Snowbally Games and Game Length

Riot has stated that the games aren't actually THAT much faster, and that they've only seen faster pace "...reflected in about a minute off the average game-time (patch 5.23 notes). Some of the changes made, especially those to turrets, were made "...to reward being decisive." However, just as winrate is only a partially accurate barometer of champion balance, I think game length is only a partially accurate barometer of game snowballing. In my experience, many more of the games I've played since the preseason changes have been decided much, much earlier than they used to be, but continue to limp on because people won't surrender. To come clean, I'm one of those die-hard "never surrender" people that have to really, really be having a shitty time to vote "yes," both because I believe in the comeback and because playing the game sure beats sitting in queue, even if you're losing.

The goal of this change was to make it so that if one team had an advantage and began to capitalize on it, they would be able to secure victory efficiently without leaving much room for a comeback. If that's the direction the game should go, then so be it, and while there's decent arguments to be made either way, I'd lean in favor of it. However, it feels to me like what these changes actually accomplished were to make it so that many games were nearly irrevocably decided in lane phase before a lot of real team play had occurred.

TLDR imo preseason felt snowbally because Riot successfully shifted the most impactful time during a game, but shifted it to too early in the game. The statistical decrease in average game-time does not accurately reflect this, just as winrate doesn't always accurately reflect champion balance.

Discuss.