July TLPD Win Rates - Page 9

Plansix

United States60190 Posts

August 08 2012 16:15 GMT

#161

On August 09 2012 01:13 Pazuzu wrote:

earlier someone crunched the numbers on what would happen if you took out Taeja (10-1) and Gumiho (12-3). result was a terran win rate of 47%

Thats really not that bad and as close to 50/50 at most winrates are going to get.

Lorch

Germany3703 Posts

August 08 2012 16:16 GMT

#162

Reflects the skill level of foreigner vs korean pretty well.
But the game looks really balanced in korea, even tvz, so I think it's fine, just give foreigners another month to adapt.

NHY

1013 Posts

August 08 2012 16:19 GMT

#163

On August 09 2012 01:12 pmp10 wrote:

Show nested quote +

Hp maybe but as long as you have buildings that's build once per game anywhere from bronze league to code S then chances are it's cost will change a race balance irrelevant to the players skill level.

Let's just say we give zerg players extra 50 mineral at the beginning. We can agree that covers everyone from bronze league to code S. But that would also affect balance differently across skill level.

Edit: And average number of spawning pools built per game would not be same across the skill level.

Greenei

Germany1754 Posts

August 08 2012 16:25 GMT

#164

It's funny how the general consensus seems to be that "herp derp foreigner terran just need to be better" while also saying "korean terrans have the same skill as korean toss/zerg".

Why wouldn't it be possible to at least try to balance every level? Just make Toss and Zerg not so easy, it has been 3 month since Terran is the worst race internationally, I think it's time for a change.

Irre

United States646 Posts

August 08 2012 16:29 GMT

#165

The balance changes shift the metagame pretty hard and have certainly made terran either be punished very severely for a few tiny mistakes or forced to have their early/mid game do lots more damage or go into the first scenario, but the issue that every terran complains about comes down to design more than anything. Sc2 design for the different races is too different in terms of units. Its a very "this counters this so build this" game at this point. Except terrans first units are so good that blizzard had to make all the rest kind of gimmicky or not scale well. This leaves once you get to a point in the game where P and Z can build (A, B, C, D units and mix them very powerfully), terran is stuck with much less to work with. Then you bring in how sc2 design is for TERRIBLE TERRIBLE DAMAGE, and how terran got the short end of that stick as well outside of stimpack, and there in lies the issues. Its what we all have to live with, and people just want to ignore that or shrug it off in the name of keeping our esport we love alive. I can't say I blame them for that, not to mention if Z and P were designed similarly to T, I don't think you'd ever see the foreign scene we have now.

hunts

United States2113 Posts

August 08 2012 16:31 GMT

#166

Maybe now Ts will stop complaining about zerg after seeing it's balanced? Nah who am I kidding people will always cry imbalance even when there is none.

Plansix

United States60190 Posts

August 08 2012 16:32 GMT

#167

On August 09 2012 01:25 Greenei wrote:
It's funny how the general consensus seems to be that "herp derp foreigner terran just need to be better" while also saying "korean terrans have the same skill as korean toss/zerg".

Why wouldn't it be possible to at least try to balance every level? Just make Toss and Zerg not so easy, it has been 3 month since Terran is the worst race internationally, I think it's time for a change.

Why is the game imbalanced for "other levels" if one region has a nearly 50/50 win rate and another does not? And if the region with the 50/50 win rate is also considered to have the most advanced metagame, why should the game be rebalanced? Why do you assume the game isn't balanced for your skill level?

AKomrade

United States582 Posts

August 08 2012 16:32 GMT

#168

The advantage of preparation is fairly obvious in the Korean TLPD vs. the "quick match" foreign tournaments. Skill difference between Korea and Western T still pretty big though =/

XiWi

11 Posts

August 08 2012 16:36 GMT

#169

nvm recalculating messed up the samples
apparently they are separate so nevermind

NHY

1013 Posts

August 08 2012 16:37 GMT

#170

On August 09 2012 01:36 XiWi wrote:
Hey all if the international games included the korea ones ( korea sample size ~2000 international 10k) I did some math and excluded the korean from the international ones and I got that international TvZ (w/o korea) is at ~42.77%

~~They are separate.~~
I was wrong. The chart says they are included.

canikizu

4860 Posts

August 08 2012 16:38 GMT

#171

On August 09 2012 01:13 lichter wrote:

Show nested quote +

On August 09 2012 01:04 canikizu wrote:

On August 09 2012 00:18 lichter wrote:

On August 09 2012 00:13 canikizu wrote:

On August 09 2012 00:00 Pazuzu wrote:

On August 08 2012 23:55 canikizu wrote:

On August 08 2012 23:45 ELA wrote:
KR TvZ winrate explained:

[image loading]

July games v. Z: 10 wins, 1 loss (90.91% winrate)

You forgot Gumiho vs Zerg 12-3 (80% winrate)

These two alone skew the Terran winrate like 10% or something.

2 players out of how many in korea? skew a winrate 10%? sorry but the math nowhere near checks out

there're 321 Korean games, so essentially 3 games ~ 1%
if Taeja and Gumiho only won half of their game (Taeja lose 4 more, Gumiho lose 4 more), that were ~ 3% more for Zerg and 3% less for Terran, and ZvT in Korea would be 53%-47%.
Oh hey, that's not nice anymore.

You can't pick stats like this, because it becomes meaningless.

If you suggest that, then what about removing the two Zergs with the best ZvT as well?

And then the cycle continues. Really bad terrans or really bad zergs skewing the numbers. Meaningless arguments because removing all outliers will leave you with the median or mean, more or less. Back to square one.

Yes you can pick stats like that. That's the very basic fundamental when you look at chart and statistic. If you trade stock and stuff, and you don't know the source of why the that stock is flying so high, you will get perceived and get burn.

The reasons Taeja and Gumiho are counted as outliners because they contributed a significant number of games in TvZ, and have siginicant higher win rate than other Terran.
Now compared to Zerg players, we have Leenock, Curious, Symbol (100% winrate), Horror (80%), Shine, Life, Sniper, Losira, DRG, Hyun (60%-70%). There are too many players that have good winrate against Terran, and most of them didn't contribue significant number of games to count as outliners. Among the Zerg there're only 2 players that contribute good number of games vT, such as Coca (14-5 73% winrate), and Nestea (4-9, 30%). But comparing between Coca's winnrate and others, he doesn't have any significant winrate that can be counted as outliner. On the other hand, Nestea can be counted as outliner because his winrrate was significantly lower than the rest of the Zerg. But that just proved that if you took Nestea out, ZvT winrate should even be higher.

I rest my case.

I don't care if it is skewed for T or for Z.

I am against this kind of approach at looking at the numbers.

In this month's case, it just happens that there are two T outliers that contribute greatly to their win percentage, mostly because they played more games and the sample size is so small that their games affect the overall numbers so much. Fortunately for that person's argument, there weren't any Z outliers, as you showed. However these are numbers for a SINGLE month, with a sample size so small one person can affect it by a significant percentage.

Larger sample size, more months (within the same patch), either outliers will disappear or other outliers will appear because they will be able to play the matchup. They will cancel each other out if the game is near balance.

I don't see how the presence of these outliers is supposed to mean that the numbers are "wrong" or "skewed". Some people are good. Some people suck. That's caused by talent, not balance.

Same stuffs happened last month and I did the same stuff last month too. And same stuff is gonna happen next month, and I might have to do the same stuff next month too.
I don't care about the number either. All I care about is the moving average on larger picture. I ran numbers to annoy people who use numbers to claim TvZ is not imbalance lol.

pmp10

3399 Posts

August 08 2012 16:39 GMT

#172

On August 09 2012 01:32 Plansix wrote:

Show nested quote +

Because the SC2 viewer base consists of foreigners and race distribution at foreign tournaments is important for the game success.
This has no meaning in trying to look at things in 2-3 months of meta-game shifts but the overall long-term trends are important.

eFonSG

United States255 Posts

August 08 2012 16:40 GMT

#173

The game should be very very balanced once Z players learn to stop this new wave of Toss 2 base all-ins. Go Balance

RandomAccount#106470

Lithuania1506 Posts

August 08 2012 16:42 GMT

#174

--- Nuked ---

IshinShishi

Japan6156 Posts

August 08 2012 16:43 GMT

#175

In the last three months zerg has won 256 more games (1430-1174) than terran (considering that this data is correct) in the entire pro scene, and didn't they win like 8 times more tournaments than terran in the last few months?Like 24 to 3 or something like that, I can't seem to find the image, but the reason is simple, it's just that the talented players aren't picking terran, they always choose zerg or protoss and those that stick to terran are never good enough, that's why for god knows how many seasons terran has been underrepresented in masters and GM around the world, only the bad players pick terran, but it's all fine and dandy, soon we will learn how to macro even better like Idra suggested and we will have many JonnyReccos and Sortofs.

ELA

Denmark4608 Posts

August 08 2012 16:45 GMT

#176

On August 09 2012 01:31 hunts wrote:
Maybe now Ts will stop complaining about zerg after seeing it's balanced? Nah who am I kidding people will always cry imbalance even when there is none.

Funny, find any Z's in here calling T's whiners, check their post history and you will most likely find something like this:

On September 10 2011 09:12 hunts wrote:
neural change is quite possibly the dumbest thing I've seen from blizzard yet. Yeah infstor is a bit too good of a unit as is, but it's basicly the only thing actually keeping Z viable as a race. If they want to really nerf ifnestors, they need to buff zerg elsewhere to keep it viable, but they aren't doing that. They nerfed fungal with no compensation, but that was fine because it wasn't a huge nerf. But now they are doing a huge nerf, and still no compensation. The overseer buff while nice still in no way helps the actual Z army. And while the ultra buff is better than nothing, it still doesn't help the actual zerg army because honeslty, ultras are still useless and will die before getting to their targets, even if they are fungald.

Keep it classy

Greenei

Germany1754 Posts

August 08 2012 16:46 GMT

#177

On August 09 2012 01:32 Plansix wrote:

Show nested quote +

Because the game makes more fun if the game is blanced. 99% of people who play this game are out of the balanced region. I can't see why they shouldn't matter at all.

ShamW0W

160 Posts

August 08 2012 16:51 GMT

#178

It's comical how many logical fallacies you find in this thread.

On topic, the Korean data is definitely interesting. On the ladder it feels like Terrans have started to really adapt to the new Zerg style of play and are more aggressive with their drops and removal of creep from the map.

Watching some of the top players actually shows you the potential each race has. It's up to us as players to unlock that potential, not up to Blizzard to make the potential easier to obtain.

Plansix

United States60190 Posts

August 08 2012 16:52 GMT

#179

On August 09 2012 01:46 Greenei wrote:

Show nested quote +

Because the game makes more fun if the game is blanced. 99% of people who play this game are out of the balanced region. I can't see why they shouldn't matter at all.

I am having a great time and I am still getting rocked by zerg, protoss and terran's alike. Have you ever considered that you might now be playing a paticular match up correctly? Or that your winrates are not that bad against people of the same skill lever? Why are your losses to "less skilled players" not the results of some bad decision you made, rather than the fault of some balance issue?

Doublemint

Austria8760 Posts

August 08 2012 16:52 GMT

#180

On August 09 2012 01:25 Greenei wrote:
Why wouldn't it be possible to at least try to balance every level? Just make Toss and Zerg not so easy, it has been 3 month since Terran is the worst race internationally, I think it's time for a change.

Just for answers like this one I like to come to these monthly threads :D

Apart from foreigner Ts the numbers seem quite ok.

Curious how long it will take blizzard, even after the beta, to come close to these numbers in HotS^^

Prev 1 7 8 9 10 11 23 Next All

Please or register to reply.

July TLPD Win Rates - Page 9

Completed

Ongoing

Upcoming