Engine championships


Ineptly Chaotic
Apr 5, 2020
A Hermit's Cave
So, the Top Chess Engine Competition Season 17 ended last month and Season 18 is underway. Leela c0 won season 17, beating Stockfish +17 -12 =71 which continues Leela & Stockfish's ongoing tussle and that of traditional engines vs neural networks.

Thing is, its popularity is incredibly niche, even among chess enthusiasts mostly because there's literally no way human players can compete, for obvious reasons (Magnus Carlsen at his best wouldn't even qualify for the preliminaries despite being rated a little under 2900 at the time given the 'traditional' criterion is being rated 3300+), and engine vs engine gameplay can be horridly dry, especially endgames (though AlphaZero does have a strange reputation for spicing things up), and are typically pointless to analyse with any real depth because of the concept of perceived 'human moves'.

My question is: for MOBA games (notably DotA), would there be any potential market or following for AI developers to compete in this way? (Or is there one now? Seeing as I'm not a MOBA fan.) As far as I know, OpenAI5 has come far enough to regularly beat human champion players. Could they stand to have competition in this regard at all?