Okay, but the premise of your post ignores the reality of Competitive Play in Overwatch. I would question the value of your “simulations,” if they are only modeling solo queue. “Frustration with solo queue” is not the point of my thread about Match Making Rating/handicapping; it’s just a one part of the picture.
So a seminal paper in this area comes from Microsoft and their Trueskill system. Most modern matchmakers share the core ideas of it, which itself is an extension of the ideas of Elo.
The “lighter” website about it:
https://www.microsoft.com/en-us/research/project/trueskill-ranking-system/
The “heavier” publication
https://www.microsoft.com/en-us/research/wp-content/uploads/2007/01/NIPS2006_0688.pdf
So you think you, as a support player, are intentionally matched with lower MMR teammates? Like the game checks if you usually plays supports, and adjusts your teammates if you do? Sounds ridiculous, doesn’t it?
Maybe you think it’s harder to earn a rank as a support. I don’t personally, but if it was true, then you could say “supports at MMR X are better players than DPS at MMR X”. Fundamentally, if you are breaking the game down by roles, your job is to be better than the enemy players at your role. The DPS/Tanks on both teams will be the same (on average), so if you are a better support than other supports at your rank, your team should have an advantage and climb.
I like how you asked why even have SR, when you already mentioned two ways they use it. Let me do the full list I can think of while I’m already making a comment.
- Punish decayed players (without messing up matchmaking)
- Punish leavers (without messing up matchmaking)
- They were able to change the rank distribution (without messing up matchmaking… I will stop repeating this after every point)
- Early seasons dropped your SR only so everyone would rank up, even though the overall winrate has to be 50%, at first. Due to the bonus SR to catch up to players MMR.
- Performance based SR.
Now you could fairly complain about some of these reasons, and people have. Some, like the placement drop in SR, have been removed.
However, it’s all temporary. The SR bonuses force your SR back to your MMR. Also, it’s all irrelevant from matchmaking. No one is handicapped based on your SR, it’s totally irrelevant. You are always matched based on your MMR, with other players who have as close as an MMR as possible.
I understood it, you just don’t want me to understand that you are literally throwing up a bunch of nonsense to try and convince people that this awful matchmaker is legitimate.
Tell me, what is like to know that your job is literally to lie to a mass of people?
You can’t have an advantage in a match that is created so that each team has a 50% chance of winning. How do you not understand that still? A 50% chance of winning is a 50% chance of winning no matter what you do.
Your game is failing, and will continue to fail and you don’t have the skills it takes to fix it. Watch it burn.
1 Like
No, that is not what I meant. My impression is more along the lines, that in previous seasons I was more successful as a Zen support main, that I am now.
Generally the DPSers in previous seasons knew when to break line of sight, when to show themselves for a healing orb, knew when to group up for tranquility.
Nowadays, the opposite is true. Despite me being more used to the game, DPSers LoS me more than they try to LoS the enemies, they rarely gather up for me to pull off a good Tranquility. Things like that. These do affect my performance.
And you can say “talk to your team”, but the fact of the matter is, I didn’t have to beg to my team to gather up for a good use of Tranq. They used to have the gamesense to do so.
If I have to micromanage my entire team at all times and to dictate everything they should do, they how are we of similar skill? Why are they in my match in the first place?
I’m not sure why is that. I highly doubt, that the game deliberately gives me worse DPSers than it used to in previous seasons. Hell, people claim, that smurfing is rampant, but for some reason I don’t get those smurfs on my teams :*(
Might be a result in changes to the META, better players leaving to othergames, or it might be totally random. Who knows.
At a certain point in time it was true. Earlier in Overwatch, many supports reported, that they don’t get enough SR compared to DPSers. Blizzard claimed to have investigated, done adjustments and so on. It was back in 2017.
However, keep in mind, that sometimes bugs do resurface in games.
As to the whole “if you do better than other supports at your tank, you’ll climb” this is yet again, because of the personal performance factor in the system for plat and below.
However, I would ask you to understand, that if the team refuses to cooperate, there’s no way for you to maintain the same performance as you would with a team, that works.
This system was removed for Diamond and above, and I believe it should be removed entirely.
2 Likes
How and why do you keep intentionally misinterpreting 50% win chance to mean that they give you potatoes on your team to counteract your better-than-your-SR skill when the truth about the 50% winchance simply means put 12 available players in the same game with the same MMR but also have to account for premade groups?
When you said that you intended to keep thinking that all your teammates are trash every single game, it became clear that you didn’t want to play this game for the right reasons.
Thanks! I’ll look into it when I have the time.
Here’s the thing, though. I have read before about True Skill. Blizzard are probably using some variation of that, considering, that they have a perfonal performance factor. Here is what I’ve read:
http://www.ets.org/Media/Research/pdf/RM-15-03.pdf
Dangauthier et al. (2007) extended TrueSkill to estimate players’ skills not only forward through time but also backward. They called this extension TrueSkill Through Time (TTT) or TTT-D when the estimation of an additional draw margin parameter discussed earlier is included. Under TTT, for example, if Player A beats Player B, and then later, Player B beats a strong Player C, TTT and TTT-D are able to adjust Player A’s ranking by going backward in the estimation.
I wonder if Blizzard are using something like this, because it would be terrible. I’m not excluding it as a possibility.
If they are using the so called TTT extension, that would explain why we can’t see our MMR… because they can be changing our MMR based not just on our games, but games by people we’ve played against… and our MMR can be changing in the time between our play sessions.
Imagine player A beats player B, then later player B decides to go and throw his next 10 games. Player A’s MMR would get retroactively adjusted (lowered), and instead of getting better allies for his wins as his SR raises, his retroactively adjusted MMR (lowered) might give him lower skilled allies.
Something like this could explain the streaky nature of the game.
In any case, I would argue, that Competitive players should demand from Blizzard to display our MMR, so we can see what’s going on.
3 Likes
This is the dumbest discussion. They won’t change anything because of this. They won’t show us their MMR system. We all have felt the effects of whatever it is, is not 100% fair games. What you should be getting people to do is stop playing the game… The numbers go down, things will change.
3 Likes
I agree 100%. Blizzard would say “no, because then players would be able to manipulate their MMR.” They prefer that MMR manipulates us, you see.
I already do everything I can to manipulate my MMR. You can bet that if they told me their exact formula, I would be doing it even more. It would probably make the game less fun for my teammates and opponents, but I don’t care about that.
3 Likes
What else could Blizzard expect ‘gamers’ to do? I appreciate your honesty, and thank you for your candor.
Of course, most players do not know what MMR is. Some don’t even know it exists. And those players pay for the gains that you make by gaming the system.
1 Like
The thing is, some players can already manupulate their MMR to some degree. A year or two ago, this used to be called “stat hunting”. They have already figured out some of the things, that affect the personal performance factor.
To make a weird analogy with World of WarCraft PvE, those players are focusing on doing the best DPS they possibly can, while ignoring every other mechanic.
In WoW, back when there was loot distribution, such players would be deprived of loot as a punishment by the raid leader, because they could have wiped the raid.
Instead, in Overwatch the personal performance factor rewards them with less points taken for a loss, and with more points given for a win, despite the fact, that what they are doing is reckless and is likely to cost the game for the entire team, but they do it anyways, because they wanna game the system.
Yeah, but that does not mean they should or shouldn’t disclose their formula. That means, that they should remove this crap from the game entirely. They already did it for Diamond and above. I don’t see why Plat and below have to put up with it.
It’s such a simple point you are missing. The match is only actually 50% if the game has you accurately ranked.
If you ARE accurately ranked, then yes you will have a 50% win-rate, and won’t climb. As you shouldn’t, because you are where you belong. If your MMR ISN’T accurate, let’s say you’ve improved a lot recently. When the system tries to make a match with your current MMR, you play better than that, and give your side the advantage.
Let’s put it another way: THE GAME CAN ONLY MAKE A MATCH 50% IF PLAYER MMRS ARE ACCURATE. It’s not “handicapping” for the game to put you in matches with and against equal players, leaving you at a 50% win-rate overall. That’s exactly what a competitive system is supposed to do.
I generally agree with no personal performance. Personally I’m glad it was removed for the ranks I play. I do know there are many lower ranked players who want it to matter more, so they don’t have to “rely” on their teammates, but I think that is a flawed idea.
However, let’s remember that your performance is compared against other players at your rank, and on the same hero. This means you’re only compared against other supports at your rank, which means they deal with exactly the same things. While bad teammates might make it harder to get good stats, it doesn’t impact you more than anyone else you’re compared against.
lol! That’s how it’s supposed to work and exactly what everyone is saying is not happening!!!
2 Likes
Why can one kid fail in a class and say the teacher was unfair? Why are there morbidly obese people who say they hardly eat? It’s easy to externalize responsibility and look for something to blame whenever you fail. People do it all the time, both in and out of video games.
Half of all players will be ranked below average. Obviously, some won’t be happy about it, and some of them will look for any excuse rather than accepting the obvious fact that they are responsible for their own rank. Blaming your teammates is the go-to for any team game. Players whine about bad teammates and elo hell in literally every competitive game.
Even the most ardent defends of “matchmaking is rigged” theories have to admit that good players climb no problem. There are countless videos of bronze to GM climbs, and anyone can go pay for a booster right now to rank you up a tier. But then you get the response “oh of course PROS can climb but not regular players!” No, anyone can climb, if they deserve too.
Honestly, how can you imagine that Blizzard went out of their way to implement a more complex system just to hold certain average players below their real rank? Why wouldn’t Blizzard just match players on their MMR, which is THE ENTIRE POINT OF AN MMR SYSTEM?
1 Like
Jeff K in a video interview with Emongg said something to the effect of “matchmaking is highly complicated engineering”, he stressed it is extremely complicated.
I don’t know why they would do such a thing, sounds like you and I both think they are stupid.
2 Likes
Woah calm down. Obviously a match maker ranking is for match making. Hidden mmr could be made null and use SR instead. Pretty simple.
Crux of the issue as always is lost in you. I still feel you are one of the designers and are defending your choices, unable to see that the complex machine you made was worse than no machine.
There are solutions to this. PBSR could be adjusted to be greater at low ranks to be nonexistent at higher ranks. This limits smurf trolling but still needs team play to be the focus for higher ranks.
Factually false. No punishment occurs.
Factually false. No punishment occurs. Well, you can get banned, but that isn’t this discussion.
PBSR can do the same, so moot. Also an unnecessary action created from a flawed system to begin with.
Even they realised this is dumb, so why bring it up to defend a system I’m calling dumb?
Um, mmr isn’t related to PBSR that way.
So as usual every point you make is wrong. Wonderful seeing you back again.
1 Like
@FriendlyFire
Here’s the thing. I do understand what you’re saying in theory, but:
- There are cases where, to use your analogy, the teacher is indeed unfair.
- There are definitely cases in Overwatch, where I can safely say I didn’t have a fair chance to win.
This could be because of all sorts of reasons.
- The might be a bug, or a change in the system, that you might not even be aware of. Specific example of Scott Mercer talking about such instance:
Overwatch Forums
As part of the 1.9 Orisa patch, we made a change to how assist scoring was handled to address what we honestly considered to be a bug. Players were getting full assist credit even if the player being assisted did very little to the target. This change, along with other more significant balance changes in the patch, meant that we needed to recalibrate the tuning for the systems that calculate a player’s contribution to the match. This was performed for all heroes several weeks ago, and we’ve already recalibrated once more after the recent 1.10 patch.
- There was a hypothesis by MHz, that you might play in a manner, that does not satisfy the personal performance factor, you might get marked for deranking, and you might get match with people, that are deranking intentionally. Feel free to read:
https://www.reddit.com/r/Overwatch/comments/7hb05p/i_have_finally_done_it_i_have_understood_the/
. - Here is hypothesis of mine, that I posted above. The technology exists and Blizzard could be using it without us knowing, and on some level it seems likely, because they refuse to show you your MMR:
So many of these are possible and they would explain, why some matches do seem like you didn’t have a fair chance to win.
However, you don’t seem to wanna pay attention to any of those. You assume, that people are whining like kids and “externalize responsibility and look for something to blame”. Basically, you assume, that “all these morbidly obese people, that hardly eat” are all lying, when there can be actual explanation when it comes to the condition of at least some of them… say insulin resistance?
How about you give people the benefit of the doubt, so we can discuss like adults?
1 Like
I can’t believe i read all that.
Don’t worry, it is not butter.
1 Like