I went through all of the posts in this thread for December, analysing data from all screenshots to identify trends.
Here are the main issues identified across the screenshots. While there were many other issues raised unrelated to balance, like cheaters and toxic behaviour, when people said the matchmaking was bad, this is what they pointed to as evidence
Overwhelmingly the main issue was one player gaining far more kills than everyone else, and thereby becoming untouchable. There were quite of a lot of instances where this was contained to just one lane, with a single feeder, but more often the problem spread across lanes and resulted in multiple players feeding that one player.
Less commonly others would join in on the stomping to create multiple stompers, but never was it shown that an entire team was stomping, or even that the majority of one team was stomping. There were also 28 matches reported as feeling unbalanced, but the scoreboard showed quite a close game. In some of these the outcome appeared to be determined by feeders, though without any stompers.
To classify as a stomper in this analysis, a player had to have at least 10 more kills than the average on the enemy team.
To be a feeder, a player had to have a negative K/D and at least ten deaths, or 0 kills and 8 or more deaths
Here are the characters that were the most often fed. So when people were upset about a player killing their whole team, this is the character they were getting stomped by
and here are the characters that were most often feeding them. So when people complained their team mates were dying too much, this is who their team mates were playing
Of course this doesn't account for how often characters are picked. Haze has a nearly 100% pick rate so is near the top of both lists. We can account for this somewhat by comparing the times they were a problem for being too fed as a percentage of total times they were a problem in these games
So this might be interpreted as saying when vindicta, yamato, shiv and haze are the characters most associated in this thread with unfairly higher skill players. At the other end, Seven, Viscous, Mcginnis and infernus are the most often associated with unfarily low skill team mates.
Finally, there is one other very clear trend in the screenshots. People who come to this thread to post screenshots of their matches, overwhelmingly have their rank turned on, by a nearly 5:1 ratio
I don't know how prevalent this option is in the general player community, but the 'default bias' in UI design would suggest any non-default option should be rarely selected.
Actually the correlation would be even stronger if you included people who mentioned rank in their posts not just their screenshots. Concerns about ranks being impacted by these perceived match imbalance issues are extremely common in these comments.
This could mean that people who turn on their rank are more likely to care about the outcome of their matches or post in this thread.
Controversially It could also mean that regularly seeing your rank makes you less happy with your matches
