So, long story short, what we’re looking to do with our website (www.Pyclid.com) is to let people vote on important elements of any game (obviously characters in USF4 and other fighters, guns in CS:GO, classes in TF2, heroes in MOBAs, etc) and then create dynamic tier lists based on those votes. We’ve got a bunch of games implemented thus far, but USF4 is particularly important to us, as it’s the most traditional and proven game as far as tier list creation goes, and it’s where I’ve spent the majority of my personal time for the extent of my gaming life. Ultimately what we’re trying to do is democratize the tier list creation system so that the views of the entire community are taken into account, and will change dynamically as updates hit or new setups/tech are discovered. The algorithm we’ve created does a solid job of normalizing the list against concensus and minimizing the effect of statistical outliers (trolls). In the future, we’ll also be adding a tier list / poll builder function so that people can create their own traditional individual tier lists, and also mini type-specific lists (grappler, shoto, etc) if people are interested in that kinda thing. As far as the main difference between this site and something like Eventhubs, is the algorithms used and Steam API integration. Obviously, for anyone playing on console rather than PC, the differences won’t be as profound, especially before we’ve integrated infographics and more rich, user-driven functionality.
So what we’re looking for now are some people to help test our model, and flesh out the current tier list. It’s an early prototype, and we’re hardly infallible, so we’re always looking for feedback. The results page is pretty bare bones at the moment, bear with us as we figure out how to make SQL queries pretty :lol:
Especially when it comes to something like matchup charts where people will be not equally selective on what they do and don’t vote for, regardless of their skill level.
I think my favorite matchup charts are the ST ones where they asked multiple top players for their opinion on each character.
Without going into too much detail about how the algorithm works, I’ll just say I agree with what you guys said, specifically as it pertains to matchups though. That’s one reason why we are focusing on voting pairs, but specifically as it pertains to higher or lower tier, not matchup. Individual matchup polls, with much smaller vote totals and slower vote decay wouldn’t be as easy to manage.
Match up based tier lists are generally a lot more meaningful than arbitrarily tiering characters. Good, thorough tier lists are based off of match up charts.
And again, democratizing tier lists is an awful idea. See Eventhubs.
I was merely explaining our approach to the process. If we need to rethink our approach as it pertains to fighters to make it specifically matchup focused, that’s fine. We want what the concensus among experienced players is.
Thanks a bunch for this! Will definitely come in handy.
One part of the problem with Eventhubs tier lists is that they open em up right when the game drops or latest major update drops. They give zero breathing room for the thing. Sure they get regular new people voting but I can bet it’s less than 10% of the userbase who ever goes back and adjusts their scores EVER. Regardless if they change their mind or a new minor patch changes something about a MU that is important. This causes players to never really adjust their views and while some of the characters properly end up floating up or down a bit they never are accurately reflected in their strength/weakness because of all of the people who voted one way on the first day of the game.
Then you have the issue of opening up it to ANYONE in the public. I’m not an exclusionary person by any means. I find it annoying whenever any high level player looks down on someone just because they haven’t proven themselves in a tournament setting. Seen some actually high level people refuse to believe something ultradavid says until he sits them down and actually does it in person to them but if a top actual PLAYER were to spit it out something most players eat it up as fact unless it’s COMPLETELY absurd like Daigo saying Akuma is a weak character which was obviously a come-on.
When you got democratic tiers everyone’s vote is equal, from the guy who programmed the game to the guy who couldn’t tell you how to perform a DP, from the guy who won Evo to the guy who can’t get past the first character in arcade mode. Leads to less of an accurate representation of characters potential and more a representation of two things:
Parroting (how much the characters are talked about by other players as being strong / weak with people just accepting those ideas as gospel.)
Placement based on personal online experiences rather than effectiveness of the character. Gimmicky shit works a lot better online and isn’t really a good representation of the potential of a character. Vortex is much stronger online while footsies tends to be weaker for instance. Characters who require reactive play in neutral (Guile, Dhalsim, Chun, Hakan, Cody, Balrog, ect) are often considered even worse than they very well may be when compared to characters who play a proactive game and require the opponent to react to their actions and properly punish unsafe things (Cammy, El Fuerte, Makoto, Viper, Yun, Abel, Akuma, ect.)
Personally, whenever a person submits a matchup chart I feel that they should at least do a BASIC explanation of some of the key points they have on MUs that don’t fall into the generally agreed upon views for that MU. Like say you believe Yun beats Zangief, at least give like a 2-3 sentence explanation as to what scenarios you feel are in Yun’s advantage and how Yun’s tools at beating Zangiefs. It doesn’t need to be one of my well known “9 paragraph” rants just give us like 3 sentences saying “This tool generally beats this other tool. Gief’s gameplan is XYZ and Yun can counter Gief’s options with ABC in these situations. The risk vs reward falls in Yun’s favor in these situations.”
At least give a BASIC explanation like that. Sadly you can’t get this from most people
