I think the game is trolling me

General Discussion
21

Wow, so a newbie player was placed into a game with higher level players, probably due to long queue time or something?

Listening the game splayed on hotslogs for these accounts might help establish more info.

Treating Hotslogs MMR as accurate okay.

Well, I’ll just be looking at my Silver practice account for bad heroes has a higher MMR has a higher TL MMR (2147 v. 2268). Despite me hitting Diamond multiple seasons on TL.

And has about 400 higher QM MMR, about 2600 versus about 3000…

But yeah, hotslogs works great. Silver = Diamond, yo!

What does solo queue have to do with anything?
What was the gamemode anyways?
Why wasn’t it said!?

looks like Hotslogs either has given you an inflated MMR, or the rest of your team has a deflated MMR.

Matchmaking is not perfect, wow!
Obviously that means it’s forced 50%.
Good logic!

No, not at all. I already put an example up earlier.

And with help of a friend, looking again:
A Gold account with 300 higher TL MMR than my actual account.
(I should also note, while TL is abusable, I’ve been in Diamond every HL season I’ve tried on main.).
And has 200 higher QM MMR.

It’s completely worthless. But hey, someone who placed Gold in HL multiple times and silver in TL == Diamond in both. And with a higher QM MMR to boot!

Hotslogs MMR isn’t helpful, even with 100% games uploaded simply having FAR fewer games with a higher winrate causes MMR to skyrocket.

Yup, my Gold friend who I’ve helped learn some heroes is 200 MMR higher in QM and 300 points higher in TL.

In HL the MMR is accurate, for one. Where I’ve got about 500 points more MMR than them.

22

I didn’t say it was forced, I said if Blizzard wanted to stop making people think it was forced they could stop creating garbage matches.

That’s ok though, the lowest MMR players just happen to be the lowest performers on the team because the whole system is just completely inaccurate. Their score is completely inaccurate too, my bad, I should have realized.

23

Sample size of 1, sorry, 2, is a great way to do things yeah!

My point is that hotslogs MMR is hilariously inaccurate. But hey, you know, my practice account totally is top 200 QM players. Despite constantly getting in QM games with people who are silver/gold ranked if you check after game.

(Thats’s where it would be if I had about 80 more QM games on it keeping same MMR).

It is inaccurate most of the time, or to a large degree.

That 1200 MMR person might have been the lowest/worst, but they could easily have been 1400 instead. Still the worst, but much less worse than stated overall on their games.

Yes, because MMR being inaccurate means that their scores in game taken directly from the replay are inaccurate.

Actually, hotslogs still says people are taking Assault Strain on Abathur… somehow.

24

It’s not a sample size of one, I have a sample size of like 12 at this point.

But it doesn’t matter, you’re going to deny what the system is actually doing regardless of what I say.

LOL, it still doesn’t matter, they’d still be the lowest MMR player on the team and the spread is still over 1000 MMR.

I don’t care if they’re 1100, or 1500, or whatever their MMR is. It’s still vastly lower than mine, and mine is accurate because HOTSLogs has my ENTIRE match history.

25

The point of concern that I brought up here is that you’re trying to shoot for confirmation bias and use information that doesn’t suit the demands of what you want it to be to try to make claims it cannot.

The fixation you present is assuming the crux of the matching is on you personally and then derive information only about yourself to suit the claim of that demand and thereby ignore any other relevant information in the matchmaking to try to argue confirmation bias.

I’m not acting like there’s a ‘magical explanation’, but to people who deny possibilities, then yea, stuff suddenly seems like ‘magic’ when it comes to explanations they don’t want to hear.

The causation and correlation of what you want to deride (bad people in your matches) is coming from MMR appraisals from a system that’s not only optional, but is bad at anything beyond personal progress and sweeping generalizations from large sample sizes.

If you were so concerned why “why blizz matches” then you’d be posting information from the game itself:
-account levels
-match history ( win/lose streaks may shift confidence)
-be explicit about team pairings
(HoTS adjusts mmr estimations for groups in ways HoTSlogs doesn’t)

If hotslogs shows 8 or 40 games for people that had more than that, then the apprasial isn’t going to be great.

Similarly, if they seeded for hots confidence a year ago, and only just came back to the game, then that’s information relevant to their skill level that mmr estimations for quick match that likely aren’t going to be shown by looking at stat pages from HoTSlogs.

The point being: Hotslogs has lots of errors and doesn’t tend to paint a reliable picture for those looking to scrutinize particular details from it. eg: on 3/27 I lost 160 mmr points for Qm because of the “season transition” that ya know, doesn’t happen for QM.

Estimation jump hit all sorts of ways on these transitions when people don’t have a game played to show for it, but its part of the corrective assumptions Hotslogs makes to suit the demands of seeing matches that are made where it doesn’t know if it exceeding queue time perameters, had team mmr adjustments, and a number of other details that do influence the actual match maker.

and none of that has anything to do with whether or not these players should be matched, but rather the point is the sort of resources you’re trying to use to make a claim as a means to circumvent the stuff that actually does influence that sort of thing.

If people want to point out issues of the flaws of the system, then they should use the evidence from the system for those claims instead of something else that happens to suit what they expect. Its like the joke/story of the guy who’s looking for his wallet under a streetlight, not because that’s where he lost it, but because the lightning is better there.

a lot of these mmr-rant topics generally ignore the details of the other players in the loop to figure out how much of an outlier something might be, or other factors to influence it – maybe the lemon in question had a good streak, maybe they had a bad, maybe the mmr over estimates are influenced from another year, mode or the like.

But that’s not the sort of details people are going to convey cuz they don’t care to look into those sort of details. But the trouble is, having more details helps because there’s plenty of examples of the 1500++ lvl accounts that play like they’re never touched the game themselves, and then there’ s the gm smurfs with lvl 1 rerolls that stomp qm games for whatever reasons.

The functional concern is expressing details pertinent to the match making if your concerns are about the match maker rather than trying to rely on reactive estimations that sometimes toss in or take a couple of hundred points to simulate adjustments it wants to account for existing, but can’t apply on a game-to-game basis.

1 Like
26

Does it also have the full match history for everyone you’ve ever matched up again, and every match they all have and every match those players have had? And so on.

No?!?!?

Well, it’s not accurate in that case. It is one of the less inaccurate MMRs, probably. But less inaccurate != accurate.

Just like how because I’ve got all the games on my practice account doesn’t make it’s MMR being in the top 200 players on QM ranking accurate.

27

I can’t believe that people will still defend to the death blizzard’s matchmaking…

3 Likes
28

It’s probably why Blizzard won’t show us MMR. They know full well they matchmake like this, and showing us the internal MMR would just prove all of the outstanding criticisms lobbied at the system.

I’m also tired of the “HOTSLogs is inaccurate” argument. Yes, it’s inaccurate if you’re looking for an EXACT value, but it’s accurate within a degree of certainty and it’s certainly not going to be off by numbers on magnitude of the thousands.

But no, HOTSLogs has to be 100% accurate before someone will accept any arguments that use it. Guess what, we’re never going to have those numbers, HOTSLogs is the best tool we have. When literally every game I can point out the one person who threw it, and their MMR is like 1000 below mine, I know damn well that the game is intentionally making matches lopsided.

Not only can I see the skill disparities in game, I can also measure them with HOTSLogs. They exist.

2 Likes
29

Don’t worry, Blizzard is working on improvements to the matchmaking system to prevent these sort of lopsided matches from happening in the future. Or maybe they’re working on a new event with some skins you don’t care about, whatever the point is this: matchmaking will be fixed… uh, sometime soon, maybe. Or never. But skins.

30

I’m pretty confident that new player protection has been disabled. I have seen more < lvl 50 accounts in the last week than I have seen in the last 2 years of playing Quick match.

2 Likes
31

Probably, that would make the most sense, but they need to start new accounts lower in MMR.

1 Like
32

if your concern is within this topic, then you’re demonstrating a lack of reading comprehension: the capacity for people to show concerns on how hotslogs shouldn’t be used is not a defense of the blizz matchmaking.

What posters like yourself, and likely droth, are doing is conflating any form of disagreement of what you ‘want’ to thus mean its an entire disagreement with the extent of what you’re trying to argue.

Creating a strawman to argue doesn’t improve the position of the point people want to make, if further egasterates the issue on their fixating on unreliable information to make claims unsuited for argument they want to assert.

That’s part of why there’s literally no denial regarding confirmation bias, posting new information to back the claim, redirecting the concern through a different perspective: its their automated assertion that they’re “right” and anything else must then be ‘wrong’ or “not make sence” and it doesn’t “make sense” because people are too busy looking for disagreement to care about details and particulars of concern to hidge this in a false dilemma of two possible outcomes.

The defense “its the best we have” is a impulsive response be people who don’t bother to look at other pertinent details, esp if they potentially conflict with what the person what’s to claim.

Here’s a very telling aspect of the crutch of hotslogs assertions: Droth wants to assert the accuracy to make topical claims (for hardline assertions, not generalize trends per statistical analyis) and wants to evaluate and judge other players… but hides their own profile.

Now, the full extent of what I imply of that isn’t to say droth, or anyone specific, is going to do the “oh you’re rank [whatever] therefore you don’t count” but… the capacity to try to make the inferences of the data they want to so do comes from, to some degree, the involuntary availability of that information from anyone else for him to judge accordingly, but is otherwise unwilling to be available for that same standard.

Part of the issue of the “inaccuracy” of hotslogs, and those that are “sick of that being argued” have a (from my ‘research’) strong tendency to disregard other information, judge people with double standards, lump ‘disagreement’ into overgeneralization, and then assert strawman as a means to ignore most anything that doesn’t agree with them.

and its not like i have some magically grounded study in this sort of thing, but there’s plenty of examples of players since alpha who did they exact same things, from the lyra/animus arguments, to lightnings forray into statistical inconsistency, to droth catching onto mmr ‘averaging’ now: a lot of these all break down into repetitious arguments of people so keen to point out the “best we have” as rationalization to deny any other pertinent information that doesn’t suit the claim they want to make.

Its disingenuous conduct by people content to argue ‘holier than thou’ to justify confirmation bias despite the evident lack of actually experience in evaluating the statistical information they essentially want to swear by.

My pointing this out has zero bearing on the quality of blizzard’s matchmaking – in case that evidently wasn’t apparent. In math, how one arrives to a conclusion tends to be more important than the conclusion they draw and that sort of notion applies to stats, argumentation, and so forth.

But the ones that claim to be some of the most ‘data driven’ tend to also be the quickest to deny problems with their methodology cuz well… that’s how confirmation bias works.

2 Likes
33

If someone doesnt know what button hearths their MMR is 200. It has been happening since game began which is why the game died.

Hotslogs is more than accurate enough to support the arguments people have been making about its broken AF matchmaking. Everybody knows a sub bronze player when they see one, but when you see their MMR hovering around whatever # that is way higher than they should be thats the only time we need to bring up inaccuracy… which in truth is close to how Blizzard’s MM works… hence the problem with the matchmaking itself.

Blizzard’s system of assuming a newbie Raynor player can play with Diamond+ players just because he logged in for the day is what is inaccurate. Not hotslogs. Always has been and always will be … hence why the game is daed.

1 Like
34

Your concerns are baseless. There is only two ways currently to divulge the winrate, playrate, and overall data of heroes.

Intuition and Hotslogs. And the latter has value. Even if hotslogs isn’t exactly blizzards algorithm, it could very well be close to it. It still provides the general data and winrates of people so it shouldn’t be hard at all to guess mmr.

So if hotslogs is as inaccurate as you say, can you provide any other options at all? Compared to nothing?

35

There is no other option, HOTSLogs is the only tool of empirical measurement we have, and the only argument put forth against any evidence that games are this lopsided is “HOTSLogs is inaccurate” nevermind the fact that the average MMR on both teams seems to be equal in nearly every scenario so it’s pretty safe to hypothesize that Blizzard isn’t attempting to put players of similar skill together, it’s trying to average skill out which has not, and will not ever work. The team closer together in MMR usually wins.

1 Like
36

Please not this again.

Hotslogs isnt accurate! Get that through your head! Geez…

I have accounts i never uploaded any games, but some that i have been playing with did, so you could see 1/10 of the games i played on them.

This very account never uploaded anything in the first half year, then uploaded a lot in the next few months and now again nothing. You think it would reflect to anything at all?

Yes, the MM can do lots of bad things, and not good at all, but any argument based on Hotslogs data is a joke and should be thrown to the bin.

37

I just get it, its QM isnt it?

Then any MM complaint is a joke, cause i belive QM has a MM like Brawl.

And btw who writes a post about their QM games? I love the style tho: “they throwong harder than i can carry them” - just to make everybody sure you are the guy who carries others.

Welp, maybe you are in these games because you deserve to be there, therefore you need to “carry” games to climb in the imaginary QM ladder. Good luck!

38

They reduced protected games from 100 to 10 i believe.

39

Do you know how bad 1400 is? 1200, 800, 1400, 1600… same thing. Terrible player.

40

Well, Hotslogs said I was 1457 or so when I got placed into Platinum 2 (note, if the next 25 or so games I climbed my MMR to 2350…)

Oh and I should note, my practice account, the one that is supposedly in the top 200 QM players per hotslogs? With nearly 3000 QM MMR… It says someone who has been silver last few seasons is near my skill level when queuing for QM. Their QM MMR on hotslogs is a bit over 2000.

So nearly 50% more, or 1000 points different in Hotslogs, but near equal skill in the actual game. Which one do you think is more accurate?

Team League MMR as an indicator is also flawed, 1800 v. 2300 (which is higher than my main account btw, despite being in silver instead of diamond).

So 500 MMR is close? Either way, it makes this stuff about hotslogs being accurate hilarious.