I think we got a glimpse of Blizzard’s true character at the last Blizzcon, where they announced their mobile game development plans, to the chagrin of their almost entirely PC-gaming fan base.
Blizzard doesn’t care about gamers. Blizzard just cares about making their products as ubiquitous and addictive as possible, to generate revenue. They are like any other mindless corporation, chasing maximum profit from quarter to quarter.
I understand your decision to give up on this fight. Thanks for all the input you’ve given.
All of my statements are firmly grounded in a direct quotation from Principal Overwatch Designer Scott Mercer. The statements of my detractors are grounded in nothing but wishful thinking and fanboyism.
Where does Mercer talk about “handicapping” in that topic? We can see your direct quotation of him starting in paragraph 3 of the above linked post, but look at the context of the entire topic, starting from the opening post, all the way to the end of the topic. Where’s the “handicapping”?
Thank you. Evidently countering the arguments of fan boy’s and white knights is against the TOS. Inconvenient truths are considered “insults”. You can’t have meaningful discussion when you censor dissenting opinions.
Blizzard has shown it’s true colors with the Diablo Debacle. They’ve done everything they can to silence the voice of the masses. I’m wondering if censoring YouTube vids might come next.
With this statement "When the matchmaker creates a match, it determines the % chance for each team to win based on the match it made. The VAST majority of matches are usually near to 50% " - From your link
MMR is a mathematically driven outcome as that is how computers would look at it.
If MMR is such, and there is a “median” player in the system. How then do you interpret a system that calculates a 50% win prediction, without averaging both teams’ MMRs?
With a median or average players, there must then be outliers to the system. If those outliers exist, then in a game, those outliers must be placed onto a team that will average out that outlier in the group.
If you take that, then there must be manipulation within the teams to meet that 50% win chance it is looking for.
Therefore, the outcome of a team is modified to incorporate that outlier.
You need to understand the context of that topic I linked, because that entire topic (including the Mercer quote in Mercer’s actual post) has nothing to do with Cuthbert’s argument that there is systematic handicapping where the matchmaker knows which players are “good” and “bad” at a certain SR and thus distributes the “good” and “bad” players evenly.
The entire crux of my argument is that Cuthbert is using IRRELEVANT information as evidence of his idea of handicapping.
Well there’s no doubt this game uses some sort of system to rig the games.
It’s anything but random.
I don’t need to prove it. We’ve had 13 seasons, and plenty of people have seen it plain as day. They’re here trying to tell the devs how frustrating it is, but Blizz knows that SR is the only carrot they’ve got. So it won’t change.
Shame, because the core game is fun. But the comp system is absolute cancer.
For the first option, no one’s actually saying that “good” and “bad” players at whatever SR can’t be mixed onto the same team. What I, myself, am trying to say is that this isn’t intentionally, and systematically done. The matchmaker itself doesn’t actually care or even know if you’re a “good” or “bad” player at, say, 2300 SR. All it cares about is that you’re an available player at 2300 SR because you clicked the Competitive Play button, and it tries to find any and all approximately-2300 SR players, or groups averaging approximately-2300 SR (groups can do funny things to the matchmaker, especially if the highest and lowest SRs of the group are very different).
Also, I know we had this discussion on what SR and MMR is, and I maintain that SR is simply a rough translation of your MMR (where MMR is a series of digits/characters not meant to be read by humans). That is why I use SR and MMR interchangeably, because it’s stated by Blizzard that MMR is the sole value used in matchmaking, and not what Cuthbert presented as finding 12 players by SR and then distributing them into two teams by MMR.
Anyways, back to talking about the first option, instead of handicapping “good” and “bad”, I’d actually argue that the matchmaker YOLOs a bit too hard and that it’s actually random, and random as all-hell (within a reasonable SR range, obviously). I’m going to have to pull out some admittedly very weak evidence myself for this argument: look at the posts on just this forum alone complaining about how way too many games are extreme one-sided stomps (even back when KOTH was best of 5), or the dreaded all-support-mains on one team.
Now, going to the second option you presented: Cuthbert took a quote from a topic that was talking about SR gains and losses in games that didn’t appear to be 50/50 (and I really hope you read that topic, particularly the opening post and Mercer’s post, which is post #3). Context matters, that’s why I’m asking you to read the topic I linked and explain, based off that topic, how you can come to a conclusion that the following happens (excerpt from Cuthbert’s opening post):
The issue with that: MMR is a value, linked but can be different from SR, a separate value, ergo using them interchangeably is inconsistent at best.
The mere fact that SR is volatile, always moving, and MMR can be consistent, can remain stagnant, hinders them from being used interchangeably.
“MMR works very similarly to SR. There are some minor differences that make it feel worse though, when you just watch that number. For example, 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 – even if it’s just a little bit – to feel psychologically rewarding.” - https://us.battle.net/forums/en/overwatch/topic/20758686566#post-2
This in particular addresses that they look at a SR range:
“The first change is limiting the maximum Skill Rating (SR) difference between the highest SR player on a team and the lowest SR player on the same team.” - https://us.battle.net/forums/en/overwatch/topic/20759648155
But because we know that SR is not how the ultimate match is made, that is MMR this implies that SR is used to make the “lobby” then MMR is used to make a match 50% win chance.
I, personally, would like a system to have SR and MMR truly linked. I in so much that SR only goes up or down based on if MMR moved. Give us the “digestible number” but forgo the psychological “reward” of seeing the number continuously moving.
Back in way older seasons, you’d see the absolute top of the SR ladder (4600+) consistently gain very little SR on wins, single digit SR gains, even to the point of having a 3-game winstreak, where each win in that streak gave them 2 SR, for a total gain of 6 SR for 3 wins. That’s the only time where MMR would “not move” on a win. It’s the same at the very bottom of the ladder (back when you actually saw the SR go below 500), where your MMR wouldn’t “move down” on a loss. MMR “not moving” does not happen in the middle ranks.
As explained in that post, the main beneficiaries to that “change” were the extreme ends of the ladder. If you watched streams back then, you would see actual OW pros/streamers at high GM being matched with low masters or even diamonds during primetime playing hours, and obviously, the great majority of them complained about, for example, having the following kind of match:
4600 (the pro or streamer)-3750-3900-4050-4100-4000
vs
4250-4200-4300-4100-4150-4200
You did not see that kind of disparity in the middle ranks without actually voluntarily grouping, because again, the middle ranks are obviously the most populated ones.
Also, SR is not used at all, obviously except for the part where you group before the party leader clicks on the Competitive Play button. It’s really nitpicky to use those Blizzard quotes to say that there can be perceptible differences in SR and MMR when they have also stated that (with the exception of extreme inactivity decay) SR and MMR are closely linked. That’s why I maintain rough translation, with MMR allowing for decay and the performance-related SR bonus/penalty attached to the base win/loss.
I like this nuanced discussion. But I still don’t see a valid reason for MMR to exist in the first place, and to my mind the debate ends there. If there is not a valid reason for this backstage tinkering (MMR) with the user-facing ranking system (SR), then MMR should be abolished.
MMR works for high ranks. If there was no MMR, i would have to play against decayed top500, GM, and masters. Instead i play at my level because they get games against their MMR.
It pretty much does nothing for plat and below, because if you have a high MMR there, it means you are pulling the stats for it.
That makes you automatically better than any plat, and the game throws 30+ points per win (75 is my record) and it takes very very few SR (15 or less) on losses. Sure, you lose games (because plat is a sinkhole), but it doesn’t matter. You can climb with a 35% winrate if you are good. I once climbed out of gold with a 35% winrate (my widow is not very good =$ )
You think MMR puts you against other “stuck” people with high MMR, but high MMR people don’t get stuck.
If you put two top500 against each other in any rank and they fight each other every game they CLIMB, because they are winning 50-100 SR each victory and losing 5-10 each defeat.
True as, at those ends the population is so thin that making a match was difficult without making radically variance leeway.
But that does not change the fact that it is something that is accounted for, just less perceptible or needed in the middle/populated ranks.
I am not inferring, outside of decay, that the disparity is huge. Just that there is one, and that disparity, while close and linked, can be different. Which is how MMR can effect your SR gain if you are too far from it. Albeit a small factor in your SR gain or loss.
If there is a disparity, then you can logically deduce that someone at 2450 may have an MMR +/- 100. As win and loss streaks do occur. I had a 6 game win streak from one day to another.
There is no evidence of this. No posts indicate that MMR stagnation is at the extreme ends of the spectrum, only that it is something that could happen and that would make people psychologically unsatisfied.
Stagnation is probably only indicative of a win you were stated to win with a wide margin or a loss your were synthesized to lose by a great margin ~40-45% and your contribution was average to your performance mean.
I will start off assuming the White Knights are 100% correct. The system is working and everyone is where they’re supposed to be based on their skill. For example, last season I peaked in plat but fell down to gold because I’m a gold-skill player. Further, in gold I had 50/50 sessions where I lost a few SR. So, I must be where I’m supposed to be or maybe a little over-ranked, right?
This season, I’ve played 13 matches and won 10 with this last bit being an 8 game winning streak. This, of course, includes placements. So, if I really was where I was supposed to be based on my skill, then why the win streak?
Theories:
I suddenly got gud? Highly doubtful. I haven’t changed anything since a few weeks ago when I was losing. So, if I wasn’t gud then, I’m not now.
Getting carried by friends. Again, highly doubtful. I’m playing with the very same 2-3 people that I lost down to gold with, and they haven’t improved lately. If they couldn’t keep me in plat, then they must be at the rank they’re supposed to be. Also, in every session I was the highest or second highest SR player. The friend who was higher was only ~100 SR above me.
I’m getting carried by randoms. Possibly, but how long are the odds that I would have 10 out of 13 games (~77%) with under-ranked people who were just that good?
Let’s say #3 is the actual scenario. Why did the matchmaker set up these teams? Seems like something is off. Here’s the foil hat part. My forum account is linked to my game account. I’ve put forth so many legitimate problems with the comp system that Blizzard has altered my matchmaking status.
They’re purposely giving me good teammates to keep me quiet. Remember how I said people who benefit from a system rarely find fault with that system? I guess they didn’t believe me when I said I’d stop complaining, and this isn’t a complaint post. I’m simply reporting what’s happening and supplying possible scenarios to explain it (with a little snark thrown in).
I guess it’ll be like the house cleaning elves in the Dreseden Files books. Now that I’ve told people about them, the matchmaking elves won’t help me any more. Oh well, I’ve been expecting a losing streak anyway.
But when would you see such a thing in the middle ranks? Let’s assume prime-time playing hours, Pacific Standard Time. Show an example of a match with two opposing teams, before the “change”.
This is where it becomes hard to discuss, because you have MMR presented something like “This player has 2400 SR but his MMR is really 2500”, which I don’t agree with. I still maintain that SR is nothing more than a human-readable translation of MMR, which was meant to be a non-readable value according to Blizzard’s statement of “SR being a more digestible number” when MMR came into the discussion. Part of my reasoning comes from this video:
I linked to a Stylosa video, and I know this isn’t hard evidence since none of it was ever documented, but according to Stylosa’s explanations of what Blizzard has internally said (Jeff, specifically), your SR may be 3000 and your MMR may be “5526639280124925563…etc”
This goes back to the first two points I made above. Because the minimum gain/loss of SR after a non-draw game is 2, you could assume that would be the MMR stagnation case if that happens (MMR would “change” since a game was played, but SR is just the visible form of what happens to your MMR, which in the case of 2, be “not much”). This isn’t counting the long-standing bug of your SR gain being 2 if you win a game after leaving a game previously during the “safe period” if one of your teammates left first. That one’s probably caused by some kind of desynchronization between game server and database server where a player who “safely left” gets double losses counted or something like that.
Under normal circumstances and in the middle ranks, you would NEVER see a movement of 2 SR unless something extreme happened, such as you and 5 other solos at, say, 2400 SR, took on a six-stack of Diamonds and above at like 6:30 am.