The reason teams can be so dominant in team games regardless of mechanics is because executing strategies with 5+ people, even if the game is easy for individuals, is much more difficult.
On the "Artosis Curse", Luck and Liquibet in SC2. - Page 15
Forum Index > SC2 General |
TheNthMemory
United States17 Posts
The reason teams can be so dominant in team games regardless of mechanics is because executing strategies with 5+ people, even if the game is easy for individuals, is much more difficult. | ||
Highways
Australia6098 Posts
| ||
Khai
Australia551 Posts
Why is it that in physical sports like tennis we see the top players with far superior winrates? Easy, unlike sc2 which has simple mechanics which pretty much most average people can get to a pro-level-skill close to the skill ceiling of the game tennis is different in that if you lack the physical talent, strength or stamina you pretty much have no chance while in sc2 150 apm can easily beat 300 apm. | ||
iky43210
United States2099 Posts
On February 23 2013 11:03 TheNthMemory wrote: It's a 1v1 game with relatively easy mechanics, unlike Brood War or to go on an extreme tangent, tennis. Golf is a little more volatile but you still see consistent winners. The reason teams can be so dominant in team games regardless of mechanics is because executing strategies with 5+ people, even if the game is easy for individuals, is much more difficult. you might have a point except broodwar pros don't have any better winrates than most sc2 pros. Even the very best of the best (flash) have only ever achieved 75% winrates while most others averages out 60%. And team games are not as dominant as you think. If you bother to break down team win statistics on games like CS, dota or LoL I am willing to bet they are actually much worse than sc2 given the ban features and very volatile gameplay | ||
iky43210
United States2099 Posts
On February 23 2013 12:15 Khai wrote: I think a lot of people don't realise that actual game winrates at the top of the top are a lot closer to 50%, in practice most of the players would have very close to 50% winrates, on the KR ladder 70% winrates are reserved for the top of the GM ladder. Basically this means 2 things imo, 1 skill gap at the top is not too large and 2 sc2 is pretty luck based and naturally makes winrates go closer to 50:50 due to things like build order choices being somewhat like scissor paper rock. Why is it that in physical sports like tennis we see the top players with far superior winrates? Easy, unlike sc2 which has simple mechanics which pretty much most average people can get to a pro-level-skill close to the skill ceiling of the game tennis is different in that if you lack the physical talent, strength or stamina you pretty much have no chance while in sc2 150 apm can easily beat 300 apm. 50:50 winrates is because that is how Blizzard design the ladder. If you aren't getting 50:50, then the system is not doing its job correctly. There are alot of sc2 pros, so even the ladder at the very top are still very packed and tiered with those that are better than the one below them. Only in a dying game without fresh blood or one with inactive ladder would you have a ladder where there is a huge gap for those at top (outside of a 1 or 2 that are exceptionally at the game). you can improve 10 folds from where you're standing and your resulting ladder stat should be 50:50. Example: you beat clide 10 times in a row because you are much better player than he is, skyrocket your winrates. Then you get push to MVP's tier and get stomp back down to 50% Again people kept talking about how they feel the game is "luck based" and very volatile, but facts just don't support it. Why do we always see the same face in code S, why do Korean keep stomping foreigners, and why is it always the same few players that keep appearing in GSL ro8, round 4, and finalist? All those players must either be extremely lucky or they deserve to be there through skill and dedication | ||
Cele
Germany4012 Posts
| ||
theinfamousone
United States103 Posts
The reasons tennis is so easy to predict (ie Serena Williams winning everything for years) is that between the matches, set, and games, you have to consistently beat someone over and over, even if by a little to actually win. The luck ends up equal, if you had 2 SC2 players play best of 39s or something, I'd imagine it would be a pretty good indication of who the better player is just like tennis. Much better than best of 3 which is usually the case. If you had tennis players go first to 30 points (two scores), it'd be a miracle to get past 60% win rates for anyone. | ||
IcedBacon
Canada906 Posts
| ||
achan1058
1091 Posts
On February 23 2013 13:10 Cele wrote: skill ceiling isnt high enough, thats why koreans still own in Broodwar. Nough said How about Kim Carrier? | ||
Dagan159
United States203 Posts
On February 23 2013 14:09 theinfamousone wrote: If you're 5% better than someone, that means that you should win about 11 games out of 20. Not 20 out of 20. There's too many ways to lose including some luck, but even tennis has luck. Theoretically if you play enough games, luck would balance out. The reasons tennis is so easy to predict (ie Serena Williams winning everything for years) is that between the matches, set, and games, you have to consistently beat someone over and over, even if by a little to actually win. The luck ends up equal, if you had 2 SC2 players play best of 39s or something, I'd imagine it would be a pretty good indication of who the better player is just like tennis. Much better than best of 3 which is usually the case. If you had tennis players go first to 30 points (two scores), it'd be a miracle to get past 60% win rates for anyone. Pretty much this, In tennis you can think of each point as a seperate event, so a set has at least 7x4 = 21 events, and a match 63 events, all theoretically independent (not really; because of serves and streaks and mentality etc, but close enough) however starcraft usually boils down to fewer points per match, there are probably 5-6 in a game, and these are definelty not independent ( much easier to block his 3rd expo if you successfully predictively defended his banshees, much harder if you diddnt.) The problem with Starcraft if it takes much longer for events to play out, and there is much more time in between events. However the more balanced the game gets (and blizz is doing a great job in my eyes), and the more players begin to engage multiple times in a game, the less "random" the games will get, resulting in higher winrates for the top players. If Hots is to WoL as BW was to SC, hold onto your panties. | ||
IMSmooth
United States679 Posts
On February 23 2013 13:10 Cele wrote: skill ceiling isnt high enough, thats why koreans still own in Broodwar. Nough said Are Koreans not dominating in SC2? I know they were absolutely untouchable in BW but wasn't that due more to the fact there was almost zero legitimate teams for foreigners ? | ||
Mozzery
United States140 Posts
In hockey and basketball, sports with half the sample size of a baseball season, you see the same thing, teams with 60% win rates being contenders for championships. Now what does this have to do with your original thought? Simple, if a team is thought of as the best team and it only acheives its intended result around 60% of the time, does that mean that skill is not involved? Of course not. | ||
TheNthMemory
United States17 Posts
On February 23 2013 15:12 Mozzery wrote: Here's something I think you are missing, over a large sample size, 64% is AMAZING as a winrate in the majority of professional sports. Here's a good example. Let's take a look at what is universally considered one of the dominant sports franchises in America, the New York Yankees. They play 162 game seasons every year so they have a large sample size. In 2009 they had the best record in all of baseball and won the world series, (for those who don't know that's the championship for baseball). They had a 63.6 winning percentage over the course of that season. In baseball just about every season there are teams with under a 60% win rate that win their division consisting of 5 teams. In hockey and basketball, sports with half the sample size of a baseball season, you see the same thing, teams with 60% win rates being contenders for championships. Now what does this have to do with your original thought? Simple, if a team is thought of as the best team and it only acheives its intended result around 60% of the time, does that mean that skill is not involved? Of course not. American professional sports have amateur drafts and some kind of salary cap (baseball's is a soft cap which means half the teams almost never win a World Series) so all the teams are relatively equal. In a 1v1 sport like tennis you see players like Roger Federer win nearly everything. An esport with a high mechanical skill cap would be similar. | ||
julianto
2292 Posts
On February 23 2013 13:10 Cele wrote: skill ceiling isnt high enough, thats why koreans still own in Broodwar. Nough said The battles as of now are too fast. The more time a player is allowed, the better they'll micro, and so the higher the skill ceiling. | ||
Cuce
Turkey1127 Posts
best players are best players when they play their best, on his avarage day best players is no better than a less good players prime. | ||
Cuce
Turkey1127 Posts
| ||
Kaeru
Sweden552 Posts
| ||
SpaceSaKe
United States2 Posts
But starcraft unlike basketball has a constantly evolving meta game, build order loses and what i believe is a much more muscle memory demanding sport (not to say basketball isn't but take a week of from sc2 and ur trash again. take a week off from basketball and ur likely the same) gives starcraft a wider margin for seemingly weaker players to win games from stronger players. Competition is good the games are exciting and the best player that season normally wins these days. WE can all agree nestea and mvp were not the best this season and that LIFE maybe the best player in the season lost to mind-games and a very strong player. SO its all good. | ||
Wrathsc2
United States2025 Posts
On February 23 2013 16:12 SpaceSaKe wrote: There is no curse. The real reason players who seem the strongest lose is the same reason the #1 ranked team in pro basketball loses all the time. The reason is because the margin that separates the top 16 teams in the world or the top 100 players in sc2 is such a small amount that even a week of bad practice lack of sleep or any one of a plethora of small events could lead to a seemingly stronger player to lose to a weaker one. But starcraft unlike basketball has a constantly evolving meta game, build order loses and what i believe is a much more muscle memory demanding sport (not to say basketball isn't but take a week of from sc2 and ur trash again. take a week off from basketball and ur likely the same) gives starcraft a wider margin for seemingly weaker players to win games from stronger players. Competition is good the games are exciting and the best player that season normally wins these days. WE can all agree nestea and mvp were not the best this season and that LIFE maybe the best player in the season lost to mind-games and a very strong player. SO its all good. People dont seems to understand that in a game like sc2 one wrong thought can ruin an entire game for a player. Its literally close to impossible to play consistent in a game like this for more then 3 months with the metagame changes. | ||
iky43210
United States2099 Posts
On February 23 2013 15:20 TheNthMemory wrote: American professional sports have amateur drafts and some kind of salary cap (baseball's is a soft cap which means half the teams almost never win a World Series) so all the teams are relatively equal. In a 1v1 sport like tennis you see players like Roger Federer win nearly everything. An esport with a high mechanical skill cap would be similar. Lets put things into perspective. Tiger Wood: PGA events played = 283; PGA wins = 73; win percentage = 25.8% Immvp: GSL played = 17; GSL wins = 4; win percentage = 23.5% Immvp is literally the tiger woods of sc2. Yes yes small sample size, but just illustrating a point about the illusion of winrates Btw, here's roger federer total statistics Total 84 39 123 0.68 (68%) Its not the same winrates as used in sc2, as it is the win rates he have in the final. But if you really want to have an exact conversion, count up the points he scored vs the one he lost over the course of a season. Don't expect anything more than 75% | ||
| ||