Matchmaking is a hoax

General Discussion
Prev 1 2 3 7 Next
12/06/2018 05:34 AMPosted by ImctrEU
12/05/2018 02:49 PMPosted by Kyobi
Terran bio will also have very high apm by merely stutter stepping. APM really means nothing before GM.

Id be playing any non-Terran race and have higher APM than any main race player of similar MMR.

I find my APM varies based more on opponent’s race. My APM against terran is always very high no matter what race I play as, while my APM against Protoss is typically quite low.

I would think though that the less your APM varies based on MU, the more you may want to look into how much spamming you are doing.
12/05/2018 12:31 PMPosted by Europe
The more you play the more you realize Blizzard algoritms intentionally force 50/50.
Dude, you do realize, that's exactly how it is supposed to work, don't you?

It's supposed to match you with opponents of equal strength. If you win, your opponent was objectively weaker than you, therefore, you get a stronger opponent the next time and vice versa. Therefore, your win-loss-ratio will idealy become 50-50 as long as you're neither an excessively good nor bad player.
12/06/2018 11:11 AMPosted by ArthurDent
12/05/2018 12:31 PMPosted by Europe
The more you play the more you realize Blizzard algoritms intentionally force 50/50.
Dude, you do realize, that's exactly how it is supposed to work, don't you?

It's supposed to match you with opponents of equal strength. If you win, your opponent was objectively weaker than you, therefore, you get a stronger opponent the next time and vice versa. Therefore, your win-loss-ratio will idealy become 50-50 as long as you're neither an excessively good nor bad player.


No, it is not how it supposed to work. Platinum players shouldn't play vs master players and vise versa! <--- JUST ONE EXAMPLE!!!

It's not even about APM, that was just an example. Matchmaking doesn't work, it works as good as Blizzard implemented it, and they implemented it bad. More, they implemented it bad on purpose with a reason(s). Some of the reasons i already mentioned before.

Matchmaking in SC2 is simply bad.
12/06/2018 01:31 PMPosted by Europe
12/06/2018 11:11 AMPosted by ArthurDent
...Dude, you do realize, that's exactly how it is supposed to work, don't you?

It's supposed to match you with opponents of equal strength. If you win, your opponent was objectively weaker than you, therefore, you get a stronger opponent the next time and vice versa. Therefore, your win-loss-ratio will idealy become 50-50 as long as you're neither an excessively good nor bad player.


No, it is not how it supposed to work. Platinum players shouldn't play vs master players and vise versa! <--- JUST ONE EXAMPLE!!!

It's not even about APM, that was just an example. Matchmaking doesn't work, it works as good as Blizzard implemented it, and they implemented it bad. More, they implemented it bad on purpose with a reason(s). Some of the reasons i already mentioned before.

Matchmaking in SC2 is simply bad.


aslong as the master doesnt play unranked or lost many games on purpose, to lower his mmr, plats wont ever play against master player. thats over 1k mmr difference.. MAYBE this can happen in 1 out of 50 games, but not more often.

OR the plat is playing with provisoral mmr, which means, its not his real mmr and the bnet still figuring out, how good the plat player is.
You'd be surprised at the amount of sub-diamond leaguers who have 300+ apm.
Meanwhile I defeat them in my sleep with like 130 apm. It's pretty silly and adorable.
12/05/2018 09:17 PMPosted by WireBender
12/05/2018 06:32 PMPosted by niteowl
Another bizarre post on this forum....... why am I not surprised?

But seriously...... the APM does not equal MMR, and regarding the 50/50 thing, yeah, that's the whole purpose of the algorithm....

I think you misunderstood what he’s saying about the 50/50.

What it is doing is trying to rank you in a middle ground where when you win you on average get theoretically marginally harder opponents and slightly easier when you lose, with fluctuations being so small that all your opponent’s may as well be equal skill level.

What OP is saying is 50% of games it’s putting you against people who are leagues ahead of you and will near certainly win, and the other 50% it’s putting you against people who are noobs compared to you and are almost guaranteed to lose.


I understand that, I was just stating that in my experience that is usually NOT true.... like I said it can happen occasionally (depending on time of the day there might be less people online and so it might be hard to find a perfect match), but usually most matches are fairly even...... (and there are also smurfs, cheese, etc, which can make matches fairly even in terms of MMR look rather swingy... for example I remember I played the same cannon rusher a couple of times, in one game I stopped the rush and won so easily it felt like I was leagues ahead, in another I messed something up and lost badly.... in both cases our MMRs were similar, even though it feels like the guy was way inferior because if his cannon rush failed he then looked like a low bronze at anything else).
But yeah, with less people online, like at night, the system has no choice but to match us with a wider range of players..... it's not perfect but imo it's not as bad as the OP claimed it is.
anything sub 80 apm is pretty average, you simply can't accomplish enough post diamond. So you team with people with 40-50 apm its no wonder you die, but on the flipside you can kill people 100apm higher than you.

making 4700mmr with around 90-130 is pretty doable, but generally you either roflstomp or get blown out, there hardly seems to be inbetweens since the system is so broken and volatile
12/08/2018 11:14 AMPosted by llll
anything sub 80 apm is pretty terrible on average, you simply can't accomplish enough. So you team with people with 40-50 apm its no wonder you die, but on the flipside you can kill people 100apm higher than you.


If you are protoss you can take games off x3 times better terrans even with 45 or 60 or 70 APM. Just camp (expands ofc but turtle defense) and mass to mothership carrier/ tempest. Then A move, there is no punish for mis-control/micro mistake on ships since such thing does not exist.

(now I dont have such game in new patch but since BC are AG not AA esp after nerfed yamato, dont expect BCs to be used like TY vs Creator)
You'd be surprised at the amount of sub-diamond leaguers who have 300+ apm.
Meanwhile I defeat them in my sleep with like 130 apm. It's pretty silly and adorable.


No I wouldn't be, because you pull those statement out of your a55. What a joker. It's like you guys play another game.
Matchmaking is soo fcked up in random team that it's not even remotely fun to play.
12/05/2018 12:31 PMPosted by Europe
When 300 APMers are matched with 100 APMers and vise versa you know its all fraud.


APM is not what determines a player's skill level. If anything it's one metric that it's inflatable by smashing the keyboard to input commands or holding a "build this unit" button. If you're trying to determine whether or not the game's rigged by using APM as a metric, you're mistaken.

Since you mentioned "ProTech" and team games, StarCraft 2 is not built around team games and they're underpopulated anyway so the matchmaking is going to be off there anyhow due to said population.

The game doesn't "let you win" since winning or losing is up your ability vs the other guy's ability. Matchmaking puts games together based on MMR, therefore you're always playing against people who are within a certain range. The only time you'll see a drastic difference in MMR is if you're in high Master or Grandmaster leagues in 1v1 or in team games where the population is just as low.

Stop blaming the game for why you lose.
12/09/2018 12:18 AMPosted by Yuurei
Stop blaming the game for why you lose.


I don't blame the game, i blame Blizzard.

Besides i have nothing against me losing vs similar skill level players, however i do not like to play vs masters as plat/dia player. And this happens in team games as well as in 1v1 games.
12/09/2018 12:54 AMPosted by Europe
I don't blame the game, i blame Blizzard.


Fair enough. Team games are underpopulated so matchmaking isn't going to be any good. It has to do with StarCraft 2's fundamental design for 1v1 competitive play.
I think the op is heavily confused here.

Team games are pretty wonky as far as match making goes, but to call it intentional is kinda foolish. There is a perfectly solid reason why team games have wacky skill match ups. The player population for any given que in team games is much smaller, on top of it also requiring more players. In a recent thread I did the quick math on this and there are something like 18 2v2 games started a minute on average if I recall correctly. That's not a lot at all and the odds that you'll find a good match in those 18 is not good. Since the mm in teams gets random based on this it's going to skew everybody who plays it towards 50% because it becomes coin flippy by nature. This perfectly explains why team games skew towards 50/50 and nothing is wrong on the programming side.

As for mm in general anyways, yes, it's actually designed to get each player to roughly 50/50. That's the point of it. If each player played 100 games a day for 100 days (10,000 games) you'd find that nearly all players would have between 49-51% winrates and that would only deviate at the extreme ends of the ladder. Each of these players would fall directly into where they should be skillwise and would proceed to go about 50/50 in games (assuming they don't get better in the time being for a clean example).
12/08/2018 11:59 AMPosted by ImctrEU
anything sub 80 apm is pretty terrible on average, you simply can't accomplish enough. So you team with people with 40-50 apm its no wonder you die, but on the flipside you can kill people 100apm higher than you.


If you are protoss you can take games off x3 times better terrans even with 45 or 60 or 70 APM. Just camp (expands ofc but turtle defense) and mass to mothership carrier/ tempest. Then A move, there is no punish for mis-control/micro mistake on ships since such thing does not exist.

(now I dont have such game in new patch but since BC are AG not AA esp after nerfed yamato, dont expect BCs to be used like TY vs Creator)

Turtle skytoss is a bad build and if you lose to it you should feel bad.
12/10/2018 08:46 AMPosted by WireBender
Turtle skytoss is a bad build and if you lose to it you should feel bad.

The PPP speaks agai (when they say you lose to skytoss? Impossible!). Personally haven't played it on the new patch yet. But I saw a game of protoss streamer vs another Terran where the Terran army got obliterated once again. I would blame it on the Terran not making enough air but there is your 'Losing to skytoss lmao rofl'. Whoever denies the power of skytoss has no clue what it is to play vs it.
I dont really ladder enough to say I agree that htis is true in SC2 but in Overwatch.....yes. Definitely true, and the ranked system sucks in general though they kinded of tried to fix it with looking for group (LFG) feature, the other big problem in overwatch is the hard counters now, not like soft counters like widow vs pharah but hard hard counters...I would say APM does matter a bit in SC2 though for sure but still it is enough of a strategy game that you can still win....but yea SPM more important and decision making, macro, scouting, BUT against good players with good builds, control of your units is really important,, like controlling your oracle or obs vs zerg etc or like pheonix or prism w/immortals vs terran pushing early with tanks and bunkers.
Exactly same crap going on in OW and Hots. Hence ppl don't want to play Blizzard games anymore. Casual wise is all about secret algorithm and less about individual skill. It's all in the code, guys. And yes, it is intentional, no doubt about that.
12/10/2018 07:52 AMPosted by DrFuzzyGlove
As for mm in general anyways, yes, it's actually designed to get each player to roughly 50/50. That's the point of it. If each player played 100 games a day for 100 days (10,000 games) you'd find that nearly all players would have between 49-51% winrates and that would only deviate at the extreme ends of the ladder. Each of these players would fall directly into where they should be skillwise and would proceed to go about 50/50 in games (assuming they don't get better in the time being for a clean example).


Exactly you basically just agreed Blizzard forces players into 50/50 crap. It's just you're confused, lack of statistic and programming knowledge and don't want to understand how it works. Plus, you're a fanboy like many around here.
12/10/2018 01:14 PMPosted by Europe
12/10/2018 07:52 AMPosted by DrFuzzyGlove
As for mm in general anyways, yes, it's actually designed to get each player to roughly 50/50. That's the point of it. If each player played 100 games a day for 100 days (10,000 games) you'd find that nearly all players would have between 49-51% winrates and that would only deviate at the extreme ends of the ladder. Each of these players would fall directly into where they should be skillwise and would proceed to go about 50/50 in games (assuming they don't get better in the time being for a clean example).


Exactly you basically just agreed Blizzard forces players into 50/50 crap. It's just you're confused, lack of statistic and programming knowledge and don't want to understand how it works. Plus, you're a fanboy like many around here.


The entire point of mmr is for you to gain mmr as you win, and lose mmr and you lose. This creates an arbitrary number (mmr) that is used to find an opponent with a similar mmr. By default and just playing the game each players winrate will trend towards 50/50 as they settle into their correct mmr. If they get better they will begin to beat players of the same mmr they used to have and go up some. I perfectly well understand the system and it's designed to work you towards 50/50 just like every other game in existence that has any kind of skill based match making. The only way to go higher than 50% is get better and play better which is the entire point of any competition. Testing your skills at the game.

None of the above is done with any malice or is done to "keep a player down for winning" as the op was claiming. I don't understand how me describing exactly how mmr works and is intended to work and we observe it working has anything to do with me being a blizzard fanboy. I do like blizzard games, they've made some poor decisions in the recent past, but I'll call out bad ones just as much as the rest. I'm at least fair about it.

Basically the op claimed blizzard would pair them unfairly if they won too much. No, that's what happens when a quene isnt full enough and you quene enough games. You'll also hit unfair match's when you're losing as well when a quene isn't saturated enough.

I understand there are blizzard fanboys who will defend them to their death, there are also tinfoil hat conspiracy theorist and haters who will condemn them no matter what, despite the fact that they still play their games which confuses me more... I like blizzard, I'll call out bad moves when I see them though. Using a standard mmr system isn't a bad move... I have no idea about ow or hots match making, barely play either, but sc2 mmr works exactly as it should.

Join the Conversation

Return to Forum