Response to Vorpal's Speedrun Cheating Discussion

Response to Vorpal's Pastebin:

Vorpal's original discussion: https://pastebin.com/nqNjTnuc

-- Personal Opinions: --

I think context is everything.  Black and white rules don't cover every case.  Guidelines can be a good starting point.  I think most communities handle this stuff internally, and I think most of the larger communities do it pretty well.

I think the wider sentiments are somewhat just people being addicted to drama, and somewhat people being uncertain on how to evaluate these things.  How does the average viewer have any concept if the run isn't cheated?  How does a person have any concept if Wikipedia is accurate?  How do you know if a guy actually did a run 30 years ago?

"1. How strong does evidence need to be before you support removing a run from a leaderboard?"

In my opinion neither "innocent until proven guilty" nor "guilty until proven innocent" are the right idea here.  I think the more appropriate question is "Does this run meet the burden of proof the community expects for the board."

When players show up out of nowhere and beat runs, it makes sense to be suspicious.  It is often on the runner to provide the burden of proof that they performed and are capable of performing the run.  On the flip side, I don't think you should call someone a cheater unless you've supplied the burden of proof that they've actually cheated.  A rejected run doesn't have to imply someone cheated.  Equating a run that doesn't meet the burden of proof to cheating misrepresents the very gray area in between.

What's really bad:
- Splicing.
- Intentionally misrepresenting the circumstances of play.
What's not so bad:
- Being bad at reading rules.
- Not understanding community norms and expectations.

The runs listed on the board should represent community consensus.  One vague way to tier this might be as follows:
1: Runners of the game.
2: Runners of similar games (Same console or series)
3: Wider community standards/expectations in general.

"2. Some leaderboards do not require videos for submissions. Others require current submissions to have video"

At SGDQ 2014, the runners with the top 3 Bomberman Hero times at the time played for the marathon.  I played the new WR at the time of 44:16 unrecorded while practicing, and the other two runners witnessed about 2/3rds of the run.  Later during the race on stream I finished about 1 minute behind that, largely due to a couple very specific mistakes.  The run is listed on speedrun.com here, with no video: https://www.speedrun.com/bhero/run/wzpwg8yq  It is still my current PB.  If the other runners wanted me to not list this run on the boards because it doesn't have a video I'd be fine with it.  As it stands, in the game's community's opinions I've supplied the burden of proof for that run.  With this knowledge, would most people actually think that I didn't do it?  Or is this terrible precedent and community nepotism and the reason speedrunning needs global rules.  I tend to think that's a bit too authoritarian and doesn't consider the context.  It's not a court of law, we can define our own precedents and expectations.

I think the size of the community and the prominence of the game is also a huge factor in burden of proof requirements.  I don't think it's practical to say that all small communities need to follow the same standards as the most competitive games.

"5. Many of the ways suspect runs were detected was through greater understanding of the game"

In my opinion, 99% of cheaters are too lazy to be thorough enough or knowledgeable enough to cover their tracks from the most knowledgeable players.  The players with skill who also cheat are the hardest to identify, but they're also quite rare in my opinion.  The reputation one would gain from being a high profile person who cheated, is most often the reason skilled people won't cheat.  The reputation of the runner is one factor that gives high confidence in many prominent records.  When cheating happens in prominent games you definitely hear about it, but it also doesn't happen that often.  Most prominent runners stream their attempts live now, which supplies a much higher burden of proof than anything more than about 7 years ago.

Additionally: Some very popular games have begun imposing high burdens of proof on top runners.  Super Mario Bros is probably the foremost example of this.  One high confidence way to make sure someone isn't cheating in this very prominent game is to more or less require them to stream hundreds of attempts so their overall skill can be compared to their PB.  A random run by a user no one knows isn't possibly going to meet the burden of proof that community, or the wider community in general expects.