Spilo Dev Matchmaking Interview TLDW

I haven’t watched the interview yet, but if the mm is doling out close games (i.e. what players should want) then win rates are going to be close to 50%.

That’s not forcing 50/50, it just statistics.

That said what I’d like to see is the magnitude of wins as a metric.
Winning or losing is just a direction, but based on the ops’ post it doesnt seem to take that into account.

I’d also love to see hero pool metrics of winning/losing teams.
Seeing their total pool and how effective their pool is would be a cool thing to see.

1 Like

That isn’t what that means. The system is always trying to find an individual match that is close to 50-50. That is how every non random matchmaker works.

What the “forced 50-50” crowd thinks is that Blizzard is out to get them by making sure that everytime they start to win, the matchmaker is trying to force a loss to happen the next game. That is the core of the rigging argument.

It is also complete insanity because simply finding 9 people of similar skill and having the winning team gain rating and the losing team lose it results in a 50-50 win rate organically. There is no need for additional layers of complexity.

1 Like

I mean what causes them isn’t really something they can fix. They can’t fix someone being good at Kings Row but awful at Gibraltar. They can’t fix someone knowing how to work with Juno but not Brigg. That’s just normal match variation.

You don’t need echo chambers to use common sense.
What other conclusion is there if the matches are completely one sided but everyone has near perfect 50% winrate.
It surely isn’t good matchmaking nor it is luck.

1 Like

since when a blizzard employee is going to admitt that games are rigged? To get fired?

games are rigged, same as in hearthstone. They have system, they know who is cheating they always put cheaters in enemy team when you are forced to lose and you get cheaters in your team when you are forced to win.

Riggedwatch2+cheaterwatch2.

shame on you devs.

1 Like

In a very loose sense, that is what all match makers do.
If you’re winning a bunch then your mmr will increase.
That increase leads to facing stronger opponents which will culminate in an eventual loss.

In general this is fine and correct. The issue arises when the mmr overshoots and puts you way above where you’re supposed to be and then takes too long to lower. That can result a loss streak.

Again, it’s not forced and not rigging, but it’s a outlier/sideffect of matchmaking.

There’s one sided stomps because people are sick and tired of being forced to babysit the bad players in the lobby to make it even.

That would make sense but that’s not the theory. They think Blizzard is intentionally pairing someone on a win streak with inferior teammates in order to force a loss, while making sure their opponents are perfectly matched.

Like I said it doesn’t make sense because as you point out if you can’t hang at a higher you will lose anyway. The system doesn’t need to force the issue.

Ah… there you are. Trolling as usual I see…

One thing I want to know is : for people who you think are using walls how do you know for sure they don’t just have good gamesense/headphones if there are no obvious tells such as perfectly tracking enemies through walls. Footsteps in this game are pretty loud, you know, and stealth has an audio cue.

But, you know, clearly they must be cheating because they outplayed you.

I love catching sombras off guard with Photon Projector especially given how telegraphed stealth is now after the rework.

Even made one ragequit by doing that. I don’t have the replay, unfortunately, or I would’ve posted it somewhere already.

No the problem is not that you are facing stronger opponents. The problem is since the mm doens’t care how mmr is distributed and you can have weaker opponents but even weaker allies. and they waste your time with losing and wining streaks.

I mentioned it in my previous post, but I don’t think the MM is taking into account the “magnitude” of a victory, so there’s no difference between a close win and a stomp.

And depending on how the magnitudes are distributed, it could be a lot easier for the MM to find a 50/50 stomp, than a 50/50 close match.

The end result is the same technically, but a stomp will feel worse and is more memorable.

This reminds me that I’d also like to see metrics on leavers affecting the results of games.

Oh, I agree with you, just adding for the sake of clarity.

10 losses in a row gets me to buy a skin. I think that is their plan

1 Like

I do not disagree with that sentiment. That was not my point. The point is that they lied to us directly and the community is so utterly incompetent that they cannot discern that fact. It just so tiring. Like, rank is not the equivalent of MMR or rank demotion wouldn’t exist as an “entirely cosmetic” thing (which it is). Nor would leaving (also cosmetic). Or resets (also cosmetic).

If they are willing to lie about something so easily verifiable, their word means nothing elsewhere. That being said, I do genuinely believe the rest of what they are saying to a reasonable extent. None of it would surprise me.

I am just saying that Bella is correct. They are not going to be directly honest with the community because there IS a component that is not necessarily “fair.”

But then again, the community is obsessed with this “forced 50/50” narrative and that is quite clearly what they actually wanted to address. I did not watch this interview or anything. I do not have context.

All I know is that rank and MMR are blatantly NOT 1-1. And nor are they in literally any game as it defeats the purpose of climbing and grouping would intrinsically change these values which they more or less allude to regardless. If it was, they would have come out and said that long ago. But you know what happened? The exact opposite. For obvious reasons. Like, I cannot even believe I need to explain this.

It is just silly. The “it’s rigged and out to get me” crowd is just as nonsensical as the “every match is fair and balanced” advocates. Both are extremists that lack nuance and have no idea what they are talking about.

3 Likes

I don’t think it’s intentionally malicious, it’s just a poorly designed system that can make certain accounts hardstuck by no fault of the player. People who play over multiple accounts know this is true.

My main account is hardstuck diamond/gold/plat yet I’ve been playing in GM/diamond/masters for the past 10 months on a different account. It’s literally impossible to budge my rank any considerable amount on my main. It just gives me trash team mates to push me down any time I start moving up lol.

Sure they may have said this stuff but another dev many years ago also said “the more playtime an account has, the more sure it is of where you belong on ladder. The less play time the less sure.”

This could indicate they have mmr that gets stickier over time and if you spend enough time at a certain rank, it becomes much much harder to change the systems opinion of your skill, to the point where it just makes some accounts hardstuck. Would certainly line up with my experience.

4 Likes

Of course they do, their ego won’t let them accept that maybe they really are worse than whatever percentile of other players they’ve made up in their head.

3 Likes

Aside from the fact that it would be unhinged lunacy for a dev to do this.

1 Like

Aren’t characters balanced around 50/50? So if Mauga wins too much, they nerf him again. That’s why tanks eat the most nerfs, one per team, so consistency is easier to see.
When balance is used like this, games should feel more 50/50. Right?

Is this a misunderstanding on the communities behalf somewhere? (at least when talking about MMR)

1 Like

I am not entirely sure about that. It’s all relative and a lot of their balancing is for the sake of changing things and keeping it fresh. But yeah, Mauga was definitely boosting people at launch. Myself included. That being said, at the time, new characters were an incentive to sell Battle Passes. Realistically Mauga players would climb to their new peaks and you would go back to a balanced 50/50 experience once the MMR adjusted.

That is exactly the problem. It seems to care only about who wins the match so it has shorter margin of error to fail in that task if one team is much stronger as possible.

They sure do feel worse but the issue is the frequency, not just having a stomp once in a while but most of the time.

50% chance of winning is not the same as forced 50/50

2 Likes