|
United States5217 Posts
On March 07 2016 03:50 BretZ wrote:
Honestly just want a new map pool before too much balance discussion gets going. I think new maps could do wonders for P
I'm going to go on the record right now and say that maps alone will not solve PvZ unless we get a bunch of maps like Dusk Towers with free backdoor expansions and easily defendable 3rd and 4th bases. And that is very constrictive to map design.
Time will tell who is right.
On March 07 2016 03:51 SNSeigifried wrote:Show nested quote +On March 07 2016 03:45 BronzeKnee wrote:On March 07 2016 03:39 SNSeigifried wrote:
but i had to limit dreamhack leizpig to only noteables because their were quite a big chunk of players who loss such as Monty/hinO who would hurt the accuracy of the data since the skill gap between them to compared ShowTimE/Snute is to big. So there is the cherry picking. The problem with cherry picking is that you never know who the next Stephano is going to be. Let's say that Monty performs poorly in one tournament, but then goes a massive streak the next tournament and is the next big thing, do you then retroactively add his statistics back into past data because you now realize he is good? That is the qualitative cherry picking I was looking for, and it is flawed. As for the map pool, as someone said, "better" maps can solve any problem. And the data is still so small that just a few series can flip a favored map for one race to a favored map for another for 6 out of 9 (66%) of the maps. Well sorry i can't fix the sample size since their have only been 170 zvp maps played at the premier level of competition on location currently. Also i would rather cherrypick then add data of players who could of been silver since dreamhack leizpig was a massive open tournament that anyone in germany could play in which intern could skew the data incorrectly. 
Stephano could have been silver too when you scan a tournament and see players you don't know and choose to remove them from a dataset.
Bad players of every race join open tournaments and lose to good players of every race. Overtime, that will control for itself. That is one of the benefits of using a large sample size.
|
On March 07 2016 03:54 BronzeKnee wrote:Show nested quote +On March 07 2016 03:50 BretZ wrote:
Honestly just want a new map pool before too much balance discussion gets going. I think new maps could do wonders for P I'm going to go on the record right now and say that maps alone will not solve PvZ unless we get a bunch of maps like Dusk Towers with free backdoor expansions and easily defendable 3rd and 4th bases. And that is very constrictive to map design. Time will tell who is right. Show nested quote +On March 07 2016 03:51 SNSeigifried wrote:On March 07 2016 03:45 BronzeKnee wrote:On March 07 2016 03:39 SNSeigifried wrote:
but i had to limit dreamhack leizpig to only noteables because their were quite a big chunk of players who loss such as Monty/hinO who would hurt the accuracy of the data since the skill gap between them to compared ShowTimE/Snute is to big. So there is the cherry picking. The problem with cherry picking is that you never know who the next Stephano is going to be. Let's say that Monty performs poorly in one tournament, but then goes a massive streak the next tournament and is the next big thing, do you then retroactively add his statistics back into past data because you now realize he is good? That is the qualitative cherry picking I was looking for, and it is flawed. As for the map pool, as someone said, "better" maps can solve any problem. And the data is still so small that just a few series can flip a favored map for one race to a favored map for another for 6 out of 9 (66%) of the maps. Well sorry i can't fix the sample size since their have only been 170 zvp maps played at the premier level of competition on location currently. Also i would rather cherrypick then add data of players who could of been silver since dreamhack leizpig was a massive open tournament that anyone in germany could play in which intern could skew the data incorrectly. Stephano could have been silver too. Bad players of every race join open tournament and lose to good players of every race. Overtime, that will control for itself. That is one of the benefits of using a large sample size.
I can fix the dreamhack data if you want and use every series but everything else is accurate and not cherrypicked 
|
United States5217 Posts
That is up to you since you collected the data. I was just giving my opinion of what you collected since you asked.
Personally, I like data with large sample sizes. It is just so hard to infer much looking alone at the data from Korea when Zerg has a 40% winrate versus Protoss in one tournament and better than a 77% winrate versus Protoss in another.
|
No surprise Protoss is struggling vs Zerg right after the pylon overcharge nerf. Partially I think there needs to be some adjustment, while partially I think they may be just flat out weaker in the MU.
|
On March 07 2016 04:00 BronzeKnee wrote:
That is up to you since you collected the data. I was just giving my opinion of what you collected since you asked.
Personally, I like data with large sample sizes. It is just so hard to infer much looking alone at the data from Korea when Zerg has a 40% winrate versus Protoss in one tournament and better than a 77% winrate versus Protoss in another.
Updated dreamhack to have an all players stat and a noteables.
With all players its actually very close and protoss actually has a map lead but a series deficit:D!!!
|
Its funny that when there was 3 terrans in GSL and terrans went 6 months in a row without a single premier tournament title, toss didnt cared at all.
Now its been only 1 month, and suddenly, we shoud all feel bad for poor protoss?
I agree that toss is UP right now (mostly vs zergs), but the last time they were, it was in WoL...................
I know they dont deserve to be UP because they were broken for the vast majority of Hots, but i still dont feel any sympathy for them
|
Czech Republic12129 Posts
On March 07 2016 04:40 Ensiferum8 wrote:
Its funny that when there was 3 terrans in GSL and terrans went 6 months in a row without a single premier tournament title, toss didnt cared at all.
Now its been only 1 month, and suddenly, we shoud all feel bad for poor protoss?
I agree that toss is UP right now (mostly vs zergs), but the last time they were, it was in WoL...................
I know they dont deserve to be UP because they were broken for the vast majority of Hots, but i still dont feel any sympathy for them
I cared the same way I care about it now. Don't say stupid things. Many Protoss cared. And I haven't seen any Terran writing there "just wait and see" (if we don't count Flash)
|
United States5217 Posts
On March 07 2016 04:40 Ensiferum8 wrote:
Its funny that when there was 3 terrans in GSL and terrans went 6 months in a row without a single premier tournament title, toss didnt cared at all.
Now its been only 1 month, and suddenly, we shoud all feel bad for poor protoss?
It has been longer than 1 month. Look at the chart in the first post. It isn't about feeling bad, it is about wanting to play a competitive game.
And frankly balance has nothing to do with who wins what tournament. FruitDealer won the first GSL, but Zerg was horrendously underpowered. Nestea won the second GSL, Zerg still was underpowered. A tournament finals is nothing more than another series when it comes to statistics.
|
United Kingdom20285 Posts
3 races, equal distribution on the top 10
That's HOTS
|
On March 07 2016 05:24 Cyro wrote:That's HOTS
It includes HotS results but the players are where they are because of LotV results.
|
United Kingdom20285 Posts
Why do they have one labaled WOL, one labeled HOTS and no LOTV?
|
On March 07 2016 03:54 BronzeKnee wrote:
Bad players of every race join open tournaments and lose to good players of every race. Overtime, that will control for itself. That is one of the benefits of using a large sample size.
no
this will depend on the balance on lower skill levels:
if protoss is imba in bronze, the bronze protosses will feel more encouraged to go to tournaments
|
France1887 Posts
it's just that we thought that for TLPD, having a common database would be better this time around.
Having a different database for LotV was having more cons that pros. Mostly it meant additional workload, and separate pages, just for the small advantage of having a LotV database.
Map stats should be good, because LotV has its own maps.
And for most of the stuff, you can filter out by the release date of LotV to get stats.
The only thing that changes is that the Elo is not reset, but resetting also means it puts every player at the same level at the start. Seeing that (in my opinion) Elo already has it flaws, it was an acceptable drawback to start with HotS Elo.
I know it's weird that we have LotV stuff in the "HotS" DB, but it will probably be eventually fixed in some way.
|
On March 07 2016 06:12 neptunusfisk wrote:Show nested quote +On March 07 2016 03:54 BronzeKnee wrote:
Bad players of every race join open tournaments and lose to good players of every race. Overtime, that will control for itself. That is one of the benefits of using a large sample size.
no this will depend on the balance on lower skill levels: if protoss is imba in bronze, the bronze protosses will feel more encouraged to go to tournaments
If protoss is imba in bronze, the bronze protosses are now silver protosses... It's not like you can play a race that is imbalanced in bronze enough to be interested in going to tournaments but not enough to be promoted to silver.
|
Is it possible to download all the aligulac data? It'd be fun to do some more detailed analysis by map and by skill level and that kind of things.
|
With a better map pool everything is gonna be fine. P is not doing half as bad as people seem to think at the highest levels, give the players time to adapt.
|
But given how reluctant they've been with changing the maps, I get the feeling they want to have a good balance on these maps, before going on to other maps. The reasoning being that yes, you can balance through maps, but as seen at the end of Wings and HotS the diversity of maps narrows as they need to achieve balance. This time they want to tweak the balance in favour of balance on as many maps as possible.
|
Protosses just need to get better.
Be more like those manly zergs!
|
United States5217 Posts
On March 07 2016 07:38 [PkF] Wire wrote:
With a better map pool everything is gonna be fine. P is not doing half as bad as people seem to think at the highest levels, give the players time to adapt.
I disagree, but hope you are right. I think Protoss has some systematic flaws with how the race is designed that prevent it from being as strong as Zerg or Terran currently. Maps with an inbase natural mask some of those problems which is why we see a Protoss doing okay versus Zerg on certain maps.
But is the whole map pool going to be maps with in base naturals and easily taken thirds? That doesn't fit with Blizzard's current philosophy went it comes to map design.
But hopefully you are right and Protoss does better with a new map pool.
On March 07 2016 08:02 ejozl wrote:
But given how reluctant they've been with changing the maps, I get the feeling they want to have a good balance on these maps, before going on to other maps. The reasoning being that yes, you can balance through maps, but as seen at the end of Wings and HotS the diversity of maps narrows as they need to achieve balance. This time they want to tweak the balance in favour of balance on as many maps as possible.
I don't think Blizzard plans to do anything with comments like this:
The State of Protoss
Contrary to right after the last patch hit, Protoss didn’t look to be struggling as much... Although there are currently no immediate, glaring issues, let’s continue doing what we can during this time so that we can be prepared for the future.
http://www.teamliquid.net/forum/starcraft-2/504638-community-feedback-update-february-26
|
I think its a game design issue... Protoss seems hardest to play of the 3 races currently. Many of their units require really precise micro in ways that cannot be done within a large control group. For instance adepts and disruptors require multiple clicks to control their spells effectively. Further units like phoenix and oracles (which seems essential to success especially against Z) always have to be controlled separate from the main army. Lastly P cannot use hotkeys to reinforce. They always have to warp-in, which requires numerous mouse clicks and looking away from the main battles!
Z on the other hand can control ravagers, lurkers, hydra, roach within one control group or large army group. I mean it does get tougher with vipers. But properly controlling a Z army requires slightly less precision.
|
|
|
|
|
|
EPT
