| Forum Index > SC2 General |
|
Fencar
United States2694 Posts
On April 06 2012 05:12 darkscream wrote: The korean stats o_O P>Z>T>P definitely not the metagame i expected to see. Yeah, I thought it would be P>T>Z>P, but clearly Terrans figured out how to all-in and exploit timings, and Zergs figured out how to use Ultras, Broodlords, and Infestors properly since the Ghost nerf. | ||
|
Zazzles
Slovakia153 Posts
| ||
|
biology]major
United States2253 Posts
On April 06 2012 08:02 Fencer710 wrote: Yeah, I thought it would be P>T>Z>P, but clearly Terrans figured out how to all-in and exploit timings, and Zergs figured out how to use Ultras, Broodlords, and Infestors properly since the Ghost nerf. all in and exploit timings? maybe they just learned how to play better | ||
|
Plansix
United States60190 Posts
On April 06 2012 08:04 biology]major wrote: all in and exploit timings? maybe they just learned how to play better Exploiting timings is playing better. All ins are key as well. You have to have one or two in your pocket. | ||
|
SovietHammer
United States166 Posts
| ||
|
Antimatterz
United States1010 Posts
Hell, it's about time! | ||
Whitewing
United States7483 Posts
On April 06 2012 07:22 Orracle wrote: You must take into account the crazy low sample size the KR graph holds this month. At 250 games, and no idea how many of games of each match up was played, these graphs can fluctuate greatly. I'm not sure how you can say a sample size of 30 is fine. Do me a favor and look at the international statistics. You see the low variance and the trend lines? That is from a relatively large sample size. Now look at the Korean graph. Still see the same trend? Absolutely not. It fluctuates greatly. Right, which is why you have to include the variance in your interpretation of the data. | ||
|
BoB_KiLLeR
Spain620 Posts
On April 06 2012 06:47 ZeromuS wrote: lol well done with this post. It would be interesting to see how many of the games are from each race's respective top players. If 25% of the T replays from KR are MKP and Maru winning then it kinda throws the stats out of whack doesnt it Actually I was reffering the metagame of ZvP and TvP  | ||
|
hzflank
United Kingdom2991 Posts
On April 06 2012 07:05 Whitewing wrote: No, you are incorrect, and you would know why if you studied statistics. A sample size of greater than 30 is probably good enough. Care to explain why? Would someone who knows this stuff off the top of their head mind telling me how many samples we expect to need to get a SD of less than 1%? | ||
|
sorrowptoss
Canada1431 Posts
| ||
|
Go1den
England116 Posts
| ||
|
blade55555
United States17423 Posts
On April 06 2012 09:58 Go1den wrote: Good to see Protoss still in last by a long shot. Warpgate un-nerf plz now that Zergs are unbreakable early game? really in last by a long shot? 55% win rate pvz doesn't seem last place | ||
|
Let it Raine
Canada1245 Posts
On April 06 2012 09:58 Go1den wrote: Good to see Protoss still in last by a long shot. Warpgate un-nerf plz now that Zergs are unbreakable early game? thats what protoss needs quicker warp gate | ||
|
Shiori
3815 Posts
1) Foreign players aren't so bad that balance doesn't affect their play. The foreign pros are still way better than anyone on ladder (with a couple of exceptions, obviously). If there's a trend of imbalance on the foreign scene, unless it's a serious metagame lag (unlikely given the integration of the scene and the amount of mixed tournaments) it's probably worth considering. 2) International winrates include Korean statistics, and there are plenty of Korean players at international tournaments. It's important to realize that Korean statistics only show Korean tournaments, not games that had a Korean participant. This means they don't include MKP winning MLG against DRG, for exampler. 3) The sample size is freaking tiny. Seriously, 245 games this month? Assuming perfect distribution that means 80 games per matchup. That means 42-38 is indicative of a 52.5% favouring for the first race. One player 2-0ing someone of a different race would be sufficient to fuck up the symmetry. Look at the foreign stats, not the Korean ones. The only conclusion we can draw from these stats is that Zerg is ahead in both PvZ and TvZ, which is more or less what everyone has been saying since the Ghost nerf and 3base Roach timings came to the fore. | ||
|
hzflank
United Kingdom2991 Posts
On April 06 2012 09:59 blade55555 wrote: really in last by a long shot? 55% win rate pvz doesn't seem last place Having followed these threads for six months, people seem to think that anything outside of 52/48 makes the matchup imbalanced. Hence why I think we need bigger sample sizes, because with these samples it will often look 'imbalanced' and there will be a lot of monthly variance. | ||
|
Let it Raine
Canada1245 Posts
On April 06 2012 10:04 Shiori wrote: K. Stop talking about the Korean statistics. They don't matter nearly as much as the international ones. Why? Few reasons: 1) Foreign players aren't so bad that balance doesn't affect their play. The foreign pros are still way better than anyone on ladder (with a couple of exceptions, obviously). If there's a trend of imbalance on the foreign scene, unless it's a serious metagame lag (unlikely given the integration of the scene and the amount of mixed tournaments) it's probably worth considering. 2) International winrates include Korean statistics, and there are plenty of Korean players at international tournaments. It's important to realize that Korean statistics only show Korean tournaments, not games that had a Korean participant. This means they don't include MKP winning MLG against DRG, for exampler. 3) The sample size is freaking tiny. Seriously, 245 games this month? Assuming perfect distribution that means 80 games per matchup. That means 42-38 is indicative of a 52.5% favouring for the first race. One player 2-0ing someone of a different race would be sufficient to fuck up the symmetry. Look at the foreign stats, not the Korean ones. The only conclusion we can draw from these stats is that Zerg is ahead in both PvZ and TvZ, which is more or less what everyone has been saying since the Ghost nerf and 3base Roach timings came to the fore. http://www.teamliquid.net/tlpd/sc2-international/games#tblt-2798-1-1-DESC look at the games making up the stats you think = balance (T)LamBorGhiNi vs (P)BloB in a bo3! thats 2 wins for terran in this months statistics, and 2 losses for protoss! then don't forget games that are pros vs joes or koreans vs foreigner pros aaah, the international stats. so useful! | ||
|
avilo
United States4100 Posts
Surprised about the TvP stats...although I'm also not surprised, because i'm fairly sure Terrans have started to do a lot more all-ins and timing attacks to attempt to take wins before the game hits lategame. Not that the stats actually comment on the metagame much, people shouldn't really point to the stats and say "omg i knew it was balanced/imba." The stats are just a very general idea, it has nothing to do with how the game is actually played in any specific sense at least. I'd say the things you can take from the TvP graphs are that it's the same ol' same ol' - if Terran can damage the protoss with their first drop timing, then they have a chance. If protoss defends that first timing into 3-base the game gets progressively harder for Terran, so the stats stay about even or flip with 5% diff or so. The drastic sudden change in the TvZ graph is most likely because of the ghost nerf. | ||
|
Necro)Phagist(
Canada6660 Posts
But in all seriousness these stats really shouldn't be used in arguments for real balance, between low monthly sample sizes and meta game shifts it's nearly impossible to use them to pinpoint imbalances. | ||
|
T0fuuu
Australia2275 Posts
| ||
|
PlacidPanda
United States246 Posts
On April 06 2012 07:45 chuky500 wrote: Protoss losing is clearly an effect of the phoenix range upgrade. Thank you, this honestly made my day go better Also guys, stop whining, even in BW certain races were stronger points of the game i.e. the rush to defilers in ZvT. Its just something you have to accept, terrans need to accept like toss do in PvZ that the strength of their army is its mobility and not its pure fighting prowess. Abusing mobility allows you not to win on the first or second battle but win in the subsequent ones. | ||
| ||
|
|
OSC
BSL 21
Cross vs Dewalt
Replay Cast
Wardi Open
OSC
Solar vs MaxPax
ByuN vs Krystianer
Spirit vs TBD
OSC
Korean StarCraft League
OSC
OSC
OSC
[ Show More ] |
|
|
EPT
