Updated wording to stop referring to Hitler as just a "world leader [some people] don't like," after finding out he was kind of a bad guy what?
QM now has the same templates as the MTL, ordered by popularity
Fair on SE. But the voting system and the fixed templates both don't reflect what casual players actually want to play on, especially when you consider that the vast majority of them are not going to even unlock the templates near the bottom that they voted to put in there.
Though that would significantly rise the amount of text which is storaged.
They already store logs for all of Global Chat and possibly Clan Chat, it's just about what's visible on the UI.
Nah, the site might get too commercialized.
Yeah, they might create a cash cow side-game, unnecessarily tightly couple it with the main game, and keep nudging everyone to play it with unwanted but unavoidable Idle notifs. Or ignore user feedback, etc. How exactly is Fizzer operating the site differently than Activision would? You can't make Classic p2w: there's a dozen other games in this genre, none of them are p2w because the p2w model does not work for any of them (for the same reason we don't have p2w chess). Instead the standard is pay-to-unlock or pay-to-progress, which is already the model of this game and especially Idle.
that skill from one template is not that much transferable to another.
There's a significant correlation, though (see: MTL - I actually checked these correlations in https://bit.ly/mdl-analysis
). Plus it doesn't matter since most players stick to a favored template.
Oh... and how is this going to be worse than having a single TrueSkill rating for every player in a clan? Is my clanmate's skill on Strat MME transferrable to mine on SEAD more than my own skill would be across those two templates? I should get paired with opponents who are around my own skill level, simple as.
CW is a casual event
Yet MME MA and Strat 2v2 are wildly unpopular templates while the popular templates on CW are SEAD, SE1W, CUMBOMB, Guiroma, and Strat MME. If it's about what casual players want, then no need to force it to have templates only Fizzer and an eccentric minority (no offense intended) want, especially when it appears to decrease overall participation.
write a better one
We had a whole thread of people doing this after the yolo swag clan got a bunch of voided games over and over again, with Derfellios writing some very straightforward pseudocode that he could prove would have desirable behavior (maximum match quality for minimum voided games). Also, I already have: https://pypi.org/project/pair/
Matchmaking has been a solved problem for longer than Fizzer has been alive. I even offered in that thread to implement a good matchmaking algorithm for him as a library in whatever language he wants, if it's a time thing. He just didn't read the thread lol and responded with some asinine reasoning about match quality that doesn't hold up to the slightest scrutiny (his greedy heuristic guarantees only maximal match quality for the lowest account ID on the highest-CW-rated clan, at the expense of potentially reduced match quality [including some VERY low quality games] or voided games for everyone else, with increasing likelihood for low-CW-rated clans; bottom-rated clans are going to frequently get voided games, and once they leave due to the frustration you get a new bottom-rated clan...).
This is not a hard problem, to reiterate. The algorithms to matchmake optimally have existed since back when Kennedy and Khrushchev were running countries- and they can support arbitrarily strict match quality thresholds and provide stronger guarantees of match quality than a greedy heuristic ever could. It probably took more effort to write a novel bad matchmaking heuristic than it would have to find and integrate a library for one of the optimal ones. Same as with rating systems for QM, CW, 1v1 ladder, etc.: instead of simply lifting things that work, Fizzer arbitrarily decided he knew better and made much worse things that required more effort than just doing the obvious thing. Which is fine, except we've been stuck with these bad systems with likely no chance of them getting changed in the near future.
Edited 2/9/2022 18:22:50