bad argumentation.
example: on pro lvl, toss have partial problems in late game. but until mid GM, sky toss is strong against zerg.
as a newbie, the games runs like this: from build arms run over, win fight win game ~silver.
1 Like
Even with those win rates and say I was on the lower side of that. You would expect my winrate to be 45%. and then say maybe 10% +/- with in that. say 35%.
I’ve had others try to help and still working with some people… That are higher ranked than me…
I wonder if I can even be helped outside of trying to refine my build order and making it tigther.
Its not really the build orders I have an issue with. I can watch any build order and mimic it. its normally leading up to and / or after that first attack… Whats next.
If it was TvT or TvZ. I know the follow up for each situation.
In toss I wonder what the follow up is for “simple” things like mass I.A.C
This hard counter mech and the longer you stay on bio the more likely you are going to lose… I am looking at my own replays in an attempt to come up with a comletly different strat outside of the norm
Obviously pure mech doesnt work and bio is on a timer.
On another note my PvT has got worst since that 11/2018 toss patch that they love to deny ever happened. but thats another discussion can’t do to much about that now…
1 Like
Yes
Yes, which is why for me it is the Playoff representation. Right now it looks good for Terran, cant judge balance by the one winner.
???
Here you are way wrong, and obv I am not a gsl player… or a GM poster like some lul.
But what you say next is what I’ve said before too - when they tell you L2P - you can improve but bnet matches you same skill so eventually you get the same problems but on higher levels.
You can also add that ‘matches you vs gold protoss while your vs X race MMR is Master’ is nonsense. Dont read such clueless ones here.
Aside of not showing particular game knowledge but the thread being like a side observer, it has some things correct.
Whining daily annoys me too but I feel them with this from personal experience. It is not normal EVERY Terran on ladder to have worst matchup vs Protoss. If it was me vs many others having e.g best mu vs protoss I would say ‘oh it’s just me’.
This was the state of the game in 2018-2019 now I dont have that data but I was looking at Protoss profiles best mu - Terran.
Kind of reminds me of war3 before the recent year changes, in the past Orc felt like Punchbag for human. Always felt Hu’s strongest mu was vs orc. Some even offraced hu to beat orc.
So if Every Terran has worst mu vs Protoss, and Pro players dont - pro players dont because they play vs same few people and if they are INFINITELY better (meaning for example heromarine is always better than Gunfubanda or Geralt… when he matches these 2 players will generate a bunch of 60-70% winrate but when he matches Showtime, then the % drops). It is all based on matching players where you are favored.
If this was Trap constantly matched on ladder, win% would also be different but the idea is skewed again by who is favored.
Even so I do not think balance should be for Ladder heroes… it should be for pros as it is now. Cant say what it is that makes it more difficult for T to play vs P on ladder but there’s something.
And every T is more OK playing vs Zerg… like why we all share the same experience?
Answer - because on ladder you match equal skill and maybe let’s not say balance but design happens - where what players do/dont matter based on how race is designed.
While on pro games, it seems you kind of know the favored player, unless we see e,g Rogue/Trap/Maru/Dark/Serral/Inno where any1 can win .
1 Like
The thread is going way better than I expected. After seeing state of the forum regarding balance talks, my expectations were very low. Thank you very much for engaging and showing me that I am not perfect ^^.
The title is an argument I face in the chapter. My conclusion is:
“It is easier and more effective to balance game around pro players.”.
I do not think I understand your opinion in that part, can you clarify please?
Also results for both GM and Diamond are similar in terms of % win-rate in PvT. (3),(5).
.
.
.
Anyone has an idea or relevant experience to check what is wrong with PvT match-up? Strength and scaling of units? Advantage of one design above another? “Unfair” units?
I suspect many would point out disruptor. I will try compare my performance in PvT match-up with some games with, and some without disruptors.
I wonder if analysis of 10-100 replays with noted strenghts and weaknesses of each player, and corelation of their performance to winning side could be useful. Sounds like work for Excel.
The same as yours.
As someone who has played this game since WoL, I’d say the design was always - terran can win at 8th min or lose late game. There were times like in 2017 where Protoss wasnt so strong to always win late game and could lose. But patch from Nov 2017 overbuffed them and made them strong in 2018 vs Terran.
Shield battery proxy, overcharge battery etc just made protoss be able to hold what previously allowed Terran to beat them early. Also they changed cyclone, proxies were nerfed, so if terran couldnt win early game with all protoss defenses, late game was even worse for terran. Thats part of the prob imo. Even if they nerfed protoss they allow them to survive and in expandcraft terran time is ticking - if you cant beat them early gg, you can if they mess up but if their economy is not damaged and you try to compete w economy as Terran, well you lose.
For me the TvP win rate went from over 60% in 2017 to always 40% or under. If it was just me fine but all Terran suddenly went bad? Pro level is different story tho
1 Like
Its how the match has to be played. you are behind in eco. (always 2 to 3 bases down. You even see this in pro games.)
Pro over come this because they have impecable timings. On the ladder this is abusive.
You cannot attack into toss even with supply advantage. you have to attack around toss ( you see this even in pro games) Pro gamers have mastered attacking around toss
Splash may not be that much of an issue in pro league because they have better control of their army via splits ( on the ladder splash damage is op and toss has it in spades)
So unless you expect me to play like maru every game. which I probably never will. toss is just a better than terran. Mech is not viable at best it is an inconsitent option and bio is on a timer
toss has had so many signficant buffs to try to get them a pro league victory it is ruining ladder play. I am starting to lose other matches out of frustration of lsses to toss and I dont believe they are warranted. I think more than 50% of my losses is do to how the match up has to be played.
I aint up to the task…
1 Like
My games even losses vs Z were more Ok as they would often be 2-sided games. But yes Protoss matchups have tilted me too to make poor on any matchups.
It’s been since WoL - Terran has strength to kill them early, late game is theirs. But now they are trying to make Protoss survive more and that brings them to victory late game.
The winrates coming from T vs P on some top GM are because indeed the Terrans that beat Protoss are the better e.g HM vs goblin/gerald/gungfupanda
1 Like
I don’t think the match-up is to bad Terran’s complain a lot about losing while being macroing and getting hit by a disruptor but spoiler alarm the same is for protoss and drops
the major issue terrans face that makes them hate this match-up is the fact that protoss can and will playing agressiv and that is something they are not used vs zerg because Zerg early power to deal with terran units is limited by range and walls with repair and resulting into massiv offensiv pressure from terran
while TvP is more dynamic and most terrans stick to long in one phase the tech race while protoss can pressure a bit with adepts and stalkers they try to rush blink or robo-tech to deal with stim/ drops or mech
AOE is always a major point in TvP because Protoss need to rush it because the protoss units are less cost efficent while building up as example colossi the protoss are researches high templar and storm since the expect terran to counter it and losing their curical AOE units terran’s always asume to stick with that and get surpirsed their counter get countered or is useless and the start to build the counter and already lost the techrace
a Random player diamond lvl play from EU ( i am also writing my race aswell in the chat at the start because i want to play as normal games as possible while not choosing my race it would be cool to have this reavelled in the loading screen)
1 Like
Thank you very much everyone for taking part in the discussion.
1st post fullfilled it’s goal and is kind of out-dated in some parts.
A few my beliefs changes thanks to the experience and I started to see what problems Starcraft faces and that it is not a perfect game.
I also found a few nice tools to observe different things, so it was well spent time ^^
I kind of close the topic.
Win-rates mean basically nothing on the ladder because players with a non 50/50 win-rate have their rank adjusted higher/lower until their win-rate balances out. The information is shifted into the actual rankings themselves. If a race has a higher average ranking in a symmetrical matchup than in an asymmetrical one, it’s proof of a balance issue. If Terrans on average have a higher TvZ ladder ranking than TvT, for example, it means TvZ is Terran advantaged.
Speaking of not just the mean, but the whole distribution: If a race is underpowered, what we’d expect is a large decrease in the tails of the distribution (small shifts in the mean produces large differences in the tails) with their representation peaking in the middle leagues (where new players are placed and who haven’t played many games, and where their rankings aren’t driven by performance yet, and where higher-league players will be placed given a performance disadvantage). We’d expect the number of players in lower leagues to reduce from race-switching (confirmed to be happening) and we’d expect representation in higher leagues to reduce due to demotion (confirmed by the data) and we’d expect the middle leagues to bulge due to the flood of higher-rank players being demoted (also evident in the ladder data).
Zerg is the least played race across the whole ladder, with zerg’s numbers going down to proportional increases in Terran/Protoss numbers (aka race switching). Zerg’s numbers in higher leagues have never been lower, in fact no race has ever been as low in GM as Zerg is currently.
Zerg currently peaks in the middle leagues which math models show is due to the increased prevalence of ZvZ. Zergs demoted from higher leagues inflate the lower leagues, causing ZvZ to be much more common. This means their symmetrical matchup ranking becomes the dominant factor in deciding their mmr. So balance causes a negative performance hit that is demoting Zergs from higher leagues, causing increased zerg numbers in the middle leagues.
In other words, the state of the current ladder is clean-cut proof that Zerg is heavily disadvantaged in balance.
One prediction that isn’t verified yet, that this theory makes, is that there should be a skew in the number of games played per-race in the middle leagues. If the rankings of the middle leagues are primarily controlled by new players, who haven’t played many games, and high-ranked zerg players are being demoted into the middle leagues, then the number of games played by Zergs in the middles leagues should be higher than T/P.
Following up on this. We can confirm this with a random sampling from rankedftw from platinum:
Zerg set: 42, 49, 60, 66, 73, 78, 81, 99, 121, 135.
Protoss set: 6, 9, 22, 35, 62, 88, 94, 98, 105, 120.
“Set” represents the aligulac page each player was randomly selected from. The higher the page number, the lower the rank of that player.
Zerg games played: 1, 44, 1, 1, 63, 3, 110, 1, 77, 231
Protoss games played: 9, 2, 21, 2, 2, 27, 92, 1, 2, 21
Correlation between games played and low set: 0.497. Thus we have proved that middle ranked players play the fewest games (confirming the hypothesis that middle league representation is controlled by the default ranking of new players who have not played many games, and not performance).
Average Zerg games played: 53.2.
Average Protoss games played: 17.9.
Thus we have proved Zergs in middle leagues play more games, which means they are higher ranked players who have been demoted into leagues populated by mostly new players.
I don’t even waste time with the weeping Terrans. There were Terrans complaining in the forum that Protoss was OP at a time when they had 70% winrate at Alligulac
1 Like
But it’s a quite nice topic ? For once that we get a constructed thread, aiming for neutrality and comprehensiveness, it’d be a pity to let it go. Now, exchanging constructed thoughts on forums is timing consuming, and it’s understandable for one to wish to trim that down.
And it’s been a pleasure to read it as well. However, there were a significant amount of statements there and there which prevented me from wholeheartedly agreeing with it. So I will comment some of them a bit :
This, the amount of smurfs accounts is indeed the main reason why GM race’s proportions should be interpreted with caution. There are only 200 players in GM for each region, and sometimes (specially in the first weeks of a season). So 50 players among those each bring one or two smurfs with them, it can considerably alter the races’ proportions.
Statistically speaking, this parameter is described as the robustness of some data. If changing a bit one of the source variables greatly affects the final result, then that means the amount of uncertainty regarding this statistic is high, and it should be considered as “fragile” and interpreted with caution. That, IMO, is the case for the GM races’ proportion.
Interesting. This should be put in parallel with the amount of protoss tournament players though, but it’s interesting.
Indeed one unit could be unfair, but that’s compensated by the other unfair units of the other races. Which is why imbalanced units could lead to a balanced result, as long as the imbalances are spread in a balanced manner. 
Debatable. All the players should be taken in account when making a choice, but there is a certain level of execution required in order for your own ingame experience to pretend at starting to be relevant. A terran who can’t split shouldn’t whine about banelings for example. At the other hand, balance shouldn’t be based on only Maru’s level of split ; and the effort of splitting execution should be taken in account by the balance team.
Those would be massive nerfs to bio and also mech harass, and reflect in my opinion an exclusively protoss QoL point of view.
The harass mechanics are a way for the developers to balance the economics mechanics :
- Z has the best economy, followed by P and then T
- Terran has the best harass tools, follow by P and then Z
So if one were to considerably nerf the terrans’ harassing tools, then that would induce that mechanic not compensating the economical gaps anymore. Z or Protoss could get their 70+ workers economy unhindered, and then roll over the opponent. I understand it feels frustrating for a protoss player to lose 10 probes in 5s of inattention, but as you said it yourself, losing 10 probes isn’t game-ending for a protoss, it could even be considered like evening things out against a terran that will land his third base much later than the protoss could. If you lose more than that you screwed up, if you lost less, you’re still in a good position.
That also applies on reverse in PvZ, where the P should try to kill some workers against a macroing Zerg, and will be in a bad position if he doesn’t slow him enough. And so the harass is an additional mechanic to balance the spontaneous economy of the different races.
More discussion about the 3 races’ economy here (look at the workers amounts at 6’00, MULES included in particular).
Dealing with mine drops openers is most of the time a L2P issue, in particular regarding map control and minimap awareness skills. I provided advice to a master league protoss player complaining about this in my community, by analysing one of his replays (in particular regarding pylons and obs placement/sieging), and some weeks later he told me he did much better against those. Never saw him whining about those again ; and that without the least balance changes in this area.
That’s part of why the previous point was debatable. When arguing about imbalances in diamond and even masters, the players should first question their skills not being up to the task rather than the game being imbalanced. There is an awesome series about Harstem doing this kind of analysis, which I can recommend to everyone complaining about balance.
But what winrates ? As far as I know, upon selling the game the original balance team said that ±5% was acceptable for full ladder daily winrates, not for top of the ladder/progamer monthly average. Having PvZ (for example) at 45% one day with bronze to masters players isn’t the same than having progamers not managing better than 45% in a full month, IMO.
Humility is the first step in being able to see one’s own mistakes. And so in being able to get closer from the truth in the future. Hence those are wise words IMO, kudos for your appeased and open attitude, Dallarian ! 
PS : I saw the smurf you referred to in the other thread. if you give me a bit of time to anonymize the data, I think we can answer to your questions by ourselves. 
1 Like
Not really, since GM league kicks out inactive players. It’s extremely difficult to maintain multiple GM accounts. You’d have to play a ton on each account.
The proof for this is quite simple:
h ttps://www.rankedftw.com/ladder/lotv/1v1/mmr/?f=am,master
Tons of masters league accounts with MMR above the promotion threshold. In fact, there are about 400 accounts with mmr above the GM threshold, who are in Masters right now. Some of them are as high as 6.6k mmr. That’s clear cut proof that GM filters smurfs quite well.
Protoss dominates GM, wins the majority of tournaments and is the only race with win-rates above 55% in premier tournaments. Protoss is definitely overpowered, and there’s no way around it.
That’s a good point. I suppose that would limit the access to GM for a lot of player. However, if I’m not out of date, I believe in order to avoid demotion one had to play 30 games/3 weeks. In other words 10 games/week. Even for two accounts (=20 games/week total), that seems easily doable for most pro players. 
Last Aligulac monthly report (October) are 50.12% PvT (which is balanced even by my criteria) and 47.6% PvZ. That’s not perfect, but doesn’t seems that bad in terms of overall balance to me.
As for tournaments victories, I feel it’s been now quite some time since we didn’t have a protoss win a Premier tournament in EU or KOR. So for my part, even if some aspects are still debatable, I don’t see protoss as globally overpowered currently. 
1 Like
Anyone who claims this loses the ability to claim anything about game balance permanently. The game has literally never been more imbalanced than it is currently. If you can’t identify imbalance even at its absolute worst, you can’t identify imbalance period.
1 Like
I permanently lost my ability to whine ?! O_o
Have you said it earlier, I would’ve done this years ago. 
Not sure permanently crying to imbalance would be more helpful identifying the worst imbalance periods, that being said.
As for general balance, GM repartition ≠ global balance. It’s only one of the many parameters to be taken in account. Let me hint you about another one :
50.12% PvT, 47.6% PvZ and 48.19% TvZ : it’s not perfect, but it’s not that bad honestly. Compare that to :
- the 53% PvT and 43,7% PvZ of December 2017.
- the 41% PvZ of December 2015, followed by 45% PvZ and 52% PvT in January 2016.
- and to the 41% PvT and 54% TvZ or early 2013, you’ll see current isn’t so bad in fact.
Say hi to Nonsense Turkey from my part. 
3 Likes
But it’s a quite nice topic ? For once that we get a constructed thread, aiming for neutrality and comprehensiveness, it’d be a pity to let it go.
My believes evolved into something what now I find more proper and accurate.
So like half of the post is outdated.
My initial idea was to go through it again and update it with what I learnt, overall improve the form and quality. (version 2.0, old version left here as reference, but basicly new and totally different pdf file)
Also mentioning all the mistakes and other things incorrect I have made (I am aware now of some things I have done wrong due to laziness, for example I used very selective data, taking into consideration actually only 1 patch and making results on it).
Also in updated version, I’d approach topics differently.
But I no longer have that much free time, so I decided to close it in a shorter post, to cut me away from mistakes I have done.
1 Like
I dont follow that… I can not succeed convincing some STUBBORN GM poster (whoz made up his mind which church will follow) with 0 clue when there’s a real issue, why is this a problem and one should waste more energy in convincing? Thats why when I know im more correct and the opposite one over there has no much knowledge, I have resorted in insults than trying to ‘convince them’. It has been a way of saying - shut up you are wrong you dont even play, im not gonna waste time/energy on you… that’s the message behind my rude posts.
Just reviving this thread as a plug to myself. Around the same time I proved quite clearly Protoss were over-performing in mirror elo rankings and now Protoss gm/tournament performance has shot off to the moon. Haters say I am wrong about everything but threads like this show I am the only person in the history of sc2 to make such accurate long term predictions; nobody else comes anywhere close. GIGApecsflex.jpg.