I'm late to this post, but I could write a book on all of this stuff. I'll try to keep it brief though. Gonna put asterisks around my conclusions.

I haven't been on Twin Galaxies all that much in recent years. I check in here and there because this is where I started. I still have a lot of gripes about Twin Galaxies, but I think Jace has done a nice job in repairing the damage to an extent. I think a number of things at TG can't really be fixed ultimately though. There's no easy solution to a lot of the scoreboard issues among other things.

*TG has a bad reputation with speedrunning and cheating on the whole. There's many (especially older) scores that can't be verified as legitimate which is problematic when people are still competing on those categories. Pole Position is one title where I feel Daniel Yamnitz is being cheated out of a tied WR due to a misprint or misconduct of former scoreboard keepers. That's just one example of very many, but it's impossible to prove the WR didn't happen (I can make a very good case for Pole Position which I plan to do eventually).*

I tried to explain to the old guards and the TG refs that their version of speedrunning was very unpopular. The arcade mindset/rulesets don't transfer well to speedruns. Glitches in high score often break the competition and so banning specific things makes sense. With speedrunning, it's kind of the opposite where the glitches in a way keep the games alive and interesting for competition. There's no reason why TG couldn't track both of these things, but TG has always been very anti-glitch. TG could have been THE place for speedrunning had they played their cards right, but no one at TG wanted to listen to the speedrun community.

*Speedrunners are unlikely to submit to TG even if TG created the same categories simply because SRC already exists*

SRC (speedrun.com) is very relaxed on submissions, but per game community there's different levels of verification from the people who know the games best. There is video for every submission and concerns can be brought to game mods who will hopefully address them. The console boot-up idea at TG isn't a horrible idea for SRC to copy even, but resets should really be acceptable too unless there's an issue on a specific title regarding resets. TG boot-up was mostly a means of proving it wasn't emulation. Emulation is tracked alongside console at SRC and those are typically the only runs I really question there. Anyhow, cheating is generally detected through gameplay and not whether someone uses emulation these days and so console power-up is a bit overkill, but still understandable at TG since each platform is tracked separately.

*Since video exists for both TG and SRC the legitimacy is fairly reasonable for both websites. Each SRC game has dedicated mods that know the game they are verifying so all being said I think 99.9% of submissions to either website is legit. It's just much harder to cheat with video, but things still slip by rarely. Video itself without live stream context makes TG more questionable, where with SRC you can often track someone's progress on Twitch history. SRC is more questionable when it comes to emulated runs*

The way to fix Twin Galaxies in my opinion would be to grandfather every score/scoreboard which has nothing to substantiate it. That would basically be anything pre-1998 as far as I'm concerned. The issue there is obviously that each game can only have one world record holder. Many 80s/90s scores are obviously legitimate, but so many of them aren't. Having modern competitors compete against scores & times that might be impossible (and are impossible to verify) means a lot of people aren't getting the credit they're due.

Again, at the same time you don't want to throw away all the accomplishments pre-1998, but many are questionable and there's no denying that. I actually do enjoy the mystery surrounding such scores, but as a competitor it would be extremely frustrating to spend months, potentially years, chasing a score that is actually impossible or just didn't happen.

Sadly too, even post-1998 scores have been verified by a lot of corrupt referees and so even those aren't completely trustworthy, but certainly more than pre-1998. Still there are barely any videos still in existence so how is one to know which are and are not trustworthy?

What does that leave? Runs that are on video and runs that were done at live events. Everything else doesn't have enough credibility to call a world record from where I stand.

EDIT:

If I were in charge of TG leaderboards I would:

- Give the option to also sort scores by just Legacy or just TGSAP

- Take the live event scores from the 1980s Guinness books' scores and classify them under TGSAP verified (upon review). Possibly include other live event scores.

- Keep ESI calculation as is, but possibly only rank/place TGSAP scores and live event scores as (1st, 2nd, 3rd...) as being the *official* records

EX:

John - Missile Command - 100,000,000 - June 1984

1st: Frank - Missile Command - 90,000,000 - June 2022

Mark - Missile Command - 80,000,000 - June 1984

2nd: Josh - Missile Command - 70,000,000 - June 2017

I think I would do something close to this, if not exactly these things.

Jace Hall thanked this post

Originally Posted by andrewg

The issue there is obviously that each game can only have one world record holder.

Speed run typically has multiple holders per game already so I'm curious if this is meant as it reads? Anyone* can create a new track at TG and therefore there can be many different holders of records per game. If the old TG rules don't fit, make a new track which does...

* OK, not anyone but verified members with sufficient CR +SP which is a lot less than anyone. :P

Originally Posted by andrewg

EDIT:

If I were in charge of TG leaderboards I would:

- Give the option to also sort scores by just Legacy or just TGSAP

- Take the live event scores from the 1980s Guinness books' scores and classify them under TGSAP verified (upon review). Possibly include other live event scores.

- Keep ESI calculation as is, but possibly only rank/place TGSAP scores and live event scores as (1st, 2nd, 3rd...) as being the *official* records

I think I would do something close to this, if not exactly these things.

One of the things that is on the twin galaxies “ to do “ list is to add user defined filtering/sorting to the database toolset for tracks as well as ESI display.

This will allow people to choose how they want to view the leaderboards, as well as see under those viewing circumstances what the ESI calculation is.

TG will retain its preferred method as default display but ultimately it will be user option to decide how they want to sort/value the data.

This should address most of the preferential concerns outlined above without impacting or taking anything away from historic performances.

andrewg, Rotunda liked this post

I really meant each category of a game can have only one WR holder. I was thinking primarily of arcade games while writing this because that's what most legacy scores are and there's usually only one category.

The more I think about it the more I think there should be legacy WR holders and TGSAP WR holders and each should be able to claim title to WR if the legacy score happens to be higher than a TGSAP score. I mean, I understand that the scores can be challenged, and in this way it means TGSAP scores can potentially replace legacy scores. If Dragster were longer than 5 seconds we would still be arguing about it. It's likely the only game that could be essentially "brute-forced" to prove that it didn't happen.

A big issue I have with the arcade scores is that many rulesets were added later. We added rulesets for how we should be competing now, but it's not always reasonable to say that those were the same settings the players of the 80s used. Literally I only trust the TG scores that made it into the 1980s Guinness books. Everything else *could* have happened, but with the assumption that: any settings, any number of players, any number of credits, breaks, or glitches could have been used to achieve the score.

----

* Twin Galaxies' database pre-1998 is mainly a compilation of scores seen throughout magazines without any regard to rules, legitimacy, or accuracy.

* The issue is just that it's not always possible to disprove a score.

Barthax thanked this post

bensweeneyonbass liked this post

More regarding the main topic: I think getting speedrunners to submit to TG would be very tough. A collaboration effort with SRC for times I guess could be possible, but I have no idea how that would be possible. Idk what Guinness is doing these days, but TG could maybe bridge the gap from SRC to Guinness or something.

I don't even think it's so much that people don't want to submit to TG as it is that people just want to submit to SRC. People don't really submit to Speed Demos Archive anymore when they could. It's just that SRC is the most popular speedrun website right now.

People will move on from SRC eventually I think since the new owner(s?) have started cluttering boards with ads and making things a bit less than perfect.

I like the direction that TG has been headed. Viewable video proof is absolutely the biggest and best change from before. Bounties have been picking up stream for speedrun achievements, but there isn't a great central hub or collaboration.

Here's one example:

https://turboboard.io/

Anyway, speedrun bounties might stir up interest. A lot of speedrunners don't seem too interested in bounties or competitive events for whatever reason, but the more serious players usually are.