Posted 12 April 2016 - 10:40 AM
REASONS WHY MATCHMAKING WILL NEVER BE THE WAY YOU WANT IT TO BE IN MWO.
1.) ”Skill-Based” Matchmaking is Impossible
I see an awful lot of “well if they had a SKILL-BASED matchmaker instead of all this crappy ELO or PSR junk, then things would be better!”
Okay Mr. Armchair Dev, riddle me this: how do you programmatically determine a player’s ‘skill’? Remember! The matchmaking system is a bunch of code run on servers – you can’t put a line of code that says “match good players with good players and bad ones with other bad ones” in your system and expect it to work. You’re an armchair programmer now – what parameters do you give the system?
‘Skill’, in MWO, is most often/best determined with intangibles such as good twisting, good positional awareness and maneuvering, proper teamplay, and other things you can’t really program a machine or a digital system to recognize. There’s a very distinct difference between ‘Good’ damage that leads to quick enemy kills and victorious matches and ‘Bad’ damage that just inflates your match score – but to a machine, four hundred points of ‘Good’ damage that efficiently eliminated three enemy units looks no different than four hundred points of ‘Bad’ damage that accomplished not much of jack monkey squat.
You can try and program behavior triggers into the system, like Piranha did for the match reward table, but those are iffy and usually very easy for living, thinking, adapting human players to game. Once they understand the system they’ll know how to milk it, and once they start milking it they’re displaying the ‘Skill’ of being able to milk the system, not the ‘Skill’ of winning games in MWO.
2.) ”I HATE STOMPS! Every game should end in 12-8 or better; why can’t Piranha get their stupid matchmaker to stop giving us stomps?!?!?!!!1!!11oneoneexclamationpoint!1!”
One, stop emphasizing. One, or possibly two, exclamation points is sufficient.
Two: ‘stomps’ don’t happen as often as you think they do. A 12-3 fight could be a 12-3 fight with the 12 side being a bunch of ‘mechs with less than forty percent integrity left, out of ammo and on their last legs, to the point where one fresh Timber Wolf could knock over all nine remaining enemies. They just managed to out-attrit you for that fight, and most of them are probably sweating buckets and for once in the history of mankind actually mean it when they ‘GG’ at the end of the match – because to them it was a hell of a thing.
Three: even when you have a legit stomp, there are a million reasons beyond matchmaking why it could’ve occurred. Mechwarrior Online is a game where one mistake made by one player can swing the balance of the entire match. The fundamental structure and nature of this game encourages snowballing – your Whale driver takes a nasty CT shot early in the match and has to play gingerly from there out, his ginger play lessens the firepower your side can throw out, and that in turn makes it easier for the enemy side to get more of those nasty CT shots in. In a sense, that one, single whack to the Whale’s CT was the shot that won the game, because it affected everything else that came afterwards.
Alternatively, maybe some of your guys are drunk-streaming, or testing new builds on unmastered ‘Mechs, or their girlfriend is trying the game on their account, or they have a cat on their head, or are for some other reason not performing to the level the matchmaker expected them to. The matchmaker knows the average performance of any given player; it is not privy to that player’s specific, at-that-instant condition and thus the likelihood of that player living up to the matchmaker’s programmatic expectations of them.
3.) ”Piranha is the DEVIL! They could fix this game if they wanted to! What the hell are they spending two hundred thousand bucks on a tournament for when the game is broke and they’re always complaining about being understaffed?! PIRATES! BRIGANDS! SABOTEURS! SCALLYWAGS!”
Calm down. Breathe. Find the little switch under your ear and flip it the other way. All right. Your brain’s on now? Excellent.
For one, a lot of that tournament money is probably a tax write-off somewhere as advertising funds. They’re not shelling out as much as you think they are. For two, two hundred thousand dollars doesn’t go very far, insofar as business staffing is concerned. Game development is a highly skilled position, with a lot of competition for the individuals available. Even assuming Piranha could find someone with the skills they needed – which is not a given, as their eternal quest for more texture artists proves – hiring that person, complete with salary and benefits and other expenses, might well run them in the region of sixty or seventy grand a year, if not north of that. Judging by my own wage, which is itself a skill position of sorts if in a very different industry.
A core systems engineer guy, one of the guts-of-the-game programmers that can work in the game engine itself and do heavy-duty updates? Those guys are rare, expensive, and usually get their pick of jobs. Two hundred thousand is not enough to buy one of those guys for more than a year or two, and nobody’s going to take a job they’re going to lose in a year or two. Hiring people is a constant, ongoing expense that will, over time, pretty heavily exceed 200k. A tournament is cheap next to a programmer.
4.) ”I suck and I don’t want to admit it! The matchmaker keeps rubbing my nose in the fact that I suck, so I hate it and want Piranha to spend time, effort, and development resources Fixing Matchmaking™ for the seventh time so I never have to feel like I suck!”
You don’t always get what you want. I want an Information Warfare system that actually qualifies as Information Warfare. They tried a couple basic ideas on it once, then shelved it forever. You’ve had your way already – half a dozen times. There is nothing more Piranha can do with their matchmaking system, not without a much larger playerbase than they’ve got. Which, by the way, is generally one of the goals of a large, publicized tournament – attract new players.
5.) Shut up, 1453! You’re Tier 4, you suck so bad I’m surprised you can find your keyboard! I’m better than you at MWO so you should let me slander Piranha and incite needless forum nerdrage because off-hours MWO matchmaking isn’t perfect!”
I’m ‘T4’ because my aim is Not Good, I don’t play very often anymore due to work and also other games that warrant a slice of my interest as well, and because I have no clucks to give where it comes to Meta Dominance and/or advancing in tier. I play what I want when I want to play it, and that includes Warhawks with mixed armaments, Shadow Cats, experimental Mad Doges, PPC-armed medium ‘Mechs, and very occasional Krabs with six AC/2s and fourteen tons of ammunition.
None of that indicates that my brain is broken. A realistic appreciation for the trials faced by the matchmaking system, and a working, functional layman’s understanding of the overall nature of game development (or indeed project development of most any sort), does not require madd skillz. Especially given the opinion that many people in this very thread hold concerning PSR – namely, that it’s an XP bar more than a skill-measuring mechanic, and that one’s PSR rating has very little to do with one’s actual skill.
Can’t have it both ways, eh man?
Anything else? Or did I cover it?
4
LordSkyKnight, on 12 April 2016 - 08:17 AM, said:
