Riot doesn't understand that Dynamic Q will be frustrating even if it's balanced
In case you haven't heard, Riot has posted an update about dynamic queue, stating that solo queue will not be coming back.
Reading through the post, I noticed that Riot seems to be approaching the problem from a purely statistical perspective. In other words, "if people playing in a premade of X size have the exact same chance to win the game as solo queue players, dynamic queue is working." While this seems reasonable at first glance, that kind of balancing won't alleviate player frustration.
Everyone has had a game where it seems like the enemy team is perfectly coordinated. Six minutes in and there's five people bot lane diving your turret and walking out with a double kill. Suddenly, your teammates are typing "open mid" in all chat and end the game with "screw dynamic queue."
In Riot's ideal world, this scenario is just as frustrating as it is now. You can show players statistics all day that both teams had the same chance to win. You can show players that everyone on the enemy team was actually in a lower division and had lower MMR to compensate for the advantage of dynamic queue.
The thing is that gamers just do not care about statistics. Losing lane to someone better than you sucks, but losing lane because the enemy team had better coordination through dynamic queue sucks even more. If your team wins but you spent the entire game being camped, you're probably not cheering over the victory. Even if you stack the odds in favor of solo queue players, they will still be frustrated.
This all leads to an even greater issue, the threat of dynamic queue. In the past, if you got killed by a 5 man dive in ranked, the enemy team was simply coordinated. Nowadays, you usually assume that they're just a premade. I'm willing to bet that most of the times people accuse the enemy of being a premade, they probably aren't. How often have you checked the scoreboard to see how many enemies have the same club tag, only to find none of them share a club?
It doesn't matter if 20%, 2%, 0.2% or 0.000000002% of games have a 5 man premade in them. Just the idea that the enemy team might have killed you because they're premade is frustrating enough on its own. That amazing Malph ult > Ori ult > Yasuo ult penta might have just been the result of a few good pings between solo queue players, but good luck convincing the other team of that while dynamic queue exists.
If Riot wants to continue with dyanmic queue, they will need some way to tackle all of these issues. The goal of a vast majority of people playing League of Legends, or any game for that matter, is to have fun. It's tempting to think these sorts of problems can be solved with some tweaks to the matchmaking algorithm, balance is only a means to an end, not an end in itself. Doing statistical analysis can be difficult, but the true challenge is using those statistics in a meaningful way.
Tl;dr: Riot seems to think that frustration with dynamic queue will disappear when premades have the same win rate as solo queue players. If you've ever had 5 people dive you under tower at 7 minutes, you'll understand why that's not the case. Just the idea that you might be up against a 5 man premade makes any coordinated action they do extremely frustrating, even if they're all solo queue players.