I think one of the best parts of speedrunning is being able to admire your own accomplishments, and look at the accomplishments of others. But looking at someone's profile and seeing "first place" doesn't mean much without the context of the game. For example, getting 10th place Portal Any% is a much bigger feat than getting 2nd place in a random mobile game.

So here's my suggestion. You should be able to change the metric for runs on your own profile (only on profiles, not on game leaderboards) into something else of your choosing. Whether that be percentile, or displaying it in front of all runs in that category, for example 2nd out of 78 runs.

I don't think the site should become more competitive, which is why I think this should only be on personal profiles. But I don't see the harm in more customization, and more ways to show off your difficult accomplishments.

Another metric could be difficulty based on perceived best possible time (which could be set by mods):

1st (5m 2s 067ms from best possible time)

3rd (1s 067ms from best possible time)

10th (067ms from best possible time)

Something like this could all be displayed in a formatted tooltip, but that might defeat your purpose @Breadn_

Really, the total amount of players within a game can be redundant.

The effect more players sometimes creates is a tendency for the best times to be ground down further due to the record being passed back and forth a lot over time. However, it isn't as if a game with a small amount of players can't have a time that was ground down, as well. This system would pretty much imply any game with a lesser playerbase has a somehow worse set of top times than a game with a larger player count, when that simply isn't always the case.

The reality is, even if you look at a game that is immensively popular, like Super Mario 64, the majority of the people within that game aren't having an effect on whether or not the top times are any better or any worse, because only the top like 20 or so are even really pushing those times down to begin with.

I just kind of hate this idea that because your game has far less players it automatically creates this idea your times are worse, when the whole cause behind your time being very good is the desire for it to be the best it can be, and not simply because your record was beaten and you want it back. That can be a leading factor, but it's not the only leading factor. It's not really the sites place to be implying stuff like this, and if it's not claimed to be the purpose... what is the purpose, then? Just comes off as tacky IMO.

Yeah there are just too many variables to say definitively that Game A has a more optimized WR or Top 3 placement than Game B, regardless of players.

To start with @Oxknifer argument, there are plenty of smaller playerbase games in which the full game world record is within a second, sometimes even a single frame away from theoretical limits. Its hard to say this isnt impressive compared to a popular game (lets say Halo 2 Legendary) where the theoretical limit is still about 4 mins away. These are equally as impressive world records, and ultimately comes down to what game you as a spectator would rather watch.

Another factor of course is leaderboard inaccuracy, especially with in-game leaderboards. This is most prevalent with racing games, but for example, one of my wrs on the site is in a category with no other submissions (so by default I have wr) but on the in-game leaderboards I'm 12th out of 800,000, so i would still say records like that are still impressive.

So yeah its very hard to compare this stuff and also why there isnt a global leaderboard called "Player with Most Impressive World Records" because that just wouldnt make sense.

This is why the site decided to allow players to pin a run and do custom game ordering/game grouping. That may not fully represent how good a run is, though. I'd be interested to see if the site dev can come up with an interesting way to customize in the way that @Breadn_ describes

A clear way to represent how "good" a run is, is by utilizing human theory TAS times, which are times that utilize RTA viable strats executed flawlessly. Or by comparing to a segmented run that uses RTA strats.

Any kind of statistic that utilizes pure player count is worthless, because when it comes down to it the vast majority of players in every game do not ever come close to reaching high level play in said game. It is usually only achieved by a small subset of players.

That all said, the main flaw I could see with this:

1st (5m 2s 067ms from best possible time)

3rd (1s 067ms from best possible time)

10th (067ms from best possible time)

Is that routes over time adapt and change as the time lowers and people start looking for riskier ways to save time. These supposed times would have to remove safety strategies, which would make the comparison in general kind of misleading. And it kind of means mods also have to know the actual best possible time to begin with, when mods aren't always the ones most adept at their game and are mainly good verifiers / organizers.

when mods aren't always the ones most adept at their game and are mainly good verifiers / organizers

I mean, the mods don't have to calculate the numbers themselves, they just have to input the numbers that other trusted community members come up with.

Anyway, yeah, determining the best possible time for any given category is kind of annoying, especially if people are actively running and/or strat hunting, so maintaining that would be a huge pain for everybody involved. For these purposes, it being horribly outdated is probably worse than it not being there at all, so I don't actually know how useful something like that would end up being long-term.

I'm not entirely against the simpler metrics even if they're technically not as valuable, but also if somebody actually really wants to know exactly how impressive your run is, it's usually easy enough to get a sense of that just by looking at the leaderboard.