The ESAM topic is a dead horse. Yes, his position is silly and it's silly that it's taken over parts of the zeitergeist. Yes, he puts his money where his mouth is and puts in incredible work; a great player by any metric. Yes, having a player not whine about how everyone overrates their top tier and instead say "Screw it--my guy is #1" is in many ways refreshing and even inspiring. (Regardless of how off-base it is or isn't)
Multi-axis tier lists make sense when you, well, have multiple axes. What would the axes in smash be? Please don't answer that, it's a rhetorical question with no answer--
blackghost
is right.
Of course a legitimate matchup chart would be the ideal, but we are so, so far away from that. It was feasible to make a full chart for Melee's 325 matchups after only a few years. If you were an active, competitive Melee player, you probably had first-hand experience playing
at least some meaningful amount against a proficient player for every character in the game. And even if Bowser or Ness slipped through the cracks for some people, a few debates on -3 vs -4 matchups isn't really that important at the end of the day.
Brawl's 666 matchups barely got a believable chart after 6 years. It was less than twice as many matchups, but in practical human terms on the order of 10x as much challenge. I played over 30,000 matches of Brawl, and while I think I did manage to check off every single character in my comeptitive experience, for almost half the cast it was only against 1 strong player each. My first-hand experience for plenty of mid-tier characters like
was driven by a single top regional competitor, with all of their biases and quirks. Hell, I think I only played a "good"
1
set.
was one of the best characters in the game, and I only played 3, maybe 4 ICs ever.
We never came close to a comprehensive matchup chart for Smash 4's 1596 matchups (if you ignore Dark Pit). That's like 10x harder
again, only now also the game is way more balanced so the matchups are up for way more debate across the entire grid. The number of people who played at least
one top 1000 player for every character is vanishingly small, maybe single digit. The number who played at least 3 of each might be
zero! At best, after a few years we were able to assemble an okayish-matchup chart for just the top ~12 characters--just those 66 matchups, barely 4% of the game.
Ultimate will end at 3321 matchups (if you ignore 4 echoes), doubling things again. And improving balance significantly, again. Personal first-hand experience playing just 10 games of each matchup would take roughly 3000 straight hours. It's just way, way too much information to fit into human memory.
I believe this is the root problem behind "the decline of theory."
We have always had three pillars of information: Usage, Results, and Theory. Each is a limited perspective into the true reality that exists between them, the space of all games that have been played, could be played, and will be played. (Which is so incomprehensibly vast that we cannot deal with it directly, only look at it through these simplified portals.)
Usage, Results, and Theory were all understood to be flawed in their own individual ways. And over time, as we got more data, more experience, and more understanding, all 3 would converge towards the reality between them. Good results and promising theory would make people switch to a character, while bad results and pessimistic theory would encourage people to drop them. Strong usage and strong theory would enable good results, and visa-versa. The mistakes in theory would be corrected over time, looking at long-term trends in usage and results to realize things we had gotten wrong, overlooked, or misjudged.
But in the Melee and Brawl days, there was a general consensus that Theory was king, or at least some sort of leader. Usage and Theory would shift more towards the tier lists (and more quickly) than the tier lists would shift to be more like them. (I don't have numbers to quantify this; all I can say is that I lived it and feel very confident that this was the true status quo.)
But in Smash 4, the complexity of the game and its 1.6k matchups seemed to handicap Theory. It felt like tier lists and theory crafting in general was starting to lag behind, struggling to merely explain trends rather than predict them. With a more balanced and bigger cast, performance was increasingly driven by individual names, continuously confounding attempts to explain the metagame holistically.
And now we have Ultimate. In
this post, when I looked back at OrionRank 1 and 2 vs Reddit community tier lists corresponding to those periods, I found that face-value results were actually now predicting changes in the tier lists more than the other way around! In fact, the tier lists were no better than a coin flip at predicting which chatacters would go up vs. down--and that's despite cases like
whose trends were obvious, basically a freebie. (Their usage/results were plummeting 6 months in.) And that's
face-value results, not even looking at the most basic
trends or any other actual analysis.
So, that's basically my conclusion, my long-winded explaination for why I'm so down on macro-theorycraft in Smash Ultimate. I believe it is simply far too complex a problem space for humans to interact with via raw experience and heuristics, and believe that the observable low-quality of content created to this end is proof.
Theory will always be one of our "three pillars"--Usage and Results will never accurately tell us how good Marth is as long as Lucina exists--but it's no longer the "leader" driving most of the future and telling us the most about reality.