Question for people that have played other MOBAs (match making)

General Discussion
09/16/2018 05:43 PMPosted by Ernestas
Please Nigma, don't talk about things you have no idea about.
He is actually right on every single post he's made in this topic tho...

People here are just like
- "If I win 5 matches in a row, I'm a good player, I'm rocking... GM here I come !!! game's balanced.."
- "Oh.. did I just lose 5 games in a row? bad MM.. this game sucks... I uninstall.. burn blizzard with torches !!"
the only other one I tried was paragon which was obviously a dumpster fire....

I have lots of friends who play league and a couple who play dota and they all say that match making is hit or miss in those games the same as here..... maybe it is a little worse here I'm not sure since I have not played the others but it sounds like it is a similar story across the board when it comes to peoples satisfaction with matchmaking.
09/16/2018 05:40 PMPosted by Nigma
The game does NOT punish you for winning streaks, it merely matches you with people with ever increasing MMR as you win more


https://www.hotslogs.com/Player/MatchSummaryContainer?ReplayID=148059508

Yeah that enemy junkrat sure had business being matched with his team and against our team. To be honest, our Gul'Dan was also an abysmal player and we had to babysit him to prevent him from dying as well but there you go.

https://www.hotslogs.com/Player/MatchSummaryContainer?ReplayID=130284339

I remember this game, that Tracer was a total dunce and had zero reasons to be matched with us. Picking Tracer after we had a Genji already picked and then raging before the match even started that our draft was bad. Then again her MMR was about 1k below the other outlier. By the way we had 2 Masters in our team so I don't know what mister silver was doing in our team but hey that player must have been better than us right?

https://www.hotslogs.com/Player/MatchHistory?PlayerID=3235331

Here's another game where we get paired with a sub 2k player while the enemy team has no one dropping below 2,4k. Interestingly enough right after a giant winning streak. Must be coincidence right? By the way I nearly out damaged that Tracer while playing Ana.

As long as MMR averaging happens the statement that the matchmaker matches you with and against better people is false. This matchmaker averages MMR and doesn't match you with and against people of equal skill.
09/16/2018 05:40 PMPosted by Nigma
The game does NOT punish you for winning streaks, it merely matches you with people with ever increasing MMR as you win more (because your MMR is rising), but because you are winning more you are matched against better players, if you aren't good enough to compete with them you start losing, if you are good enough to compete with them you'll get around a 50% win rate.


Read this research article published by EA then come back and say matches aren't intentionally manipulated. Gaming companies no longer design games around "fair matchmaking". They no longer believe it is the best way.

https://arxiv.org/pdf/1702.06820.pdf
09/17/2018 06:50 AMPosted by Lightning
Read this research article published by EA


and how much of that did you bother to read before posting it on here?

Cuz the basis of using another article for a post should include a comment from [you] on where the information is not only applicable, but on what parts apply to the discussion at hand.

Otherwise it just comes off that you skimmed for something that caught your eye and want to be all "tinfoil, tinfoil, tinfoil!" and leaves too much room to assume what you're trying to convey than ya know, you actually conveying it.
I simply love how clueless and arrogant people are. Nigma just talks out of his !@# commonly know myths and will call everyone a liar or a fool who knows otherwise.

Pro tip which might blow your mind off. When you are on a winning streak, your teamates do not become worse. They become better. Shocking, I know.
People will always complain about matchmaking in a teambased pvp game.
I've seen basically the same threads in LoL, Dota, *insert game here* forums multiple times.

"Elo hell".

But tbh QM and Unranked is really weird and care about averaging MMR way to much. In those modes people are playing together that should never meet, like golds and masters at prime times (yes, I know HL and QM MMR is different).
I've played roughly 1500sh games on League of Legends and from my experience I can tell you that the problems there are the same here. A friend of mine once even said that it seems like LoL had a forced 50% win rate. I've even found people discussing a possible forced 50% win rate for LoL over reddit.

I got tired of playing LoL for the same reason I get tired of playing this, which is that allies are frustrating to deal with. My previously mentioned friend also quit LoL after a much longer period of playing it than me for the same reason.
09/17/2018 10:06 AMPosted by Xenterex
09/17/2018 06:50 AMPosted by Lightning
Read this research article published by EA


and how much of that did you bother to read before posting it on here?

Cuz the basis of using another article for a post should include a comment from [you] on where the information is not only applicable, but on what parts apply to the discussion at hand.

Otherwise it just comes off that you skimmed for something that caught your eye and want to be all "tinfoil, tinfoil, tinfoil!" and leaves too much room to assume what you're trying to convey than ya know, you actually conveying it.


Dude...just read it. It's only 5 pages (not including title and references) and took me all of 15 minutes. I honestly don't care how my post comes off to you; the reason I didn't include specific quotes from the article is because the entire article is very telling about how gaming companies are approaching their games (and gamers) now, and people should read the entire thing. Also, the entire article is relevant to this discussion (as you would know if you actually read it too). If I was only going to quote the important parts of the article that are relevant to this discussion...I would quote the whole thing. So instead of doing that, I just posted the link to it.

But if you need a specific quote from the article because you refuse to take the time to educate yourself, here you go. Pardon the formatting, pasting from a pdf does weird things sometimes...

"Current matchmaking systems depend
on a single core strategy: create fair games at all times. These
systems pair similarly skilled players on the assumption that a fair
game is best player experience. We will demonstrate, however,
that this intuitive assumption sometimes fails and that matchmaking
based on fairness is not optimal for engagement."

"Are fairly matched games always beneficial for player experience?
This fundamental, yet intuitive, assumption is worthy of deep
investigation."

"Player engagement can be seen as an objective measurement of
user experience in games [6]. Player engagement can be embodied
by many specific metrics, such as time or money spent in the game,
the number of matches played within a time window, or churn risk.
We define churn risk as the proportion of total players stopping
playing the game over a period of time."

But I'll stop there and just encourage you again to actually read the whole thing. This article is very telling about how gaming companies view their gamers. They see us as complex math equations, and instead of creating good games to engage players, they are resorting to research into psychological manipulation to engage players. They have completely lost sight of the heart of gaming in their pursuit of greed.

Why did you assume I haven't read this article? It's literally a 15-minute read and the information it contains is more than worth the time.
09/18/2018 02:59 AMPosted by Lightning
<span class="truncated">...</span>

and how much of that did you bother to read before posting it on here?

Cuz the basis of using another article for a post should include a comment from [you] on where the information is not only applicable, but on what parts apply to the discussion at hand.

Otherwise it just comes off that you skimmed for something that caught your eye and want to be all "tinfoil, tinfoil, tinfoil!" and leaves too much room to assume what you're trying to convey than ya know, you actually conveying it.


Dude...just read it. It's only 5 pages (not including title and references) and took me all of 15 minutes. I honestly don't care how my post comes off to you; the reason I didn't include specific quotes from the article is because the entire article is very telling about how gaming companies are approaching their games (and gamers) now, and people should read the entire thing. Also, the entire article is relevant to this discussion (as you would know if you actually read it too). If I was only going to quote the important parts of the article that are relevant to this discussion...I would quote the whole thing. So instead of doing that, I just posted the link to it.

But if you need a specific quote from the article because you refuse to take the time to educate yourself, here you go. Pardon the formatting, pasting from a pdf does weird things sometimes...

"Current matchmaking systems depend
on a single core strategy: create fair games at all times. These
systems pair similarly skilled players on the assumption that a fair
game is best player experience. We will demonstrate, however,
that this intuitive assumption sometimes fails and that matchmaking
based on fairness is not optimal for engagement."

"Are fairly matched games always beneficial for player experience?
This fundamental, yet intuitive, assumption is worthy of deep
investigation."

"Player engagement can be seen as an objective measurement of
user experience in games [6]. Player engagement can be embodied
by many specific metrics, such as time or money spent in the game,
the number of matches played within a time window, or churn risk.
We define churn risk as the proportion of total players stopping
playing the game over a period of time."

But I'll stop there and just encourage you again to actually read the whole thing. This article is very telling about how gaming companies view their gamers. They see us as complex math equations, and instead of creating good games to engage players, they are resorting to research into psychological manipulation to engage players. They have completely lost sight of the heart of gaming in their pursuit of greed.

Why did you assume I haven't read this article? It's literally a 15-minute read and the information it contains is more than worth the time.


Don't bother man. Even if he read the article he probably wouldn't understand the implications. I actually read this article a while back after a Youtuber made a video about it. It blew my mind how far down the rabbit hole gaming companies are going to get players engaged. Like you said, they aren't interested in making good games anymore. They don't care about fair games (as this article points out). And they are 100% willing to rig your win/loss rate and actually believe that doing so will make players more engaged. This article, in a nutshell, makes the claim that there is a specific ratio of win/loss that makes players more likely to keep playing (see Table 1), and that artificially creating this ratio by manipulating matchmaking towards wins or losses will keep players more engaged than just creating fair matches.

It's actually a bit disgusting to read.
And don't forget how EA and ActivisionBlizzard are going back and forth patenting various matchmaking algorithms to screw players in one form or another ..

EAs patent is "matchmaking that drives user engagement"

ActivisionBlizzard's patent is "matchmaking that drives microtransaction revenue"

https://www.google.com/search?q=patent+matchmaking+activision&rlz=1C1SQJL_enUS794US794&oq=patent+matchmaking+activision&aqs=chrome..69i57j0l2.7754j0j7&sourceid=chrome&ie=UTF-8
This should go to show just what's going on behind the scenes with this stuff.
Teams of data analysts and psychologists are working for these publishers and their jobs are to find ways we can be manipulated by the software, it's no longer about making a game that is fun to play - the software has a primary purpose now to try and get you to spend as much money as possible.

And for all of you blizzard white knights that endlessly and zealously defend this stuff with excuses like you can just ignore it all, or "tinfoil hats!" they are literally using match making to manipulate you to buy microtransactions

These links/pdfs/resources align with exactly what players are experiencing during match made gameplay.

It's high level manipulation of the play experience in undetectable ways all with the single goal of increasing player spend.
If I told you I was always paired with players of my same skill level in LoL, I would be blatantly lying.
MMR here is skill based with some composition rules (healers/tanks) at both sides.
Engagement based MMR is a study subject that aims to make sure you are having fun and, therefore, keeps you playing the game while buying stuff you like.
If you are not having Fun, would you keep playing?
The article is good, but is simulation based (of real data that probably uses skill Based MMR!).
Patents can be used to protect you from being legally processed.

In regards to W/L, what people observes here is Regression toward the mean.
https://en.m.wikipedia.org/wiki/Regression_toward_the_mean
Hotlogs must be used carefully as it contain many false conclusions. Sampling bias, as pointed by Nigma (in his particular way, lol) is a good example.
Something that many ppl forget is the MM is based on MMR + ranking (bronze, silver...). The ranking makes a udge impact on the MM which produce a lack of volativity.

For example, during a bronze to master, the MM didn't manage to set a match because the GM who played in gold league has a plat or diam MMR due his 99% winrate. The MM didn't find player with the worst MMR in gold to associate with him vs best MMR gold players available.

The rank freezes the MM, it increases the feeling of 50%. If you have a win streak you will be associate with worst players of your league (+/- one div). If you have lose streak you will be associate with best players of your league.

Of course a GM will win his match in gold even if he is associated with bad player but for a normal player it will be almost impossible.

This is not relevant for Master or higher as there is no league for them.

Overwatch has not this system, your MMR defines your rank and only your MMR is taking in account.
A lot of ppl wrote a lot of smart stuff about why the MM or Blizz or HotS itself doesn't punish you so I just leave here my personal experience:
09/16/2018 06:07 PMPosted by Zarbane
In HoTS when you win consecutive games after about 3-5 wins the game starts to match you with teammates that are worse as well as opponents that are better.
I have a 54-55% winrate in this season over 200+ games and I had 8-12 winstreaks without getting from those so called "potatoes" to make me loose. Winstreaks just end. End of story. You can't win forever and ever and as others mentioned, you see others' mistakes more clearly so you assume they're bad/dumb which probably just isn't true.
Hots itself had a lot to do with it. It's easy to accept the end of a winning streak after you just got blasted out of lane and went 1-6-6 by someone a rank above you.

In Hots it's very easy to start losing and getting outplayed more even if your statistics remain outwardly the same.
09/18/2018 02:59 AMPosted by Lightning
Why did you assume I haven't read this article?


Because I don't trust anyone to read anything when they can simply pretend they did. Furthermore, what you get out of the article is very different than what anyone else would get, so having what caught your eye on it is better to convey for actual discussion instead of relying on assumption, to which you do.

here's the funny thing: I read the article and that's why I think you didn't. Given your choice of summation is pretty much early on, I think you chose to fixate your bit onto the early parts of the info because the basis of what you got out of it is a few select terms, to which you then skimped the rest because it already suited what you wanted to find, and then you can just assume the rest, as you evidently did with me.

While the phrasing and use of the article itself is 'disgusting' the thing is, its not 'wrong'. I wrote pages and pages of replies back in the alpha in of this game — most of which will be 404 but still in my post history XD — because of the semantics of "fair". Players don't agree on what's "fair", but they act like they do and just assume all the other particulars to suit they nay say.

However, a key thing of an research article is looking at the particulars of who did the test, what they found, and what they can do with it.

The test there is a side project fun by an intern to see if a simulation of replays of player data could be 'optimized' by shifting the 'engagement' particulars to theoretically get more out of players. So its an assumption on an assumption, and how much did it improve the yield of the figures given?

Given the basis of what system they used to ascertain the skill metrics for the replay simulations contrasted against the system they used to try to 'optimize' the matching is an inconsistent lump and probably a hefty reason why its the intern doing this and not a more grounded research team. Granted, for a company, their goal is to optimize 'engagement' which would mean getting people to feel better about playing the game, and thus spending on it, but realistically, if people are going to be more 'engaged' with a game, that's a better outcome than just hitting a quota and being done with it.

Esp given the data of that article was 1v1, maybe 2v2, and only had notes regarding the issues of handling multiplayer — which ya know, has bigger variables to deal with.

asking for your take on it lets me know where you're going to read into this, wether it be the spirit of the matter cus of "optimization" problem, or on the specifics of the test, cuz the whole basis of any gameplay experience is "manipulation" that's part of how games work in the first place. With the flood of content being accessible, companies want people to remain invested in their projects, and given that "fair" is an argument in semantics that doesn't appease anyone cuz they don't bother it think it through, companies should look at a better way to set their matches.

a key part of that article is how it event defines what is 'skill' to use for how it evaluates its matches, as that's easier to do in 1v1 games than it is in larger ones, so if the manipulation evident for 'optimizing' the gameplay is going to work, companies are going to have to better define their metrics for that in bigger games, or its going to be a crapfest that undermines the 'optimization' they're trying to get.

Or as game trends come and go with crowds genre jumping, some of that might not even matter. Funny part of small studies run by interns? It may not even be worth their while to do more with it. That's kinda why some groups use interns to run simulations instead of other ways to test the 'optimization'

But sure, I'm going to assume you just read into all of that, and not assume that the reason you were linking the article was to not be all 'woah is us, company with a bad rap is showing why they have a bad rap, and implied logic jump'.

If the fundamental core of your 'competitive' playerbase is 'bad' at the game, is their any amount of 'fair' metrics that can be applied to make the game seem compelling for anyone?

I'm not suggesting complacency, but if people aren't going to be arsed about the particulars on how they want to whistle blow to anyone else, then they're probably not in a position to do much more than qq about what they think may be the boogie man in the closet.
09/18/2018 06:54 AMPosted by Goiaba
MMR here is skill based with some composition rules (healers/tanks) at both sides.
Engagement based MMR is a study subject that aims to make sure you are having fun and, therefore, keeps you playing the game while buying stuff you like.
If you are not having Fun, would you keep playing?
The article is good, but is simulation based (of real data that probably uses skill Based MMR!).
Patents can be used to protect you from being legally processed.

In regards to W/L, what people observes here is Regression toward the mean.
https://en.m.wikipedia.org/wiki/Regression_toward_the_mean
Hotlogs must be used carefully as it contain many false conclusions. Sampling bias, as pointed by Nigma (in his particular way, lol) is a good example.


more "bla bla bla its all in your head everything is fine with the match making system" nonsense

Join the Conversation

Return to Forum