Rating a basketball player based on how many shots they sink, or a football quarterback based on how many passes they complete is a common measure of individual talent or skill… There is nothing “arbitrary” about those statistics, at all. I would have thought this self evident.
To clarify: you are arguing from a terrible position. You are ignoring the basic truth that a coordinated effort is always better than an uncoordinated one… The only time that is not true, is when the gap in power (skill in the case of HotS) is so vast that it doesn’t matter-- And this is really one of the issues with the current mmr system; there should NEVER be a gap that wide, which is why what you are arguing doesn’t help your case.
I gotta agree with you there. I think about 4/5 of all my games end with one team with all forts and the other team not even touching a wall. Hardly ever seems like both teams might even for a second be equally skilled. Especially some recent games where the enemy team only got a single kill compared to our 22.
I’m now on 75% winrate over the course of 3 seasons, started Ranked last year climbed all the way from silver 3 (deranked to Silver 4 first game because haha thrower goes brr) now Diamond 3, very little grouped games and most of them are solo Q.
QM is a fiasco, it has a completely different law of rules that does not work and actually ruins the concept of how the game should be played for best performance (the amount of people who still fail in understanding that the solo laner job is to soak and not to push is astonishing here), even though the MMR calculated is equal, it’s always comes to who’s the weakest of the chain and “how easy i can abuse it”, if you want to win you.
Additionally there’s certainly also unfair advantages of the nature of it, for example, 5v5 assassins, the. Side with the highest self built sustain and poke tend to win more than those who lack it.
Yeah… QM match maker is a major issue. You can be on a GM account, a plat account, or a master account. And you know what you get? Bronze and players below level 10 regularly. Cause it gives up too quickly and just throws bad games. If I wanted to play QM and not have train wrecks I’d have to be in a very specific MMR range or bring a party so it takes less time to fill the rest of the game.
And I think further the rumor is that QM is falling in popularity heavily. So even less players in pool. More players smurfing in SL or playing brawls instead. Which further erodes that mode.
I already posted this yesterday, but thought it might have relevance in this topic for archival purposes.
If the matchmaking system ever evolves, please don’t calculate average MMR again like this (by creating a 2000 rated match with one super high 4000 MMR player and a bunch of low rated ones). It is not good for retaining players and keeping them happy.
Though it mostly happens out of peak times, it is still not a fun gameplay design.
I get it, there aren’t many good players to match those high MMR’s with. But if no suitable match is found, just don’t create a match at all. The exact same thing happens in QM, we just can’t see it because MMR is hidden (we have to find it out from Heroesprofile and third party sites).
I appreciate the response and I get the point you’re making but my main problem with SL is the other players, e.g. knowledge, response to pings, general conduct, stat shaming.
While I get that it’s a team brawler, I don’t feel that I should have to “grind to the level of my incompetence” … that’s just not fun for me TBH.
I do appreciate all the work you and the other devs do though which is what keeps me playing the game. But these days I mostly just play QM just to experience the game and keep my reflexes tuned. Sadly I just don’t have the patience to deal with the potatoes that end up on my team depending upon the time of day or day of the week. If I’m being matched with equivalent players then paying Friday nights for instance shouldn’t be an exercise in futility.
A lot of people for example use Master+ as a mere argument of specifying certain character being OP but ignore how inflated the numbers of winrate for these players.
Master+ is like a whole different type of players that I don’t think I belong there, yet (probably not).
Due the MASSIVE skill gap, it all comes down to which chain to abuse the most.
You know very clearly that the Jaina is a plat player and the Tank is a diamond, you tell your team and yourself focus on those the most because they are the mostly that going to screw up in the process.
Sure drafting plays a role but this also plays a big one by itself.
When a main tank is a bronze five playing in a game of goldies you know that side is going to struggle, very badly because the tank will play very poorly.
They will match you with players with very low win rates to force you to lose games. If you are in the upper 60% in win rate, youwill get players in low 40 or 30s and they will force you to lose.
I won 12-15 games yesterday and today i have lost 5 of out 5 cause of players just diving into team and chain feeding kills. Forced win rate is real
The biggest problem with the Matchmaking system is that there aren’t enough players still playing the game for Matchmaking to work the way is is “supposed” to. The game can only choose to match people who are currently in the queue, and if there is very few people in queue, well then the matchmaking will suck.
With all that talk of MMR in your post, why is it not shown anywhere to players? I can’t think of a single game on the market that uses a MMR system that keeps it hidden from the player besides HotS. I think it’s a fairly common belief that you guys don’t show MMR because you know your system doesn’t work. Whether or not that’s actually true is certainly debatable, but if you’re confident that your matchmaker is indeed working as intended, why hide everyone’s MMR?
This would be my only real complaint about MM, what you displayed above MMR averaging in an extreme form. It only bothers me in Ranked as you shouldn’t have such a huge skill gap where a GM is matched vs a Silver. Sami stated the problem with this clearly.
Another problem is unless you inspect everyone’s profile while drafting, the lowest rank might take a key role such as tank or main damage dealer and be a huge detriment to your team. I wish ranks would be displayed on the SL draft screen.
I know this also happens in QM, but I really don’t get bothered by what happens in QM.
that would be a much better metric to put out in the first place; but even then, the values on these things do change. If the mera or anub upload another game, the averages are going to swing.
The anub is listed as a 0 level account, which can’t really be ‘possible’ if he has more than one game played. heroesprofile is actively trying to assess a weighted value on the least known player (the anub) to suit the metrics of that group’s value because they’re were all matched together in game with other values.
If you look at more detailed information on those accounts, you’ll find percentages that don’t add up, hero highlights that are inaccurate (one profile says a 0% winrate hero is their most successful pick) and, as a mentioned before, account levels that don’t show the correct value of the heroes shown on the profile (the site should at least be consistent with it’s own information)
For a number of players in that match, profile would be left to think it was the best match they’ve played based on how it evaluates what it knows of those players
And yet, I frequently get games with 3 2k+ players against my team where I, at 650, is the highest ranked player or the second highest ranked play. I generally only play QM, but it makes no sense that a leveling system used to suggest skill merely by time played could be so grossly inept at determining teams. Let alone how the game determines the hero compositions.
If the game forced a tank class and a healer class in the match, hands down, the game would drastically balance out a great deal. If people whine about how long a match takes to get into, they will start playing a wider array of heroes to combat the time sink. And, in this way, players grow towards being skilled again, because they won’t play the few extremely OP heroes game after game.
Rank is not the same as MMR. Because you mostly play QM, your MMR changes without your rank changing.
Think of it this way: If a player only did their placement matches in SL, got put into Bronze, and then NEVER touched SL again, do you think that they should only ever be placed in matches against other Bronze league players, even if they have played enough QM to be as skilled as a Master level player?
There are some things that affect how long it takes for the MM to make a match though:
Population
Players queueing at the moment
How long the other players were already queueing
Sometimes you will get those weird games (at least in QM) duo to these things above. Sometimes you get instant queues but the other players were already queueing for way longer and you were just what the MM found would somehow fit considering the amount of time they were already waiting. Sometimes you will get instant queues just because there is less people queueing at the moment so the MM can’t really wait too long.
These things will happen way more if you are talking about QM, just because QM was never meant to be a balanced gamemode (which doesn’t mean the system doesn’t try to make balanced matches, it does), but quick is the key word here. The most fair games we can ever have is in ranked duo to how MMR and MM work and also because this is a moba.
It is a little off the topic. In order to move up the rank, you need to improve your skill and knowledge.
skill including but not limited to:
micro (mechanical) skill. i.e. land KTZ combo, knowing how to dance with Tracer/Valla, follow up CC
Macro skill (decision making). This has the largest impact in the lower tank. : i.e. be at the right place at right time, when giving up the obj, when to boss, when to take the camp, watching the minimap, when to soak, when to do the wrong thing with the team, when to do the right thing without the team.
Drafting skill. There are multiple correct answers in the draft but there are also many bad answers. (no damage isn’t everything.)
communicate skill
behaviour skill: i.e. not getting tilted, not to be toxic, accept the loss and move on, not getting a fight with a toxic player
If this is the case, that’s not necessarily a bad thing. Having your account permanently influenced by games you played when you were a complete noob 3 years ago, and having those games weighted equally to games played at your current skill level, would in my mind be far worse than recent games having “too much” influence. The belief that ancient matches are too influential on MMR (a sort of “hardstuck”) is also a major cause of smurfing based on my experience speaking with smurf players in various games about their reasons for creating a new account.
There are plenty of legitimate criticisms of the matchmaker, but I wouldn’t fault it for giving long-past matches less weight than recent ones, if indeed that’s the case.
