Speedruns to Watch

Have you never watched biathlon? The rules speedgolf needs are already there for us to see. In biathlon, for each missed shot, you ski an extra loop of 150m before leaving the shooting range area. This means it’s still all about whoever gets across the finish line first (in the shortest time), and nobody cares if how many targets you missed, as long as you missed fewer than the competition, or made up the extra loop distance faster than they did the 5km laps between the shooting.

So for speedgolf, we can treat each hole, or set of three holes, like a biathlon lap plus a shooting.

Each hole has a par. For each stroke above par, you have to run an extra loop of about 200m. These loops could be done between every third hole, and you run all the loops you must before moving on to the next hole. So between hole three and four there would be a 200m loop. If you complete hole one below par but hole two above par, and they equal out, you don’t have to do a penalty lap. This means below par shooting is still encouraged to “bank” some shots to use up later if needed.

In biathlon, you sometimes get a few extra bullets, so if you miss one of your shots, you can spend those bullets to reduce the time spent doing the penalty loops. It takes some time to reload the gun, but it’s less of time penalty than a 150m loop.

Speedgolf could have three “free shots” that you can take at some point around the course, which you could save in case of really wayward hits.

The length of the penalty loop could be adjusted, of course, to balance for golf skills vs running speed, but both would still be dependent on each other.

The issue with having to do a penalty loop after the 18th hole would be the that final moment would be running across a finish line. This wouldn’t do at all! The game must end with the ball going in the hole!

So I propose there being a loop at the end of the 17th hole, where you get all your previous penalties out the way. Then there would also be a penalty loop right at the 18th green! If you reach the green at one below par, you just go for the putt, or first do a penalty loop for how many shots above par minus one.

Then for each miss on the 18th green that puts you above par, you have to do a 200m penalty loop. This means that your clock stops the moment the ball goes in the 18th hole, and no matter what, you don’t have to do any running following that moment.

By the way, biathlon is super fun to watch. In Germany it’s way more popular as a spectator sport than golf, as far as I can tell. The balance of ski fucking hard mixed with can you get your body calm enough to do precise target shooting is a really cool concept.

Biathalon isn’t a bad idea for speed golf. The problem is that a biathlon course is designed for several simultaneous skiiers and shooters. A golf course, you have to wait behind the people in front of you so as to not interfere. Everyone would have to run the course individually and then compare times at the end.

I also think golf won’t need any side loops to run around. There would be an easier way. When a player arrives at a tee, their score drops equal to the par of the hole. So if the first hole of the course was a par 5, everyone would effectively start at -5. After getting the ball in the cup, players can only proceed to the next hole if they have a score of 0 or less. So if you make par you continue. If you go under par, that’s one or more bonus shots you can miss in the future.

Players that go over par will have to run back to the start of the hole. When they arrive at the tee, their score will decrease by par +1 since it is their second time doin the hole. Third time doing the hole their score decreases by par +2, and so on. The hole gets easier and easier each time you redo it. But just one really bad hole, like a +3 might get a player stuck on one hole for quite awhile. Meanwhile an eagle is going to let a player zoom through.

The only problem is I think that this could end up being more running than a marathon, even for incredible golfers who rarely go over par. Maybe just keep it to a very small number of holes. Like maybe 3-5?

And also so you don’t get beaned with a Small, hard, plastic-and-rubber ball. Those suckers are really moving when they’re hit, you don’t want to be in front of that, it fucking hurts. And unlike a Biathlon, courses are not designed so that you’re never shooting in the same direction someone is moving, they handle that with rules instead.

Speedgolf is a speedrun, not a race.

I think all the extra biathalon stuff would get in the way of the pure speedrun nature of the OG idea in terms of making it fun to watch and fast paced. Slowing it down by adding extra distance for people to run around in circles or whatever just seems kindof arbitrary.

Easier to just have the par 5 for each hole and if you go over there’s a standardized time penalty added up. Chances are everyone will end with 3-4 penalty strokes on their time, so knowing when to strategically use them in a round might be vital. But IMO the idea should from a spectator perspective be a person hitting the ball as quickly as possible in a direct path towards the hole until sinking the ball, and recording that time live. Add/subtract from the raw time at the end with penalties/bonuses and get the final score.

And I think by and large Rym is right that if someone wants to try just dribbling the golfball down the entire fairway, fine. It probably isn’t a WRP tactic, but for some courses it may be a useful tactic to employ in certain areas, just like grenade juggling was vital to Runway records in Goldeneye. Maybe that ends up being the fastest tactic by and large? But is it repeatable or only for a TAS? It would take testing to see if the game is that broken and then you can have categories such as “no dribble” or “no woods” or whatever emerges.

Definitely I’d break it into single segment/hole runs where you might compete to get world records on a specific hole, in addition to cumulative round games.

The other thing would be to take a few other cues from shooting sports with different classes. do you have, say, Caddy Class where you have a caddy who shares the load of the gear and runs with the golfer, BUT you only have access to what the both of you carry on your person through the whole round. No golf cart, no refreshment tables or tents. If your caddy doesn’t have water bottles in the bag, you might be pretty thirsty by hole 10. So you’re running hole to hole with your team mate, deciding the strat for the next hole while in transit, caddy is swapping clubs on the fly from a specially designed speedgolf bag and so-on.



This is real good.

I want to make one on “fastest time from faceoff to goal” in hockey. There are some sub-second ones out there.

I knew Mark Buehrle’s name and not much else. Thank you Internet for giving me some appreciation.

The style of the video was good too. Borrow the framing of speedrun talk but lay it over sports. I love that kind of cross-pollination.

Foolish Baseball is a really good channel. I’m not that deep into baseball as a sport itself, but his videos are giving me a some deeper appreciation for it. While he does use the 8-bit aesthetic for all of them, this is the only one that does explicit comparison to video gaming contests, but he does talk about other obscure superlatives, like for example the guy who gets on 1st a lot because he gets hit by pitch a ton.

Anyway, I strongly recommend the channel.

Center-ice or zone? Do own-goals count? Or are these different categories. Anyway, SBNation has a series that goes in a similar direction with High Score.


Board game speed running…

My manager is now the world record holder for single player money puzzle exchanger.



1 Like

There’s a video about the new Any% run of SMB3 in 3 minutes. It’s very satisfying from an engineering point of view, but very unsatisfying that this is in the same category as, and taking the record away from, runs that complete the game via gameplay, not via code execution.

I’m sure the debate has been had many times in the speedrunning community, but I’m very much of the opinion that when someone is using knowledge of memory allocations to load end-screens, that should be its own and different category. It’s not using an in-game-discoverable glitch, but an on-cartridge-memory-hack.

I was struggling to find an analogy to this in other sports, and the closest I could come to was motorsports. But really, it’s an esport specific problem, as all other sports “run” on the physics engine and programming of reality itself, and only esports has a way to hack the rules of the universe.

So an analogy would be something like an “any% of an F1 season” where Sebastian Vettel drove his car in exactly the right pattern to stop time, transport himself to Abu Dhabi, and then show everyone a photo of him holding the driver’s championship trophy taken in 2012. Sure, he’s in the right place and showing the right image…


I understand where the frustration comes, but I also think it makes sense that any% is the “anything goes” category and any restrictions are then their own modifiers and categories. If that’s how the people running the games like it and find working, that should be enough.

Right. But does “anything goes” include tool assisted runs?

I’m not frustrated, just unsatisfied that code execution is considered closer to “any%” and not closer to “tool assisted” because it feels way more like that to me.

I know how it usually works, in that a new category will be “any% no code execution”, but then, for me that one becomes the interesting category, and the categories that allow code execution become as interesting to me as tool assisted runs.

Except clearly this was not tool-assisted, in the jargon of speedruns. Throwing out runs that use human knowledge of the game state would be far too limiting. Human execution is the key—and the limit being tested.

Speedrun classes get renamed and clarified all the time, especially when something new breaks a beloved class. That said, Any% seems to fit this particular run.

I know it’s not tool assisted. I know the jargon of speedruns.

I’m saying using code execution feels like as big of a break away from normal gameplay glitches as tool assisted runs.

The “human skill” required for Ocarina of Time code execution is things like “prepare a third controller in advance with buttons held down with tape and elastic bands” and “cut notches into controllers to allow the joysticks to be moved beyond their normal boundaries”.

And I know that this is also not tool assistance, but it feels to me to be as big of a gap from normal gameplay to be able to enter values of six digit precision with an analog joystick than the gap from normal gameplay to tool assisted runs.

Knowing the interior workings of the game to improve chances of shaving seconds off a record feels (to me) fine.

Knowing the interior workings of a game, then exploiting those to sort through in-game memory to find where the code can be executed, then using programs to discover what the horizontal values of an enemy must be to load the final screen after the game crashes feels (to me) both incredibly difficult, but also a fundamentally different activity. An activity that feels (to me) as different from normal any% speedrunning as normal any% speedrunning is from tool assisted runs.

Yes. What you’re saying is correct. Any% speedrunning is basically a completely different thing than the other categories for most/many games. It’s often more about understanding computers than about being good at video games. That’s why they have the categories.

1 Like

I guess it’s that only the biggest games (like Zelda and Mario and Pokémon) have the dedicated communities to spend the resources to find the opportunities and dig into the technical information to get code execution hacks running.

It means that for these big, popular titles, any% becomes about code execution and pre-set controller inputs and modifying controllers.

And for smaller games (literally every other game) any% remains a human-skill oriented category.