Algorithmic Handicapping (MMR) is Wrong for Overwatch

Please try to refrain from Ad homonym attacks. This does nothing.

If you have a real counter argument other than “He’s a crazy conspiracist!” Please enlighten us.

2 Likes

There’s no point in arguing with any of you if you’re just going to dismiss it and blame “handicapping” regardless. Honestly if any of you got a vod review and got to see where your mistakes are you would climb and forget about this nonsense but nobody actually wants to work to improve. There is no handicapping to keep you down, only your own play. What rank are you? Gold? Plat?

I could get you out of your low rank in a day on your account, yet you can’t get out. Ie you’re not handicapped, you’re not objective about yourself, and you don’t want to improve so you blame the system when the system is actually relatively good at placing people where they belong. I’ve been from gold to Masters, and there is a very clear trajectory of play that rises with the tiers

Wow, your arrogance is…astounding. A VOD review and fixing mistakes in your play isn’t enough to climb on it’s own. Been there, done that…three times. Three separate, unbiased opinions by people ranked higher than me. People who climbed from gold/plat to diamond/masters.

Here is what you and the White Knights are not admitting, or are just ignorant of:

In the ranks of silver-plat there is a large variation in individual skill within each rank. That’s why you can run across really good players in silver and really bad players in plat. And it’s not just a variance in skill, it’s a matter of players being grossly over and under-ranked.

Odds are one of these players will frequently be in your comp matchs, but they may not be actual smurfs/throwers. The disparity between their SR and their actual play skews the performance for everyone else in the match. The only way to compensate for this is to play a metric crap-ton of matches so that the sheer volume compensates for outliers.

I joke about it needing to play 40+ hours a week, but really 20 hours/week might be enough. Still, that’s equal to time spent at a part-time job, not what most people want to spend on a hobby. Your SR is heavily influenced by other people because you’re judged as a team but given a rank as an individual. Why is this so hard for people to grasp?

TL:DR
I don’t really believe active handicapping is occurring, but the system is deeply flawed. Wide variance in player skill within the lower ranks along with a system which doesn’t properly measure individual skill leads to a lot of frustration. Since Blizzard cannot or will not fix this system, the only way to climb is to improve your skill AND grind hundreds of matches.

4 Likes

I don’t observe any of these things.

this is true in the 65%+ win rates you start getting paired with afking people, people who will jump in die, jump in die, jump in die and it’s like really? is this really happening right now? the last two games I had were just mind boggling that even the enemy team reported the person. It was a forced loss and it sucked

1 Like

No. I smurf in those ranks, and thats simply not true. There isn’t a single (not one) plat skilled player stuck in silver. Mid gold? sure, that can happen, but not lower.

i had like 85% winrate on genji last season, and like 90% in brigette. The season before i did >500 to 2.7 without losing a single game.

1 Like

Either they weren’t good at vod reviews or you didn’t listen. There is no such thing as a good player stuck in silver. It’s a myth. It doesn’t happen.

And what do you base this on, the odds of an over ranked player being on your team? You say it’s high based on nothing but your own biases.

In truth, over time, the other team has a higher percentage of having one of these players if you are in fact not over ranked. This is also why if you’re actually better than your rank you will climb inevitably, because the game looks to make even matches with 50% chance to win on both sides, but if you are truly better than your rank then you already, on average, will be more likely to swing the chance to win in your advantage

The system may be flawed in the same way literally every competitive ranking system for online team based games is… There are only so many variables that can actually be accounted for. The system cannot and will not for example ever be able to tell if someone is going to deliberately throw or pick a hero they never play in the next game. Just because it’s flawed doesn’t mean it isn’t relatively accurate in general. Like i said I’ve played in every rank from gold to Masters and have played with bronze players on fresh accounts. Let me tell you there is a very clear difference between the ranks. It’s clear as day really if you’re looking from a vantage point outside gold

1 Like

Another lovely person who says a person is wrong and expects that to be good enough. Another wonderful soul who automatically thinks anything that goes against their world view is flawed without any further info needed. Who needs proof, a good theory or even rational conjecture?

I’ve seen good players stuck in silver simply because they don’t play enough matches to…wait I’ve already said this. Just read above. If these forums are anything to go by, then we’ve definitely seen people who are “stuck” in too high a rank. How else could they be causing so many losses on their way down?

Only in a truly randomized selection. Comp teams are not randomized. So, you can’t make that assumption.

So, we should just accept it? If we acted like this, then no improvements would ever be made.

Flawed but accurate? Considering the flaw relates to the system’s reliability in assigning an accurate ranking…how exactly is this possible?

And I’m telling you, I’ve played in silver, gold and plat. You could combine all three into one big tier, and you’d have about the same experience. That’s how much variability there is in skill within those three ranks.

Here’s something to think about. Let’s say the White Knights are right, and everyone who complains about their teammates are the problem. What happens when one of those people is on my team? Obviously if they’re the problem, then they will cause me to lose.

In that way other people are keeping me stuck in my rank. Causing losses when they’re on my team, and helping me win when they’re on the opposing team. The net effect is no net change in my SR. Thus, keeping me stuck in my rank.

6 Likes

Cool story bro. Shame that if you drop any random plat player in silver he will destroy the enemy team, hard carrying the game, even if he’s like a mercy main.

1 Like

All Cuthbert is arguing is that there is a matchmaking algorithm that attempts to dictate match outcomes beyond just randomly picking twelve people by SR.

There is no conjecture in this argument, it is a fact that can be easily checked by reading previous developer comments, which Cuthbert provided in his OP.

This doesn’t mean the matchmaker fails to sort players over the long term - it does achieve a roughly normal distribution of SR. It introduces, however, a pronounced feeling of being on a treadmill, which in turn induces deep skepticism in the player base about rigged matches and what Cuthbert has called “handicapping.” You can observe this skepticism yourself by reading through the hundreds of posts in this thread. That’s literally the argument, it’s not the complicated conspiracy theory you make it out to be.

1 Like

…he thinks mmr is used to rig matches essentially, which has created this narrative in the minds of bad players that they aren’t the reason they’re bad, the system is. He also thinks there should be no fair matchmaking and that teams shouldn’t be matched evenly, and he says this based on incomplete information and faulty assumptions, so yes there is plenty of conjecture.

This is just a bad attitude/mindset to spread to naive players who are prone to confirmation bias in general. It’s very clear low ranked players would rather blame the system than improve themselves

1 Like

What I don’t get is… You REALLY need to try to stay below gold.

The skill floor required to climb to low gold is so absurdly low, I’m convinced if you are below that is because YOU WANT TO.

The matches are rigged… by a matchmaking algorithm. That’s his whole point. His “narrative” is around the consequences of this kind of matchmaking system, based on his observations. Whether you think the system is “bad” or not is obviously subjective - he has very eloquently laid out his argument for why he thinks it is.

You, on the other hand, are just attacking him and, using your words, other “naive players.” The irony is that, despite your claims to the contrary, you are the less productive person in this discussion.

3 Likes

It’s funny you baddies still call it rigging. So if a 50% chance to win fair game is rigged? Doesn’t even make sense. Giving a team a disproportionate chance to win would be rigging.

His observations are obviously biased, because he’d rather blame the system for the “treadmill” as opposed to the obvious answer: he’s bad, hit his peak and can no longer climb . He’d much rather blame the system than get gud, just like most of the baddies on these forums

Expected 50 percent chance to win. Keyword: “expected.” It manipulates the distribution of players on both teams based on hidden metrics to create an “expected” match outcome. That’s the literal definition of rigging.

Again, he has made a series of cogent arguments against doing this and instead adopting a hands-off matchmaker. This has nothing to do with “getting good” - relatively good and bad players will rise and fall, respectively, in the current system. It does successfully achieve a normal distribution of SR. That doesn’t mean, however, it is immune from criticism, as you suggest. Using “git gud” as the panacea argument against any discussion of the matchmaker is just brain-dead and lazy.

1 Like

Wheres my 50% chance to win then? Why did I end last season at 85%+ winrate?

2 Likes

Cuthbert only provided a single developer statement in the overview section of his opening post.

Overview
Overwatch’s designers say they “balance” matches with MMR. The system sorts the twelve players from each match into teams, based on the merit each player has shown in matches past. Matchmaking uses merit-tracking algorithms (MMR) to keep matches from being ‘uneven.’ Principal Overwatch Designer Scott Mercer explains:

"When the matchmaker creates a match, it determines the % chance for each team to win based on the match it made. The VAST majority of matches are usually near to 50% (especially if you’re a player closer to median skill rating and you’re not in a group). When we do put you in a match that we know isn’t a 50/50, we adjust your SR gain or loss based on your calculated change of winning.

"We model the synergistic effects of players being together in a group. Based upon the data we see in groups, we predict the win % for each team. We try to match similar sized groups together.

"The amount of MMR (and SR) you go up or down isn’t simply a matter of whether you won or lost, and what was your predicted chance of winning. There’s a couple of other things at work. One is the matchmaker’s confidence in what your MMR should be. Play a lot of games, it gets more certain. Don’t play Overwatch for a while, it gets less certain. You go on a large win or loss streak, it gets less certain. The more certain the matchmaker is about your MMR, the less your MMR will change in either direction based on a win or loss.

"We also do evaluate how well you played the heroes you used in a match. The comparison is based on historical data of people playing a specific hero (not medals, not pure damage done), and we’ve done a lot of work to this system based on the community’s feedback.

“While it’s a minor factor compared to wins/losses (The best way to increase your SR is still to play together and win as a team!), doing so does help us determine your skill more accurately and faster.”

Notice how he does not link to the source of Mercer’s post though. The reason he does not is because Mercer was not trying to say that they distribute the good and bad players evenly at a given SR for a match (I don’t even know how you can read that as such).

Below is the source of Cuthbert’s Mercer quote. The topic was about why matches had some weird groupings such as 6-solos vs 2 3-stacks or 6-solos vs 3-solos and a 3-stack, and the resulting SR gains/losses. It had NOTHING to do with any kind of handicapping.

Kaawumba has collected all Blizzard statements on how Competitive and Matchmaking works in his topic, so go there:

1 Like

Yeah, check kawuumba, he has all the data and there’s no biased assumptions in his post.

Good point, Mercer doesn’t outright say anything about distributing players. The language he uses about “synthesizing an expected win %” and “creating matches with a win % that is close to 50%” leaves a lot to the imagination, though. It certainly doesn’t sound like the matchmaker is just grabbing twelve players at the same SR and calling it a day.

1 Like

Well, there’s no point trying to talk you out of it.

But, if that was the case, how can I climb over and over without fault, ignoring meta and playing whatever and some times tilting my team on purpose?

1 Like