Anomalocaris, on 18 December 2020 - 04:07 PM, said:
First of all, 3% is a lot. If the govt raised income taxes from 30% to 33% I guarantee people would be screaming about it and trying to figure out how to deal with that much less money every month. Going from 30% to 33% stomps means that person is experiencing 10% more stomps than they used to in a given period of time. For the average player that's about 3-4 more stomps per month. A large enough change that most people would perceive it. Never mind that PGI's definition is, by necessity, rather basic.
3% tax is a lot because of margins. 3% in the situation of wins/losses is not. 3-4 extra stomps a month sounds imperceptible. Do most people even remember what matches they won or lost for more than a couple hours? Personally, I remember really clutch wins, but nothing else specific. Most people just have a vague impression of if they're winning more or losing more, and vague impressions are unreliable.
Anomalocaris, on 18 December 2020 - 04:07 PM, said:
Second, while confirmation bias is definitely a thing, there are strong groups out there that are winning 80% of their matches in quick play and stomping about half the time. I expect there are really bad groups that have the stats reversed, but they don't tend to stream or post their results. The problem for the average solo dropper is when they start to observe these groups showing up repeatedly in their matches, and they come to the (logical) conclusion that whichever side that group is on will be winning the vast majority of matches. That makes those solo players feel like they really don't matter. That they can't affect the outcome, and that they are little more than NPCs in those matches. Again, probably the same thing if there is a crap group dropping as well, only reversed. No matter how hard you carry, if you get those guys on your side you're kinda screwed.
I'm willing to accept that there are groups that have 80% stomp rates, and that they don't show up in the data for whatever reasons. That seems totally plausible to me. But that brings up a new problem: Are these players stomping because they are a group, or because they are good? If the same people solo dropped instead of grouping, matches with an unbalanced amount of skilled players would probably still have the same stomp ratio.
Consider we have 2 groups of 4 skilled players. They usually match against each other, which balances out. When they aren't there to balance each other out, either group will stomp pugs. Now, remove group drops. The same number of skilled players are dropping, just mixed with each other. They lose whatever advantage teamwork within their group provides, but are also free of tonnage limits. You'll still generally have the situation that whatever team gets more of these experienced players is much more likely to stomp. And now you are actually more likely to have more than 4 of them on the same team, which could easily balance out their lost teamwork advantage in the overall stats. (This effect could also explain why adding groups in only changed stomps from 30% to33%.
Stomps could stay the same or even increase, but people wouldn't blame pre-mades anymore. Would there be the same level of frustration? What if we kept the premade groups, but people weren't part of the same unit, so it was harder to notice them?
The big assumption people make on this topic is if you didn't allow groups in with the solos, that the high skilled players who play in groups would cease to play quickplay at all. Some would, but other high skilled players would start again.
Without an ability to separate the performance of a group of skilled players from the performance of the same players dropping solo, its an invalid comparison. Is the source of stomps the skill imbalance? Probably. But how much of the power of a group is the skill of its members, and how much is the advantage of them being grouped? This must be addressed if you want a clear picture of what's going on.
If we had a matchmaker that was better at balancing skilled players between the teams, would groups matter?
Edited by Heavy Money, 18 December 2020 - 04:25 PM.