LMAO
Get a clue.
llllll
‘Balancing matches’ is the definition of handicapping, so you are either being ignorant or illiterate. I wrote all about this in the original post, please go read it.
The original post was made over 2 years ago and had links to developer quotes. I lost permission to share links when it Blizzard changed their forum domain.
Yes! Thank you
1 Like
Anyone who reads Scott Mercer’s original post can SEE that you 100% fabricated the entire idea of 50/50 meaning intentionally handicapping “good” players with “bad” players. Reading context is actually a thing, which you intentionally do NOT do.
Show the exact line in that topic that states that “good” players are handicapped with “bad” players.
Maybe you should answer this for me then: Do you think it’s acceptable to have a team of 6 solo-queue players ranging from SRs 1800-2000 vs a team of a full 6-stack at 2300-2600? And since that’s obviously a pretty extreme example, what about a full 6-stack vs 6 solo-queuers even though all 12 players in that game are equally ranked?
1 Like
there is no way to have a game that has 50 50 projected win odds if people have varying ratings other than what effectively becomes handicapping
in a 2v2 game, with 4 people selected with ratings 2300, 2400, 2450 and 2700, the 2700 player and 2300 player will always end up on the same team if game is trying to synthesize a match with projected win odds as close as possible to 50-50 or whatever possible least mismatch between the two is
1 Like
This is correct, yes. But you will NOT see that unless the 2700 and 2300 players are grouped together, OR the online player count is so low that it has to do that. That’s how you got the XQC and Bronze on the same team clip on Twitch that one time. He was not thinking straight and immediately queued for comp the very second after maintenance was over, and the game he ended up in was basically the very first 12 people after maintenance to press the “Competitive Play” button on the menu in the NA Region.
What Cuthbert has been saying is something totally different. He has been saying that you have your SR, but you also have a hidden MMR that differentiates between “good” and “bad” players at the same SR. While it is correct that you have a hidden MMR, it is 100% false to say that MMR is a differentiator between “good” and “bad” players at the same SR, considering everything Kaawumba compiled into his topic about Competitive Matchmaking (How Competitive Matchmaking and Rating Works (Season 19)), and especially in light of the video that reveals that MMR ranges from -3 to 3 (technically you can go slightly below -3 and go above 3), and that those are standard deviations from the mean rank of 0.
Where the SR/MMR decay talk starts, but the relevant part is at 1:13:30
I mean, yes, that’s exactly what MMR is
Someone at 2700 SR who is on a 10 game losing streak likely belongs lower than he does and yet his SR will adjust very slowly. His MMR on the other hand might be moving in much larger increments depending on how the algorithm is implemented. So it’s entirely possible that game adjusts MMR too rapidly and that results in you having teammates with MMR in the other direction which results in you feeling very powerless and teammates being very very bad.
considering everything Kaawumba compiled into his topic about Competitive Matchmaking (How Competitive Matchmaking and Rating Works (Season 19)),
There is absolutely nothing in there that in any way disproves this theory in any way. + that guy is operating from the perspective that the system is fair and pretty much only tries to find data that supports his narrative (hint, he doesn’t, he just interjects his own opinion as fact at crucial moments)
No, Cuthbert is incorrect. MMR is NOT a differentiator between “good” and “bad” players at the same SR, and his evidence for that statement is non-existent at best.
You didn’t read that topic then. Never mind the topic for now though, watch this video first:
In particular, watch and listen to the segment at 1:13:30, where Jeff states that MMR is a value from -3 to +3 (it actually can be slightly below -3 and slightly above +3). The highest ranks at GM are +3, so you can guess that the lowest ranks at low Bronze are -3. As explained in Kaawumba’s topic, your MMR is your rank compared to the standard deviation from the mean (MMR of 0, so somewhere around the mean SR of around 2400 or so, whatever the mean actually is), and that’s why you have a public-facing SR, because it’s simply easier to read SR 2500 than read MMR +0.25 or SR 4750 instead of MMR +3.025.
2700 player with MMR of 0.91 is a better player than a 2700 player with a MMR of 0.75
So it pretty much is
SR is irrelevant
SR is just a pretty number that may or might not coincide with MMR
MMR is the real rating and as such is likely prone to more variation and adjustments than SR is and as such would be inadequate to display to the user
It still doesn’t disprove what Cuthbert says here in any way, shape or form
As long as game tries to make even teams, it means that handicapping is occuring
If I start to smurf in bronze and then try to climb to diamond, I will start off easily and several games after that I will start to get matched against cheaters or other smurfs doing the same thing in an effort to make the games even. That should not be happening. It breaks the point of ranking up. You can even see this in some Bronze to GM runs where player occasionally gets completely unwinnable games and these should never happen unless handicapping is occuring given that the player has 2500 SR more than his actual rank is and chances for a loss are minuscule. All of these suggest that MMR is adjusted more rapidly after streaks which likely results in the fact that you can be at 2700 SR and have a MMR that is noticeably above that which soon results in you getting paired with potatoes and that results in loss streaks because if you can’t hard carry, you will lose and the fact that your MMR is still high, there will be a series of these handicapped games until your MMR swings the other way around at which point you might just go AFK and you will still be winning because reverse handicapping will start to occur.
As for Kaawumba, nothing he says in there actually disproves or debunks this theory. And he does make a leap of faith when it comes to certain points.
It’s up to debate whether this “handicapping” is occuring intentionally or is it just a pure artifcat of a bad matchmaking system, however, Blizzard has every reason from a business POV to actually implement it intentionally and they would never admit that it’s actually there. And if it somehow got revealed, they would claim that it’s not intentional and that it’s just an artifacting effect of their algorithm, not an intentional one. Blindly listening to what a developer says (who also might be engaging in very carefully crafted doublespeak) while also being a firm whiteknight of the system (which kaawumba seems to be) is probably not the best course of action when it comes to assessing the truth here.
1 Like
The burden of proof is on the handicapping believers to show where that exists. I know people like to throw the word “white knighting” all over the place, but there’s a lot of ego and immaturity coming from people that play video games in general, let alone people who play multiplayer games over the internet where you do have a veil of anonymity.
For the record, you might as well say that I’m “white knighting” as well, because I don’t believe that your supposed scenario of “2700 player with MMR of 0.91 is a better player than a 2700 player with a MMR of 0.75” exists. If you have the same SR, then you have the same MMR.
Again, burden of proof is on Cuthbert. Games being 50/50 simply means you play with your own ranks, and reconcile groups. That unless, you think that a 6-stack of Diamonds vs 6 solo-queuers that are Silver to Gold makes an acceptable match.
Nope. You can go the entire way without seeing a cheater or a smurf even if they all queue at the same time you do. That is, unless you’re streaming it and they happen to be stream snipers or it’s set up. You can even get multiple smurfs on the same team vs legit ranked players. The matchmaker only cares about your current rank.
The leap of faith is 100% Cuthbert. Again, it’s an ego/immaturity thing combined with the internet.
Been there done that. Two accounts at super close SR (difference less that 12 I think), same network, same exact time for queuing, but only one got the long time for a match warning. Thus mmr =/= SR to the level you believe.
So yes, proof to the best that can be shown has been done.
Just quoting this as one of the best explanations of the system i’ve seen.
1 Like
I think one thing that’s pretty obvious is that MMR is being used to handicap specific players to ranks, particularly streamers. Once the MMR handicap was taken off, they were judged fairly among their peers, and were placed in Plat and Gold ranks. They of course, didn’t like being judged fairly, so they threw temper tantrums at the idea that they can’t stay in Masters and GM if they play like Gold players.
Objectively speaking, MMR is what’s causing disproportionate matches. It’s why smurfs can start so low at Bronze and take ages to climb up through the ranks. All it does is judge you based upon your wins and losses, which is a very poor metric to rely upon in a team based game.
Other than showing you that MMR exists, is hidden and is not 1:1 equate to MMR as MMR and SR have separate but important differences in their algorithms.
Jeff “it’s possible to win a match and not gain any MMR. We make it so that if you win a match, you always gain SR” -‘Overwatch Forums’
This shows us that for a given SR an MMR might be LOWER or HIGHER than the digestible SR we are given.
It might be only ~50 to 100 SR equivalent difference, but that alone tells you that there are people that sit at an SR that are MMR wise Better or worse than someone at the same SR.
The burden now shifts to you to point out where they explicitly counter this and state that MMR is exactly SR.
Actually, I think wins and losses are the best possible metric to score performance, in a team-based game. But not when matches are being systematically handicapped.
No one’s denying that MMR exists. In fact, I just showed a recent video featuring Jeff Kaplan himself explaining that MMR ranges from -3 to +3 (actually you can go slightly below -3 to slightly above +3).
Yes, and there were situations where this legitimately happened. If you watched all GM streamers during Seasons 2-3, you can see them gain a legit 2 SR on a win without the bugged SR gains. That was because matchmaking was looser at that time and many times, there were games where you got a very high GM on one team with Diamonds and low Masters as teammates vs teams of mid-Masters, in particular when they played at non-peak times when the rest of the GM playerbase were not playing.
That being said, do you really think that a Plat player at 2600 whose entire team is also at that rank, taking on a team of all 2600s, can gain 25-30 SR while NOT gaining any MMR at all? A ~50-100 “difference” is literally a matter of 2-4 games, and MMR isn’t even that in the first place. Do you think there exists players with a “difference” of what would amount to 400 or more?
No. I’m still waiting on you or Cuthbert to show us the line where Blizzard states that they handicap “good” players with “bad” players.
Show us the line/lines in this topic:
Mainly this topic because this is his SOLE “evidence” of “handicapping” and it’s non-existent at best.
See patent quoted above:
So why didn’t it kick in for this player?
Or why didn’t it kick in for MallSanta when that situation blew up on the old forums?
The patent you reference has everything to do with microtransactions. Are you telling me that they would have stayed at their ranks if they bought lootboxes?
You still need to address the deranking and boosting services. Also the fact there are different ranks in the game and that there are players in all different ranks.
“Good” and “Bad” players are relative to the match itself. The very fact that the system is designed to make 50/50 matches is the method to which the MMR makes a match where each side is equal. This is handicapping.
Your bias is in thinking all 12 of the people at 2600 are of equal MMR.
People on the forums talk about 8 game streaks, thats 200 in one session. Add to that the factor of 50/50 may push you further and further away just on coin flips and the fact that streaks occur naturally. It is not unfathomable to state that someone could be ~400 off their base over the course of a few bad days.
You are too stuck on this one topic that was quoted at the beginning of all this. Just because Cuthbert used this as the original springboard, doesn’t mean “handicapping” is non existent.
Though it clearly states the 50/50 match making. You just don’t like that quote.
I’m not sure what you want with matchmaking. We keep going over the entire 50/50 semantics. I asked this multiple times already, but do you think it’s acceptable to have matches where you have a 6-stack of Diamonds and up taking on a full team of solo-queueing Plats? Before you tell me that I’m strawmanning Cuthbert’s arguments, Cuthbert was the first to strawman Scott Mercer’s arguments. And going by some of the other posters on this forum, it seems like there are a lot of people that really do want to rank up to higher ranks but do so by playing stacked matches in their favor rather than playing with and against people of whatever rank they currently are in.
And yes, people deal with winning and losing streaks with matchmaking. You’re playing with and against other human beings. I’m not sure what you’re supposed to be expecting, if you think everyone should be able to “rank up” or if you honestly think that Blizzard intentionally “handicaps” players rather than people honestly trying their best to work with other people despite their differences.
Maybe instead of thinking that Blizzard intentionally gave you bad (or good) teammates for winning (or losing) too much, try working with other people, and use VOD reviews to honestly see what went wrong.
Reporting in.
I don’t play Support often. But i decided to queue up with my support duo partner and try it out. So, we’re both supports all the sudden.
We really surprised that matchmaker! Which usually just tosses the worst support it can find on our team to balance out my duo partner, who’s a great support. Can’t have a great 2nd support, even though they are out there, or it wouldn’t be an “even match”.
Long story short, we won 4 in a row. Easiest games in ages. Know why? Becasue we surprised the matchmaker, which just gave us the typical games. We are better than that, so it was easy.
Then, exactly as we predicted, the system caught on and figured we needed a thrower in our team, then a leaver.
Suddenly, two auto losses.
And that’s how Blizz thinks they can force even matches. By having the system find horrible teammates to balance out any anomaly.
So yeah. It’s in line with exactly what we’ve seen for 4 years. Handicapping matches, holding players back from ranking up too fast, allowing idiots to ruin people’s games.
By the way… the dead giveaway is longer queues. We had very quick queues at first, because the game didn’t know what to make of us being a support duo. Now that it wants to balance things out, it has to search harder to find idiots for us to play with. Always happens.
Edit: Wow. Yest another crap team. Three in a row, and i KNOW the reason is because we too too big, too often. Unbeilvable how fast this matchmaker canceled us out. We mix it up, take it by surprise. It sees that it’s “unfair” cancels us out within 4 games.
What SHOULD happen is… we rank up!
Instead, we’re artificially stomped down by the MATCHMAKER.
Horrible idea, Blizz. Glad your game is dead.
3 Likes