|
On November 21 2013 02:45 Courthead wrote:Show nested quote +On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Show nested quote +On November 21 2013 02:42 Wingblade wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Exactly this, there's almost no way to determine which races are being chosen, especially for players who are playing for the first time. I originally chose Protoss because they were super advanced aliens and had a guy which was really tall and shot freaking lasers from his eyes(dead serious). Determining a causal relationship between the two isn't possible, because there are a variety of factors which influence the race a player chooses. There is a potential correlation, but even that can be skewed by the potential exodus of Terran players at the end of Wings of Liberty because of broodlord/infestor. The bolded statement is something that I would love to be able to test. Why do people keep making this tired argument? Racial distributions are basically even across ladder. It's 31% Zerg, 31% Protoss, 30% Terran, and 8% Random. There is basically no difference in the number of people playing each race. The F2P version of the game is Terran only though, right?
|
On November 21 2013 02:37 Wingblade wrote:Show nested quote +On November 21 2013 02:33 keglu wrote:On November 20 2013 09:25 vRadiatioNv wrote:On November 20 2013 09:22 TheDwf wrote:On November 20 2013 09:19 vRadiatioNv wrote:On November 20 2013 08:21 Ghanburighan wrote:On November 20 2013 08:06 vRadiatioNv wrote:On November 20 2013 07:56 Ghanburighan wrote:On November 20 2013 07:47 vRadiatioNv wrote:On November 20 2013 07:32 Courthead wrote:
[quote]
So we have to ask ourselves, if it's literally 100% easier for certain races to achieve success on ladder, what does that say about game balance? What does that say about how we should be dealing with those expressing frustration at ladder play? What does that say about Blizzard's repeated pronouncements that "everything looks good"? Except you cannot conclude that from these numbers at all. Maybe the reason most people don't play Terran is that they think it's stale using Marines as the primary unit in every single match up (which is exactly what most people complain about). There can be plenty of other reasons as well but you cannot conclude from those numbers that Terran is underpowered or to what degree it may be underpowered. That doesn't make sense. That would suggest that there are less terran players all together, but that's not true. Terrans are only underrepresented at higher levels, which suggests that it's harder to reach higher levels with terran. The sad thing is, I can barely remember a time when this wasn't the case. Except the people who are higher on the ladder most likely practice more often. If the playstyles are stale they will be less likely to continue playing. If I only play an hour or two every day at a low level I probably wouldn't get bored with Terran. Again, you cannot draw conclusions of balance off these numbers alone. That's mere speculation at best; making it up as you go along is more likely. As others have said, there are reasons why the ladder is dominated by P and T is underrepresented (all-ins help if you are anonymous and bo1) but there are unanswered questions. First, most pro players know each other's barcodes, so anonymity is not guaranteed. Second, Korea is as bad as the other regions, despite having a strong showing of terrans. Third, zerg doesn't benefit from all-ins, and it's especially vulnerable to them, yet it outperforms terran on ladder. The bo1 theory does not account for this. Lolwhat. My argument is at least as reasonable (imo moreso) than "omgwtfbbq look at % of ppl playing terran = UP." As many people have already said the winrates do not show Terran being absurdly underpowered. My argument makes much more sense and I only said there are plenty of other explanations and, again, low % of terran players does not directly translate to imbalance. It is you who are trying to make these figures seem like more than they are. Your argument about Terrans being underrepresented because their high level players would be more bored by their race is totally far-fetched. You don't even have the beginning of a proof for that. Show me your proof that low % of Terrans in GM = imbalance. Oh that's right, there is none. The winrates are within 1%, you have nothing. But keep complaining if you want. Drop ladder levels and put all players on one level. In this scenario average T player winrate would be lowest of all races . Ladder level race distribution is as much balance indicator for ladder like winrates are for tournaments. Also for people mentioning WoL. In WoL situation never occured that on GM level one race had over 40% and other below 25%. It was more like 35/35/30 and the end of WoL Where are you finding these numbers on Wings of Liberty GM distribution? I have my own theory regarding GM race distribution and I need access to the GM race distribution numbers, particularly early 2012 and late 2012. Also, it would help with your point if you actually had a source to back it up.
Unfortunately i can't find it now but i often was checking it back then when sc2ranks allowed to veify distribution for previous patch/seasons.
|
4713 Posts
I think the racial distribution at the top of the ladder, particularly GM and Masters should have stabilized by now, because the cases of people choosing a certain race because of how it feels or how cool they perceive it to be, should have already occurred and then the ladder started being balanced around skill level.
By now it doesn't matter if someone chose Protoss because its cool if he is stuck in Gold because that is what his skill level is. Its the skill level that should matter when factoring in ladder distribution.
Now how you guys correlate win rates to race distribution on ladder and what additional data you bring to the table is your business.
|
On November 21 2013 02:42 Wingblade wrote:Show nested quote +On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Exactly this, there's almost no way to determine which races are being chosen, especially for players who are playing for the first time. I originally chose Protoss because they were super advanced aliens and had a guy which was really tall and shot freaking lasers from his eyes(dead serious). Determining a causal relationship between the two isn't possible, because there are a variety of factors which influence the race a player chooses. There is a potential correlation, but even that can be skewed by the potential exodus of Terran players at the end of Wings of Liberty because of broodlord/infestor. The bolded statement is something that I would love to be able to test.
Good thing we dont talk about race distribution which is very close to even (something like 32-31-29) but about ladder level distribution inside playerbase of specifica races.
|
On November 21 2013 02:45 Courthead wrote:Show nested quote +On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Show nested quote +On November 21 2013 02:42 Wingblade wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Exactly this, there's almost no way to determine which races are being chosen, especially for players who are playing for the first time. I originally chose Protoss because they were super advanced aliens and had a guy which was really tall and shot freaking lasers from his eyes(dead serious). Determining a causal relationship between the two isn't possible, because there are a variety of factors which influence the race a player chooses. There is a potential correlation, but even that can be skewed by the potential exodus of Terran players at the end of Wings of Liberty because of broodlord/infestor. The bolded statement is something that I would love to be able to test. Why do people keep making this tired argument? Racial distributions are basically even across ladder. It's 31% Zerg, 31% Protoss, 30% Terran, and 8% Random. There is basically no difference in the number of people playing each race.There was no mass exodus of Terrans. That's total bs. And if you're planning to advance some conspiracy theory like "a lot of pros switched", then you'll need to provide some actual evidence, otherwise it's meaningless.
That number of terran mostly comes from bronze league, where terrans are majority. Every other league and region terran has far less than other 2 races. Check for yourself:
Sc2Ranks race Stats
|
On November 21 2013 02:53 DaRKMaTT3r wrote:Show nested quote +On November 21 2013 02:45 Courthead wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? On November 21 2013 02:42 Wingblade wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Exactly this, there's almost no way to determine which races are being chosen, especially for players who are playing for the first time. I originally chose Protoss because they were super advanced aliens and had a guy which was really tall and shot freaking lasers from his eyes(dead serious). Determining a causal relationship between the two isn't possible, because there are a variety of factors which influence the race a player chooses. There is a potential correlation, but even that can be skewed by the potential exodus of Terran players at the end of Wings of Liberty because of broodlord/infestor. The bolded statement is something that I would love to be able to test. Why do people keep making this tired argument? Racial distributions are basically even across ladder. It's 31% Zerg, 31% Protoss, 30% Terran, and 8% Random. There is basically no difference in the number of people playing each race.There was no mass exodus of Terrans. That's total bs. And if you're planning to advance some conspiracy theory like "a lot of pros switched", then you'll need to provide some actual evidence, otherwise it's meaningless. That number of terran mostly comes from bronze league, where terrans are majority. Every other league and region terran has far less than other 2 races. Check for yourself: Sc2Ranks race Stats
Global racial distribution at bronze:
Zerg
13,135 (8.62%)
Protoss
16,775 (11.00%)
Terran
17,769 (11.66%)
Random
5,674 (3.72%)
Am I missing something?
|
On November 21 2013 02:53 DaRKMaTT3r wrote:Show nested quote +On November 21 2013 02:45 Courthead wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? On November 21 2013 02:42 Wingblade wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Exactly this, there's almost no way to determine which races are being chosen, especially for players who are playing for the first time. I originally chose Protoss because they were super advanced aliens and had a guy which was really tall and shot freaking lasers from his eyes(dead serious). Determining a causal relationship between the two isn't possible, because there are a variety of factors which influence the race a player chooses. There is a potential correlation, but even that can be skewed by the potential exodus of Terran players at the end of Wings of Liberty because of broodlord/infestor. The bolded statement is something that I would love to be able to test. Why do people keep making this tired argument? Racial distributions are basically even across ladder. It's 31% Zerg, 31% Protoss, 30% Terran, and 8% Random. There is basically no difference in the number of people playing each race.There was no mass exodus of Terrans. That's total bs. And if you're planning to advance some conspiracy theory like "a lot of pros switched", then you'll need to provide some actual evidence, otherwise it's meaningless. That number of terran mostly comes from bronze league, where terrans are majority. Every other league and region terran has far less than other 2 races. Check for yourself: Sc2Ranks race Stats
I don't think the difference is statistically significant. Even if it was, and we decided to erase Bronze league from the equation, the distribution discrepancies you see at the top of the ladder still show imbalance. Look at the process of going from Diamond to Master to GM:
Diamond: 29% T, 34% P, 37% Z
Master: 27% T, 37% P, 36% Z
GM: 24% T, 42% P, 34% Z
So Zerg stays relatively constant, but Protoss begins to dominate and Terrans begin to fall off. The difference is even more pronounced on the NA server, where Diamond is 30% of Terrans but GM is down to 21%. In a balanced game, there would be similar levels of success regardless of race. No matter how you look at it, the fact is that the race you choose significantly affects your chances of advancing into the higher leagues.
|
This is again discounting that the F2P version of the game only has Terran available as an option for newcomers. So that will bring a disproportionate swell of Terrans into the lower leagues.
|
On November 21 2013 03:06 RampancyTW wrote:
This is again discounting that the F2P version of the game only has Terran available as an option for newcomers. So that will bring a disproportionate swell of Terrans into the lower leagues.
That's why I only analyzed the 3 highest leagues in my analysis.
|
Nvm, F2P version has no 1v1.
|
On November 21 2013 03:06 RampancyTW wrote:
This is again discounting that the F2P version of the game only has Terran available as an option for newcomers. So that will bring a disproportionate swell of Terrans into the lower leagues.
In WoL most common argument was "Terran campaign so more Terran casual players". I don't know what is argument for HoTS.
|
On November 21 2013 03:07 RampancyTW wrote:
Nvm, F2P version has no 1v1.
And I don't know how much F2P version of Starcraft2 would bring in 1v1 players. I can see arcade and team games numbers increasing more than 1v1 (as it is the case in the paid version).
/edit
But seriously, Terrans please be patient and see how the patch plays out in professional games. The combined mech upgrades are exciting, as a Terran and definitely a buff to my bio/mech style on ladder <3.
/edit
Also the "conspiracy" about David Kim's race is ridiculously stupid and most likely a troll post.
|
On November 21 2013 03:12 plogamer wrote:
But seriously, Terrans please be patient and see how the patch plays out in professional games. The combined mech upgrades are exciting, as a Terran and definitely a buff to my bio/mech style on ladder <3.
We already could see the difference in Homestory cup and it is negligible.
Sadly mech needs to wait for a new expansion or a new design team.
|
On November 21 2013 02:38 Big J wrote:Show nested quote +Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame?
Distribution data should be assessed on the basis of whether a given race is under or overrepresented in a given league. If used correctly, it is the best indicator of balance that us outsiders have access to - Especially once we take into account two factors which could skew the distribution data.
1) Terran was in early WOL chosen by player of weaker caliber (due to new players being familar with them from the campaign). Zerg on the other hand was more likely to be chosen by players of higher caliber. This means that in early WOL zerg distribution should be skewed more towards the higher leagues than terran.
2) Terran as a % of all active players has since declined and is now the least played race. This means that more terran players have quitted the game than zerg players. I believe it is logical to make an assumption that the current distribution data is impaced by survivorship-bias as the players that have left the game in general used more "gimmicky/unreliable/non-sustainable/less hard-working" strategies and were of weaker quality than the one's who kept playing the game.
Now, let's assume that the survivorship bias (2) equals out the chosen race-bias (1). The implications should be this;
- If the game isproperly balanced across all skill levels today, distribution should be equal across all leagues. This means the the top x% of terran players playing in a given league is the same as top x% of zerg players playing in a given league.
- In the scenario where Sc2 is imbalanced, simple logic tells us that that the race that is underpowered will be overpresented in the lower leagues (bronze/silver) and underpresented in higher leagues.
Since the latter is in fact what we have seen with terran for the last year, it is is reasonable to assume that terran is underpowered. At least that's the best interpreation we can get from the available data, but ofc there are still some uncertainties which could impact the results.
The problem though is that I am not sure that David Kim has any clue on how to interpret statistics since he keeps mentioning matchmaking w/r as an indicator of balance - That is pure nonsense as it always will be around 50-50. W/R can only be used to show whether PvT is more favored towards protoss than ZvT is favored towards zerg for instance. But it can't be used to show whether a race is underpowered, balanced or overpowered.
If one remembers back to early WOL, late WOL beta, Dustin Browder made comments that ZvT was balanced as W/R was around 50-50. That was pure nonsene and was simply a result of superior zerg players being matched up against inferior terran players. Dustin Browder showed no signs up showing he understood how to interpret the statistics back then, and neither does David Kim today. Today, we are likely seeing the reverse situation.
An alternative interpreation is that David Kim is aware that terran likely is underpowered in terms of raw ladder statistics, but that they simple perform too well in tournaments in order for a nerf to be warranted. But instead of admitting that they are downpriotizing ladder-balance, they instead "manipulate" the general public into believing the ladder is balanced by showing these irrelevant W/R-ratio's.
|
Just had an idea, why not make a "spread units apart" button? Either by spreading more and more while holding it down or making them move out within a bow made by the mouse. Ofc, all aoe would need serious buffs, and a lot of things would need redesigning (marinerange?), but clumping is advantageous quite often (dps/focusfire), so it would take skill to use such a function.
|
On November 21 2013 03:01 Courthead wrote:Show nested quote +On November 21 2013 02:53 DaRKMaTT3r wrote:On November 21 2013 02:45 Courthead wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? On November 21 2013 02:42 Wingblade wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Exactly this, there's almost no way to determine which races are being chosen, especially for players who are playing for the first time. I originally chose Protoss because they were super advanced aliens and had a guy which was really tall and shot freaking lasers from his eyes(dead serious). Determining a causal relationship between the two isn't possible, because there are a variety of factors which influence the race a player chooses. There is a potential correlation, but even that can be skewed by the potential exodus of Terran players at the end of Wings of Liberty because of broodlord/infestor. The bolded statement is something that I would love to be able to test. Why do people keep making this tired argument? Racial distributions are basically even across ladder. It's 31% Zerg, 31% Protoss, 30% Terran, and 8% Random. There is basically no difference in the number of people playing each race.There was no mass exodus of Terrans. That's total bs. And if you're planning to advance some conspiracy theory like "a lot of pros switched", then you'll need to provide some actual evidence, otherwise it's meaningless. That number of terran mostly comes from bronze league, where terrans are majority. Every other league and region terran has far less than other 2 races. Check for yourself: Sc2Ranks race Stats I don't think the difference is statistically significant. Even if it was, and we decided to erase Bronze league from the equation, the distribution discrepancies you see at the top of the ladder still show imbalance. Look at the process of going from Diamond to Master to GM: Diamond: 29% T, 34% P, 37% Z Master: 27% T, 37% P, 36% Z GM: 24% T, 42% P, 34% Z So Zerg stays relatively constant, but Protoss begins to dominate and Terrans begin to fall off. The difference is even more pronounced on the NA server, where Diamond is 30% of Terrans but GM is down to 21%. In a balanced game, there would be similar levels of success regardless of race. No matter how you look at it, the fact is that the race you choose significantly affects your chances of advancing into the higher leagues.
Not sure where you have your stats from, but that's what I get from sc2ranks.com
Bronze: 33.29%
Silver: 28.33%
Gold: 26.75%
Platinum: 27.70%
Diamond: 27.59%
Master: 26.49%
Grandmaster: 23.77%
source:
http://www.sc2ranks.com/stats/race/hots/global/1v1/bronze
http://www.sc2ranks.com/stats/race/hots/global/1v1/silver
http://www.sc2ranks.com/stats/race/hots/global/1v1/gold
http://www.sc2ranks.com/stats/race/hots/global/1v1/platinum
http://www.sc2ranks.com/stats/race/hots/global/1v1/diamond
http://www.sc2ranks.com/stats/race/hots/global/1v1/master
http://www.sc2ranks.com/stats/race/hots/global/1v1/grandmaster
Looks like Terran is overproportionally* represented in Bronze and underproportionally* in GM, while from silver to master the distribution does not change apart from a little statistical noise.
|
On November 21 2013 04:12 Big J wrote:Show nested quote +On November 21 2013 03:01 Courthead wrote:On November 21 2013 02:53 DaRKMaTT3r wrote:On November 21 2013 02:45 Courthead wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? On November 21 2013 02:42 Wingblade wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Exactly this, there's almost no way to determine which races are being chosen, especially for players who are playing for the first time. I originally chose Protoss because they were super advanced aliens and had a guy which was really tall and shot freaking lasers from his eyes(dead serious). Determining a causal relationship between the two isn't possible, because there are a variety of factors which influence the race a player chooses. There is a potential correlation, but even that can be skewed by the potential exodus of Terran players at the end of Wings of Liberty because of broodlord/infestor. The bolded statement is something that I would love to be able to test. Why do people keep making this tired argument? Racial distributions are basically even across ladder. It's 31% Zerg, 31% Protoss, 30% Terran, and 8% Random. There is basically no difference in the number of people playing each race.There was no mass exodus of Terrans. That's total bs. And if you're planning to advance some conspiracy theory like "a lot of pros switched", then you'll need to provide some actual evidence, otherwise it's meaningless. That number of terran mostly comes from bronze league, where terrans are majority. Every other league and region terran has far less than other 2 races. Check for yourself: Sc2Ranks race Stats I don't think the difference is statistically significant. Even if it was, and we decided to erase Bronze league from the equation, the distribution discrepancies you see at the top of the ladder still show imbalance. Look at the process of going from Diamond to Master to GM: Diamond: 29% T, 34% P, 37% Z Master: 27% T, 37% P, 36% Z GM: 24% T, 42% P, 34% Z So Zerg stays relatively constant, but Protoss begins to dominate and Terrans begin to fall off. The difference is even more pronounced on the NA server, where Diamond is 30% of Terrans but GM is down to 21%. In a balanced game, there would be similar levels of success regardless of race. No matter how you look at it, the fact is that the race you choose significantly affects your chances of advancing into the higher leagues. Not sure where you have your stats from, but that's what I get from sc2ranks.com Bronze: 33.29% Silver: 28.33%
Gold: 26.75%
Platinum: 27.70%
Diamond: 27.59%
Master: 26.49%Grandmaster: 23.77% source: http://www.sc2ranks.com/stats/race/hots/global/1v1/bronzehttp://www.sc2ranks.com/stats/race/hots/global/1v1/silverhttp://www.sc2ranks.com/stats/race/hots/global/1v1/goldhttp://www.sc2ranks.com/stats/race/hots/global/1v1/platinumhttp://www.sc2ranks.com/stats/race/hots/global/1v1/diamondhttp://www.sc2ranks.com/stats/race/hots/global/1v1/masterhttp://www.sc2ranks.com/stats/race/hots/global/1v1/grandmasterLooks like Terran is overproportionally* represented in Bronze and underproportionally* in GM, while from silver to master the distribution does not change apart from a little statistical noise.
Your looking at the "wrong" data. Instead, you should look at the data which lets you compare whether terran is underrepesented in a given league relative to zerg/protoss. You can gather this data from the below link;
http://www.sc2ranks.com/stats/race
next thing you need to do is calculate the percentage of race x players in a given league divided by the total amount of active race x players.
By doing that we can calculate the global terran league distribution;
GM: 0.4%
Master: 2.7%
Dia: 4.9%
Plat: 7.6%
Gold: 17.7%
Silver: 27.1%
Bronze: 39.5%
Now compare that to the distribution for zerg;
GM: 0.6% --> overepresented relative to terran
Master: 3.4% --> overepresented relative to terran
Dia: 6.1% --> overepresented relative to terran
Plat: 9.5% --> overepresented relative to terran
Gold: 23% --> overepresented relative to terran
Silver: 29.2% --> marginally overepresentedd relative to terran
Bronze: 28.1% --> Very underrepesented relative to terran
Now if you read my above post, I argued that while there can be bias'es which can impact the distribution we should see in a balanced game. But today - if the game is balanced - the distribution should be relatively equal across leagues. But since they aren't, it is a clear indication that terran simply is underpowered.
|
Btw, while Blizzard does calculate and adjusted skill-based W/R, chances are that it doesn't look at the distribution in a proper way. Look at the below video from October 2010;
http://www.teamliquid.net/forum/viewmessage.php?topic_id=250789
Notice that when we go from unadjusted to adjusted, these changes occur;
- TvZ w/r is unchanged
- PvZ goes up by 2% in favor of protoss
- PvT w/r goes up by 7% in favor of protoss
So in the adjustment data, the skill-indicator goes up alot for terran in TvP. In PvZ it goes marginally up for zerg. So basically Blizzard believes protoss raw ladder data didn't fully take into account how strong the race actually was.
But how on earth can they get to that conclusion. The ladder distribution was similar to that of protoss and terran right? Further, the fact that there were a lot fewer zerg players and that it is likely that there was some bias amongst lesser skilled players to pick terran over zerg, wasn't taken into account, since the skill-adjustmented in TvZ was unchanged.
The only skill-adjustmented proces I can see that makes a minimal amount of sense is for blizzard to assume that those who play the game more should be weaker than those who practice less. So if a terran player has played 10K minutes and a protoss player has played 8K minutes, and they were of equal skill when they started playing the game and still are equally skilled today --> Protoss is OP.
That's, however, a much less efficient adjustment proces than just relying on an estimated w/r based on "the balanced ladder distribution".
|
![[image loading]](http://i.imgur.com/TWhk2Lx.png)
![[image loading]](http://i.imgur.com/V6VNQA9.png)
![[image loading]](http://i.imgur.com/NB6F4L8.png)
![[image loading]](http://i.imgur.com/gY69QJs.png)
![[image loading]](http://i.imgur.com/XvKmOzJ.png)
David Kim:"Ladder is balanced and we are going to nerf Terran again."
|
On November 21 2013 04:23 Hider wrote:Show nested quote +On November 21 2013 04:12 Big J wrote:On November 21 2013 03:01 Courthead wrote:On November 21 2013 02:53 DaRKMaTT3r wrote:On November 21 2013 02:45 Courthead wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? On November 21 2013 02:42 Wingblade wrote:On November 21 2013 02:38 Big J wrote:Sorry, but it's extremely obvious that distribution matters. Pretending it doesn't just because you don't like the results degrades the quality of the discussion in this thread. The distributions themselve are worthless. People keep on saying that "winrates aren't everything", the plain race distributions are much worse, since we don't know the reasons why people play which race at what level. E.g. when the game came out the (overall summed up) distributions were something like 38-37-25 (P-T-Z). So initially people did not choose races 33-33-33. So was Protoss initially overpowered? Far from! Most likely there were other reasons why people chose Protoss and Terran over Zerg and same thing is probably true right now. Of course racial strength/balance is most likely a reason to play a race. But we don't know how important it is. More likely I'd say (which is of course connected to balance) it's a question of gameplay as well. Do you like how you have to play the matchups of a race? Being the aggressor or the defender? Having cheeses/tricks or rather play straight up everygame? Exactly this, there's almost no way to determine which races are being chosen, especially for players who are playing for the first time. I originally chose Protoss because they were super advanced aliens and had a guy which was really tall and shot freaking lasers from his eyes(dead serious). Determining a causal relationship between the two isn't possible, because there are a variety of factors which influence the race a player chooses. There is a potential correlation, but even that can be skewed by the potential exodus of Terran players at the end of Wings of Liberty because of broodlord/infestor. The bolded statement is something that I would love to be able to test. Why do people keep making this tired argument? Racial distributions are basically even across ladder. It's 31% Zerg, 31% Protoss, 30% Terran, and 8% Random. There is basically no difference in the number of people playing each race.There was no mass exodus of Terrans. That's total bs. And if you're planning to advance some conspiracy theory like "a lot of pros switched", then you'll need to provide some actual evidence, otherwise it's meaningless. That number of terran mostly comes from bronze league, where terrans are majority. Every other league and region terran has far less than other 2 races. Check for yourself: Sc2Ranks race Stats I don't think the difference is statistically significant. Even if it was, and we decided to erase Bronze league from the equation, the distribution discrepancies you see at the top of the ladder still show imbalance. Look at the process of going from Diamond to Master to GM: Diamond: 29% T, 34% P, 37% Z Master: 27% T, 37% P, 36% Z GM: 24% T, 42% P, 34% Z So Zerg stays relatively constant, but Protoss begins to dominate and Terrans begin to fall off. The difference is even more pronounced on the NA server, where Diamond is 30% of Terrans but GM is down to 21%. In a balanced game, there would be similar levels of success regardless of race. No matter how you look at it, the fact is that the race you choose significantly affects your chances of advancing into the higher leagues. Not sure where you have your stats from, but that's what I get from sc2ranks.com Bronze: 33.29% Silver: 28.33%
Gold: 26.75%
Platinum: 27.70%
Diamond: 27.59%
Master: 26.49%Grandmaster: 23.77% source: http://www.sc2ranks.com/stats/race/hots/global/1v1/bronzehttp://www.sc2ranks.com/stats/race/hots/global/1v1/silverhttp://www.sc2ranks.com/stats/race/hots/global/1v1/goldhttp://www.sc2ranks.com/stats/race/hots/global/1v1/platinumhttp://www.sc2ranks.com/stats/race/hots/global/1v1/diamondhttp://www.sc2ranks.com/stats/race/hots/global/1v1/masterhttp://www.sc2ranks.com/stats/race/hots/global/1v1/grandmasterLooks like Terran is overproportionally* represented in Bronze and underproportionally* in GM, while from silver to master the distribution does not change apart from a little statistical noise. Your looking at the "wrong" data. Instead, you should look at the data which lets you compare whether terran is underrepesented in a given league relative to zerg/protoss. You can gather this data from the below link; http://www.sc2ranks.com/stats/racenext thing you need to do is calculate the percentage of race x players in a given league divided by the total amount of active race x players. By doing that we can calculate the global terran league distribution;GM: 0.4% Master: 2.7% Dia: 4.9% Plat: 7.6% Gold: 17.7% Silver: 27.1% Bronze: 39.5%
Now compare that to the distribution for zerg;GM: 0.6% --> overepresented relative to terran Master: 3.4% --> overepresented relative to terran Dia: 6.1% --> overepresented relative to terran Plat: 9.5% --> overepresented relative to terran Gold: 23% --> overepresented relative to terran Silver: 29.2% --> marginally overepresentedd relative to terran Bronze: 28.1% --> Very underrepesented relative to terran Now if you read my above post, I argued that while there can be bias'es which can impact the distribution we should see in a balanced game. But today - if the game is balanced - the distribution should be relatively equal across leagues. But since they aren't, it is a clear indication that terran simply is underpowered.
The distributions are somewhat equal as I marked green in the above post, if you use a (outlier robust) median estimator for the distribution. And it says that Terran is only played by ~27% of the (for their skilllevel) competitively playing players.
The rest could be explained with statistical noise and certain impacts, e.g. theories like the one from your other post (more Terrans in bronze due to WoL campaign. so we now have overproportionally many terrans in bronze)
What you do is not a robust (i.e. it can be heavily skewed by outliers) estimator and that's exactly the reason why you get those numbers. You have two numbers that look like outliers (Zerg negative outlier in bronze; Terran positive outlier in bronze) that lead to this effect. If you could explain those differently than balance (and history tells us that balance unspecifically it has always been like that for those numbers, Terrans overrepresented in bronze, Zerg underrepresented in bronze. So why blame balance for that?) so they get more in line with the other numbers, in your methode the numbers you calculated would correspond as well.
Also in general it is hard to explain the players at the bottom. Since the skillspectrum is much, much broader there. And I don't think it makes for a good argument for balance to begin with, since getting your basic mechanics up will lead you out of bronze regardless of balance.
|
|
|
|
|
|
EPT
