Originally posted by Claushouse
As I understand it, the issue we were seeing was that at the beginning of a match, many vets were competing to wipe out the new players as fast as possible. Only huge scores end up counting, so it's an absolute necessity to get as many easy kills as possible before the chance dries up. Meanwhile, if you're an average player, running away from a top player is the best course of action if they're worth the same as anyone else. As you point out, this mirrors normal strategy -- take out the newbies, save vets for the end. But it was becoming more extreme. Risking death on the chance that you might not die and perhaps will fluke into a great score... because the point is, only a good score is meaningful in TSL. You need to win, of course, but also have a rare performance. Both of those are 100% essential. Just trying to win the game and do right by your team isn't enough to actually excel in TSL. And the factor you have more agency over is getting that fluke performance, particularly at the start of the match when the competition is lighter.
This wasn't such a problem in the beginning of TSL, because I think this strategy hadn't quite evolved yet. I'm not sure just going back unweighted kills would make everything better. Certainly could be tried, though. But I'm guessing we'd see similar issues once more. As I understand, this is one of the reasons TSL began feeling less and less satisfying. Felt more like gambling. Not sure how to get away from that. It's played as a team game but when it's all down to personal score, everyone's going to (have to) rec. Even if the win bonus is 50, and you receive 0 rating for losing, the optimum strategy is still to rush the newbies at the start, even if you risk careless deaths. Hope you don't take too many deaths scraping up the easy bites ... and then hope your team wins. If you don't get those easy kills at the start while taking a small # of deaths, it doesn't matter whether your team wins or not. The performance is null. So that means your win actually kinda feels like a loss. (Again, this is all couched within the non-committal "As I understand it...", which is to say, my understanding is that of an outsider, garnered mostly from impressions I've taken from various people in the community.)
It might be worth looking at simple match W/L % as the determining factor. Given roughly even teams, who at the end of the season (and having played at least X games) wins the largest percentage of their games? Just brainstorming. Because as long as getting as many lucky personal performances as possible is the key to winning TSL, it's probably just not going to feel much like winning. And you're going to have a lot of frustrating games along the way to that "win." If we can reframe it more toward winning matches somehow, as just about every other team-based league runs, we might be able to get away from that underwhelming feeling.
On a separate note, I believe the WB and Jav matchmaking ratings have been automatic for the past two seasons? I sent Qan the formulas to remove human input after Season 1. I know they weren't implemented in Season 2 due to time constraints, but I think they were slated for Season 3. I didn't play S3 or S4 so I'm not sure if it got added.
A group of TSLOps still adjusted scores. One would put in a rating, and another Op would need to approve. One of the biggest problems in early TSL was the dark horse phenomenon, where players unknown to the TW competitive scene, often pros from other zones, would play games at a rating of 40. It was honestly way too much work for one person. So Ops put in a ton of work giving many people initial ratings, and after that point the bot could take over.
Comment