PhoenixFire55, on 08 July 2015 - 12:24 AM, said:
Piss poor pathetic excuse is still piss poor. We'd have vastly better matchmaking if people responsible for programming it actually had a clue what they were doing. A MM that for all intents and purposes treats a Dragon-1C exactly the same way it does a TimberWolf is a fkn fail. Not only this is a poor excuse, but it is also full of lies and BS, beacuse your mr. Russ has said not very long ago that game poplulation is at its highest, so either he was lieing, or you do when you say that MM queues don't have enough people for quality MM'ing.
*shrugs* We've got a small player pool. The game population may well be at it's highest, but that doesn't mean there's a massive pool to draw from. You can wrap your mind around that, right? That you can have higher population than before, but still not a massive number? I've shown the math in a lot of threads before, but assuming you have 20,000 people
concurrently playing(that is, constantly in matches or requeuing, not in the mechlab or afk or whatever), with NO map select, there's still only enough to get reasonable matches for average Elo people, and medium-high and above (and the related below) will have sketchy matches at best. Add map select, force the MM to try to match not just Heavy vs. Heavy but more precisely? If you go battlevalue, you've got a massively complex balancing task ahead of you that would be awesome if done correctly but is totally impractical to do, and would STILL ultimately make it far slower to build matches. Right now, the MM just tries for 3/3/3/3; to try to match mechs by some more precise rating system would limit options exponetially.
Given that, stop ranting for a minute and THINK about it. If there isn't enough people to quickly get good matches for everyone with the current set of parameters
at prime time (though there is enough to get decent matches for the bulk of average players then); adding more parameters makes it exponentially more difficult for the MM. And by "more difficult" I don't mean harder to code (because ultimately it's a simple algorithm), I mean it's going to take a LOT longer to find the appropriate people.
Perhaps you feel there should be much longer queue times for everyone, and that's a discussion that's fair enough to have. Longer queues = better matches, but there's a point where your average Joe player just wants to get into a game rather than stare at "searching for match" endlessly.
It would make match finding times at off hours and for very high/low ranked players many times worse than it currently is.
Quote
Plus, I am still waiting for you to tell me exactly how "wrong" I am and exactly what was the problem with the original RiverCity map.
Until then, ggclose.
How about ridiculously tiny size, hideous color filter, and poor overall design? It wasn't a good map to play on at all. RCN in particular was one of the most hated maps overall. There's a reason River City was viewed as the #1 map to be redone. Now, this is a subjective thing, certainly, but going by response to the new River City, pretty damn close to 100% of people like it a hell of a lot more than the old POS map.
TLDR version:
It doesn't matter how well the MM is coded, if there aren't enough players to make great matches quickly, the MM cannot make great matches quickly. Period.