racial balance. somewhat. the correlation can never be perfect with individual player skill in the equation. there are people who are better at mathematics than the rest, there are people who are better at learning piano than the rest, and there are players who are better than the rest even if the race they play is at a bad spot. to make the statistics mean even less, the metagame is constantly shifting and there are sometimes bugs/map flaws etc that can be abused that have nothing to do with racial balance. statistics like that can't be used with MVPs and MCs and Lifes who are obviously more skilled than the rest and grab much more championships even if their race is not in a good spot at the moment. so please don't ever try to use numbers like that to prove anything because you're making educated people's brain hurt. thanks
don't be such an ass though to someone when you are obviously not good in statistics either.
Low sample sizes have low power but that doesn't mean they can't be indicative.. It also depends on the deviation from the null hypothesis.. Besides individual player skill can be corrected for..
|
oh stats, how useless you are and have always been.
how long till people realize that there's no sense using stats in regards to balance, since it says NOTHING USEFUL.
The only meaningful balance talk is 100% strategy related and done by people who actually understand the game, not just watch or copy builds.
|
On January 12 2013 02:26 Let it Raine wrote:
oh stats, how useless you are and have always been.
how long till people realize that there's no sense using stats in regards to balance, since it says NOTHING USEFUL.
The only meaningful balance talk is 100% strategy related and done by people who actually understand the game, not just watch or copy builds.
Just because you can't understand stats, it does not mean they're bad at explaining balance.
|
On January 12 2013 02:15 Markwerf wrote:Show nested quote +On January 12 2013 02:00 totauksz wrote:On January 11 2013 22:02 Ryuhou)aS( wrote:On January 11 2013 21:21 Qwerty85 wrote:
Even in a perfectly balanced game, if there is like 14 terrans in a tournament and 9 zergs in a tournament, statistically, terran as a race should have higher chances of winning 1st place.
Now we see exact opposite. Even though terrans are still most represented (they get eliminated in code S then get back through code A and Up&down), they have lowest chances of winning 1st place in gsl.
and yet, sourceThere have been 8 Terran first place, 8 Zerg first place, and 3 protoss first place in 1st and 2nd places there've been 17 terrans, 13 zergs, and 8 protoss. seems to me they had the highest chances of winning 1st place in GSL, since in the finals there's been 4 more terrans than zergs, and 9 more terrans than protoss it's quite obvious that you know nothing about statistics, i don't even know what you want to show with those numbers. you trying to imply that the % number of medals is in direct connection with racial balance? lol a good statistical sample is huge, not 19 medals, and the individual samples have to represent the whole perfectly. in what fucking world does MVP- who by the way is nicknamed Game Genie Terran and is probably more skilled than the others, who has a significant % of the medals himself alone- represent the rest of the terrans? with this nonsense you're implying the fact that MVP won 5 medals means that terran is overpowered overall? lol to enlighten you, if the distribution of z/p/t wins would be absolutely and totally random,(which it isn't even slightly, since it's directly connected to INDIVIUAL PLAYER SKILL, which makes any kind of statistics of sc2 borderline useless in a mathematical sense btw), even then the racial representation wouldn't ever be 33/33/33% with such a small number as the number of sc2 medals. you would need thousands and thousands of medals to make it start to be perfectly even. in truth the overall winrates are calculated after analyzing thousands of game results, so they are somewhat correlated to racial balance. somewhat. the correlation can never be perfect with individual player skill in the equation. there are people who are better at mathematics than the rest, there are people who are better at learning piano than the rest, and there are players who are better than the rest even if the race they play is at a bad spot. to make the statistics mean even less, the metagame is constantly shifting and there are sometimes bugs/map flaws etc that can be abused that have nothing to do with racial balance. statistics like that can't be used with MVPs and MCs and Lifes who are obviously more skilled than the rest and grab much more championships even if their race is not in a good spot at the moment. so please don't ever try to use numbers like that to prove anything because you're making educated people's brain hurt. thanks don't be such an ass though to someone when you are obviously not good in statistics either. Low sample sizes have low power but that doesn't mean they can't be indicative.. It also depends on the deviation from the null hypothesis.. Besides individual player skill can be corrected for..
very cute. low sample sizes have NO power, and this whole thing has nothing to do with the null hypothesis. also enlighten me how would you "correct for" individual player skill, i'm very curious.
|
Terran? What is an Terran...never seen or heard of this race for months in Tournaments...^^
I know you Zerg guys enjoy laddering and Torunaments atm...but Blizzard hopefully will fix this somehow...
|
I don't get how this can be used for any balance discussion.
Players get to Code S through skill and a some luck (Up/Down Group draws). If you're going to do a graph for the Ro32, you might as well do it for every other round cause just plainly looking at ro32 is looking at just a piece of a big picture. Just because there's lots of terrans in ro32 doesn't mean the games balanced. Correlation =/= Causation; I could be using this statement wrong since I haven't taken stats in almost 4 years.
|
|
|
|
|
|
EPT
