Excuse the slightly provocative title and let me explain what this is about... I'm writing another front-page update that will become public in some days' time. I was writing about one particular genre of games that I felt was the most deserving, if any, to claim the hypothetical crown amongst all the genres that people do speedruns on. NOTE: this is only inside the sphere of speedruns. I realized I didn't want to leave it at that but instead would like to hear if others are in agreement with me. However, you will not hear my own opinion until the update is out!
What is the "King of All Genres" for speedrunning?
EDIT (19.8.)
Here's a summary of what everyone said. First all the genres named.
FPS
[J]RPG
Tetris-like
Platformer
Metroidvania
Racing
I didn't count the number of "votes" because it was a bit vague how that went. A few people pointed out that it's a pointless exercise, but... can't we just have a bit of fun? "It depends on how you define it" means you're doing it wrong.
So here's my own reasoning in some depth. As you'll see when the update comes out tomorrow, I chose something no-one here mentioned and I'm not actually incredibly surprised that they didn't. My choice is the RTS, real-time strategy games. There aren't a lot of runs for those games and there's a lot less of them to begin with than some other genres, and perhaps the crowd is a bit esoteric, off on their own orbit so to speak. My first RTS run I saw was for Dune 2000, which, in many ways, stands out amongst all SDA runs to date. Firstly, the runner 'Shaokhan' is a long-term competitive player and only started work on the speedruns much later as a challenge or demonstration. He even had an extra stipulation he set himself (to try to never have to use turrets, which are very powerful in defense) to see if it could be done. And when I say "if it could be done" it really was NOT a trivial run to produce, even by his high high standards of play. The game's hard mode is very difficult to succeed at without save/loading a lot. So yeah he was running hard mode, but also did easy mode, all the three factions and every alternate "form" of missions that had one, which totalled 84 missions in about 14 hours combined time. That's the first thing that catches your eye about this run though it's not really what's relevant to the discourse at hand.
When I started verifying it, I was doing the usual thing... I of course didn't really have any in-depth understanding of strategies, and though I had played the game and the original from 1992, I didn't really even know what the different strengths and weaknesses of each unit type were. I was looking for moments where the runner might have taken a little bit of time placing a building or where they lost units that they maybe could have saved etc. but I was really struck by how much game knowledge the guy was positively exuding in his in-depth replies to my many many probing prods. In fact if you'd like to see it, here it is! All the 5 pages of it. It ran for 6+ months though that's partially because it also had A/V issues and such.
The thing is, there is no way to fake this kind of knowledge or copy somebody else's work (I'm not saying you can do that with all other games). Every competitive match the guy had ever played, every new strategy he had discovered over the years, would have an impact on the way the runs shaped out. And the actual difference between an RTS like that (and I'm talking about missions with the traditional "just kill everything" objective) and all the platformers and all the FPS games and really almost everything I can think of, is because no two attempts are going to go exactly the same, the amount of improvisation you're being asked for is higher than anything else you can think of, other than maybe non-real time strategy games.
Think about it: the A.I. is governing when the enemy sends in their attack troops, which determines when you need to be home defending. Depending on its complexity and scriptedness (I'm still looking as the ideal of an RTS one where there is no scripting, which does, after all, happen), it might have taken different bases if it's like Starcraft, or maybe built different types of units. In order to beat a mission in Dune 2000, to return to what I know the best, you CAN have a large-scale plan of doing it in two big attacks, who you attack first, how many different types of structures you build etc. but the individual influence of many factors - namely the outcome of individual skirmishes, which buildings the enemy has bombed down, which way the enemy's harvesters are (it's good to get rid of them early if you can!) and generally whether you can have favourable trades in the area your troops are stationed at or whether you should retreat - will make decision-making a hugely complicated task. Even just knowing what your general plan needs to be might well come down to just experimenting, and it's very very difficult to call whether or not your approach is the most prudent one. What you do at all times during the attempt will slow or speed you up in a cumulative way for all the rest of it.
Now, I know, having seen runs from virtually every genre in public verification and during marathons etc., that other genres may well have aspects of this, but I cannot see one where it's all intermingled in such a Gordion's knot as in the ideal RTS. Everything is laid out in front of you from the beginning. It's not area-per-area. It's not corridor-per-corridor. It's the same table from start to finish. Like a game of chess, you can only really try to split it into phases like opening-middle game-ending, where the opening is basically doing your chosen build the same way every time (if you're still confident with it), middle game is where it gets messy, and ending is the part where you've basically won and only need to direct your ascendant troops to the remaining bases to eradicate them in an efficient way.
Further yet, and this is what I think starts to really set the ideal RTS apart from anything else, is the incredibly far-emphasized execution. The two pillars that produce a pro match between two players of Starcraft II (who I might just as well nominate as THE archetypal and iconic competitors in all e-sports, thanks to an entire nation dedicated to supporting them) are named the "micro" and the "macro". The macro is things like making sure you have all the buildings you want to have, your workers are always doing something, you have the right balance of workers gathering each type of resource etc. "Microing" is the low-level issuing of commands to troops to make sure they're taking good positions for each fight and wounded units are sent back.
There simply is NO TIME when the runner can idle during a run. There are hardly any moments at all when you can sit back and wait a bit. Unless you're using advanced movement techniques (which actually happens in many of the most popular platformers or in Quake), in other genres there tends to be times when you're JUST running down the hall, when you're JUST standing still in the ideal spot and aiming, when you're JUST falling down or climbing up a ladder, or when you're JUST waiting for the elevator to show up. In an unscripted RTS you switch focus between two or more places giving commands, all the micro, all the macro, at up to 400 actions-per-minute (which is the kind of maximum the Korean pros are sometimes seen at). It's almost like speed-typing, and each of those actions is still important. They're not all game-changing, but in the context of a speedrun, every one of them tends to have some small effect on your finishing time.
And that's why I think RTS is the "king of all genres" in speedrunning. It's not a bit of cruisin' here, and THEN a hard trick (even frame-perfect) there, and then an easy bit and then a hard bit. Heck, there isn't usually ever anything frame or pixel-perfect at all. But it's a continuous flurry of activity lasting through the entire mission that draws from such a major portion of your general game-knowledge, not just doing one thing exactly right at the exact right places. Nothing is skipped. Nothing is ignored.
So that's my words for you all! Maybe you'll agree, maybe you won't. I think that's enough of this discussion for me though but I might start another thread like this if something else inspires me later on.
What is the "King of All Genres" for speedrunning?
EDIT (19.8.)
Here's a summary of what everyone said. First all the genres named.
FPS
[J]RPG
Tetris-like
Platformer
Metroidvania
Racing
I didn't count the number of "votes" because it was a bit vague how that went. A few people pointed out that it's a pointless exercise, but... can't we just have a bit of fun? "It depends on how you define it" means you're doing it wrong.
So here's my own reasoning in some depth. As you'll see when the update comes out tomorrow, I chose something no-one here mentioned and I'm not actually incredibly surprised that they didn't. My choice is the RTS, real-time strategy games. There aren't a lot of runs for those games and there's a lot less of them to begin with than some other genres, and perhaps the crowd is a bit esoteric, off on their own orbit so to speak. My first RTS run I saw was for Dune 2000, which, in many ways, stands out amongst all SDA runs to date. Firstly, the runner 'Shaokhan' is a long-term competitive player and only started work on the speedruns much later as a challenge or demonstration. He even had an extra stipulation he set himself (to try to never have to use turrets, which are very powerful in defense) to see if it could be done. And when I say "if it could be done" it really was NOT a trivial run to produce, even by his high high standards of play. The game's hard mode is very difficult to succeed at without save/loading a lot. So yeah he was running hard mode, but also did easy mode, all the three factions and every alternate "form" of missions that had one, which totalled 84 missions in about 14 hours combined time. That's the first thing that catches your eye about this run though it's not really what's relevant to the discourse at hand.
When I started verifying it, I was doing the usual thing... I of course didn't really have any in-depth understanding of strategies, and though I had played the game and the original from 1992, I didn't really even know what the different strengths and weaknesses of each unit type were. I was looking for moments where the runner might have taken a little bit of time placing a building or where they lost units that they maybe could have saved etc. but I was really struck by how much game knowledge the guy was positively exuding in his in-depth replies to my many many probing prods. In fact if you'd like to see it, here it is! All the 5 pages of it. It ran for 6+ months though that's partially because it also had A/V issues and such.
The thing is, there is no way to fake this kind of knowledge or copy somebody else's work (I'm not saying you can do that with all other games). Every competitive match the guy had ever played, every new strategy he had discovered over the years, would have an impact on the way the runs shaped out. And the actual difference between an RTS like that (and I'm talking about missions with the traditional "just kill everything" objective) and all the platformers and all the FPS games and really almost everything I can think of, is because no two attempts are going to go exactly the same, the amount of improvisation you're being asked for is higher than anything else you can think of, other than maybe non-real time strategy games.
Think about it: the A.I. is governing when the enemy sends in their attack troops, which determines when you need to be home defending. Depending on its complexity and scriptedness (I'm still looking as the ideal of an RTS one where there is no scripting, which does, after all, happen), it might have taken different bases if it's like Starcraft, or maybe built different types of units. In order to beat a mission in Dune 2000, to return to what I know the best, you CAN have a large-scale plan of doing it in two big attacks, who you attack first, how many different types of structures you build etc. but the individual influence of many factors - namely the outcome of individual skirmishes, which buildings the enemy has bombed down, which way the enemy's harvesters are (it's good to get rid of them early if you can!) and generally whether you can have favourable trades in the area your troops are stationed at or whether you should retreat - will make decision-making a hugely complicated task. Even just knowing what your general plan needs to be might well come down to just experimenting, and it's very very difficult to call whether or not your approach is the most prudent one. What you do at all times during the attempt will slow or speed you up in a cumulative way for all the rest of it.
Now, I know, having seen runs from virtually every genre in public verification and during marathons etc., that other genres may well have aspects of this, but I cannot see one where it's all intermingled in such a Gordion's knot as in the ideal RTS. Everything is laid out in front of you from the beginning. It's not area-per-area. It's not corridor-per-corridor. It's the same table from start to finish. Like a game of chess, you can only really try to split it into phases like opening-middle game-ending, where the opening is basically doing your chosen build the same way every time (if you're still confident with it), middle game is where it gets messy, and ending is the part where you've basically won and only need to direct your ascendant troops to the remaining bases to eradicate them in an efficient way.
Further yet, and this is what I think starts to really set the ideal RTS apart from anything else, is the incredibly far-emphasized execution. The two pillars that produce a pro match between two players of Starcraft II (who I might just as well nominate as THE archetypal and iconic competitors in all e-sports, thanks to an entire nation dedicated to supporting them) are named the "micro" and the "macro". The macro is things like making sure you have all the buildings you want to have, your workers are always doing something, you have the right balance of workers gathering each type of resource etc. "Microing" is the low-level issuing of commands to troops to make sure they're taking good positions for each fight and wounded units are sent back.
There simply is NO TIME when the runner can idle during a run. There are hardly any moments at all when you can sit back and wait a bit. Unless you're using advanced movement techniques (which actually happens in many of the most popular platformers or in Quake), in other genres there tends to be times when you're JUST running down the hall, when you're JUST standing still in the ideal spot and aiming, when you're JUST falling down or climbing up a ladder, or when you're JUST waiting for the elevator to show up. In an unscripted RTS you switch focus between two or more places giving commands, all the micro, all the macro, at up to 400 actions-per-minute (which is the kind of maximum the Korean pros are sometimes seen at). It's almost like speed-typing, and each of those actions is still important. They're not all game-changing, but in the context of a speedrun, every one of them tends to have some small effect on your finishing time.
And that's why I think RTS is the "king of all genres" in speedrunning. It's not a bit of cruisin' here, and THEN a hard trick (even frame-perfect) there, and then an easy bit and then a hard bit. Heck, there isn't usually ever anything frame or pixel-perfect at all. But it's a continuous flurry of activity lasting through the entire mission that draws from such a major portion of your general game-knowledge, not just doing one thing exactly right at the exact right places. Nothing is skipped. Nothing is ignored.
So that's my words for you all! Maybe you'll agree, maybe you won't. I think that's enough of this discussion for me though but I might start another thread like this if something else inspires me later on.
Thread title: