I prefer to spectate long games, where I think the players have a chance to take the game more seriously. It’s hard to find an “objectively good” way to filter, because byo yomi and fischer time could have very short main time and very long periods, or people could play with absolute time, so the most flexible option would be to provide some kind of open menu for filtering so the user enters the data themselves.
As a starting point, though, we might be able to get to a “good enough first pass” with the following criteria for filtering:
“short” game has under 15 minutes of main time
“medium” game has more than 15 and fewer than 30 minutes of main time
Do you mean Malkovitch filter? Yes … I did end up doing that, and it did need back-end support.
I also didn’t get as far as I wanted with it … you can filter “yes/no” whether they have logs, but you can’t currently sort them by the ones that have the most logs - once again, a bit more tricky to do than it seems.
Not impossible - and this filter is not impossible either - I was just giving heads up to anyone diving in about the nature of the work involved.
Yep, games do have a time_per_move field, which I think means that this would be easier than Malkovich? But afaik all the filtering for this page happens on the backend. While a frontend-only filter is possible, it might be unnecessarily complicated.
If I’m interpreting the original request correctly, the seek graph won’t help since it displays only challenges, not active games.
The seek graph code actually looks like it used to support showing active games as well. That might be a cool feature to bring back (on a different page), but only if there’s enough interest.
At first I thought: “that could be a nice cloud!”
But then I realised that those games shoud mostly share same settings, so instead of a cloud they should look like a bunch of lines
You’ve been pushing for this feature since 2022. Requests are done in priority and yours doesn’t seem to have any demand from community except one person. You need to convince developer that your feature is useful and, importantly, urgent to implement or to find large enough number of like minded people to voice your request so that it can’t be ignored.
It doesn’t seems like you achieved any of those if you need to bump this topic singlehandedly not to get it archived.
If seek graph already has the info, then it may be a front-end easy job to put the filter in place.
I should not be allowed to post before morning coffee 
