Twin Galaxies recently brought their database of records back online. I browsed through some of them. There are plenty of issues + reasons to not take their leaderboard seriously at all. Among the many reasons, one reason is the huge amount of clutter and useless dupe categories everywhere. This got me thinking about SRL Leaderboards, as we have been making progress not only in the planning, but the design, coding, and implementation.

One concern I have that hasn’t been figured out is the issue of category prestige. In a perfect world, categories would be easily definable with a blanket rule across all games. SDA sort-of tries to implement this a little bit with their Any%/100%/Low% ideology. Unfortunately due to the huge numbers of exceptions to the rule & the amount of “forced” 100% and “vaguely defined” low% categories, I’ve come to the conclusion that blanket rules do not work for separating out categories across all games. I think most people understand this.

The problem is that many games do warrant having more than one category. The question is how far do you go? This joke page shows what might happen if you went wild and allowed a category for every possible thing in a game.

This page, despite being a joke, actually got people to go turn on their game and snatch up a couple of the “records.” This shows me that if it is built, people will submit for it, despite it being ridiculous. This puts the responsibility in the hands of the people adding categories to the site to maintain a leaderboard system that encourages competition, which is a big ideal of SRL.

The huge cluttering at TG is a great example of their being so many “world records” to go snatch up, that it spreads the competition ridiculously thin, and makes getting the high score or a good time pretty meaningless. On SRL we want the amount of categories to be limited to encourage competition and make more exciting stuff happen such as the dxtr vs funkdoc battle for the Batman record, or Garrison vs kottpower trading the Super Metroid record back and forth. This is exciting not just for the runners but especially the speedrunning audience, which is growing larger every day (at an alarming rate).

After much discussion, I think the blanket rule (I know this is dangerous, but I have not seen a counter-example) of requiring the game to start at the beginning and end at the end is a good first step. This prevents things such as ALTTP Master Sword or Pokémon Red/Blue Beat Misty, which is obviously good. From there it gets more tricky.

Categories like SMS 78 Shines or OoT No RBA/WW have had their runners in the past (and some even recently) working hard to get good times, despite it not fitting as cleanly into a “fastest completion” type of goal. It is hard to say where exactly this line would be drawn. If all of these types of categories are allowed, it would spread out the competition and risk starting to clutter the leaderboard. If they are not allowed, that is somewhat annoying to the actual runners of the game whom might want a resource to find such times.

The truth is categories have differing levels of prestige. From an outside observer’s perspective though, this may not be evident. To them, if it appears on the leaderboard it will be “legit,” and if does not appear it won’t be. In the interest of appealing to people “not in the know” one solution is to axe the stuff that is “less legit,” which can be somewhat subjective at times.

I’m not sure if a tiered category system could somehow get onto the SRL Leaderboards that would imply correctly to an outside observer that “yes, this is one way people play this game, but it isn’t taken as seriously—here are the fastest times for it anyway.”

Not sure how to handle this problem, but with the growing speedrunning audience, it must be handled correctly so we don’t end up with the PAL 200% Speed Emulator ALTTP Glitchless Master Sword WORLD RECORD and turn into TG.