After finding several new tricks, I decided to make an automated run to see what is now the absolute minimum possible time.

Several notes:

• Everything you see in this TAS is possible to do by a human. Some tricks require you to press or hold 3 or more arrow keys (at the same time), so they require some practice. I might do a guide for advanced movement/tricks if there is an interest.

• Almost all tricks here use the "mini-pirate" glitch in some variations. When in mini-pirate state, you can enter any crate from the side, and also do a big jump outside of crates.

• Level 12 route (left chest first) is probably the hardest trick to go for, I didn't manage to find any way to do this consistently.

• The start of level 15 in the TAS takes exactly the same time as an optimal regular route (put down the fire, jump to the right crate and go to left chest). I mainly did that trick in the TAS to show off.

Final times (in terms of milliseconds by counting frames): Level 1 - 1.666 Level 2 - 2.283 Level 3 - 1.733 Level 4 - 2.583 Level 5 - 2.95 Level 6 - 2.35 Level 7 - 2.966 Level 8 - 4.3 Level 9 - 4.25 Level 10 - 2.566 Level 11 - 4.466 Level 12 - 3.4 Level 13 - 2.3 Level 14 - 6.9 Level 15 - 4.683

This is just a name. In this game there are also "normal" and "hardcore" subcategories. Don't forget that you can define variables with any name and values.

https://www.speedrun.com/david0912/allposts

@ShadowFlare7799 There are speedrunning-related podcasts, the page link is in the main site menu.

That depends on how you intend to call the API. If it's just a direct call for testing in the browser, you can do something like this: https://www.speedrun.com/api/v1/leaderboards/9dowp9k1/category/n2yooy7d?var-yn217z28=zqo22g5l

Game "9dowp9k1" is "Mr. Bree Returning Home" Category "n2yooy7d" is "NG+" Variable "yn217z28" is the "NG+ mode" subcategory variable The value "zqo22g5l" is the "reversed" subcategory

1. If you want to get ALL runs of a certain game and category (which means also getting obsolete runs), use the "GET /runs" endpoint: https://github.com/speedruncomorg/api/blob/master/version1/runs.md#get-runs

You can filter by the game and category (and level for ILs) (and probably status for only verified runs). After getting the results, you can split the runs by the values of the subcategory variable (inside the "values" property). You can get a list of variables for categories in the "GET /categories/{id}/variables" endpoint: https://github.com/speedruncomorg/api/blob/master/version1/categories.md#get-categoriesidvariables

2. If you want to get only the latest runs in the board (which means verified non-obsolete runs), use the Leaderboards endpoint, which also groups and sorts the runs by places. You can get runs by full-game categories or IL categories. https://github.com/speedruncomorg/api/blob/master/version1/leaderboards.md#leaderboards

Instead of getting all results and splitting by the variable, you will have to do multiple calls to the API (one for each subcategory). In each call you filter by the relevant subcategory variable value.

Right click on the Livesplit window, you will see "Compare against", leading to options "Personal Best" and "Best Segments". You likely want to choose the first option. In the Livesplit settings, under "hotkeys" you can see hotkeys for "Switch Comparison (Previous/Next)". You might have pressed on one of them.

@Tomanyapples If you are talking about ILs, and a certain level don't need (or can't) be completed with glitches; then I think that allowing to submit this level for a glitched category is pointless.

@Bob-chicken You need to go to the request game page (as you would for any other game), and request "a new game" of category extensions to an exist game. The request can be accepted or rejected.

@MushySR There is no point in pointing at PvP games that were added in the past, before the current rules. The request rules don't apply to games retroactively.

[quote=Danone]I was looking for games so difficult that the number of runs was low because of the games difficulty and unforgiveness[/quote]

I see what you mean now, but that raises another interesting question: Are we able to tell why a game on SRC is popular (or not popular), and what are the reasons for that?

Also, you need to set a certain threshold for what you define as unpopular or "low number of runs". Lets say for the sake of discussion... 5 runners or less?

Well, I can think of several reasons for why some games on SRC might be not popular. I will sort them from what I think is the least relevant to the most relevant:

• As you said, difficulty. Game is so "difficult"/"hard" to play or run, that nobody wants to run it
• Bad accessibility in terms of speedrunning - Slow or obnoxious resetting process, unskipable cut-scenes, heavy RNG which you have zero control of, bad controls or bad game design in general (even in casual playthrough), lots of crashes or bugs
• Bad accessibility of game - Not easy to get to play the game; can't play on emulator; need old consoles or special equipment; game is pricey
• No exposure or publicity, game is obscure

The last point is the major reason for (un)popularity - if very few people have played the game or even heard of it, don't expect the masses to come. We can also discuss about what might make a game more popular than the other, but that's irrelevant right now.

Some extra things:

• I mainly run web-games right now (so in terms of accessibility, they're "perfect" cause anyone can play them and they're free). Many of the games from the past (especially flash games) were just not meant for speedrunners at all, but there are still some that are popular.

• There are some games that I personally wanted to run sometime (I have a large "to-run" list), but after trying to play them casually I ditched the idea because they were too difficult for me, or I just knew that I won't have any fun with them. But they are also popular. For example: https://www.speedrun.com/give_up - every obstacle or hazard is on global cycle, so dying doesn't reset the level, and can ruin your entire career https://www.speedrun.com/sik1 - Only one button to press, but there are several places where you need to press it in the PERFECT spot or else you die

• Some games can be difficult to play casually, but easy to speedrun (with relevant speedrunning tricks or something). It can also be the other way around. When you talk about difficulty, I guess you mean difficulty for speedrunning? Also, everything gets easier the more you practice it.

• As I said before, a game can have several time-ranges of runs that represent a certain skill level. So when you describe a game as "hard to speedrun", are you talking about getting runs to be below a certain threshold, or getting the world record, or something else?

If we take Celeste as an example, a (very crude) partition of time ranges of Any% can be: 1:30:00 or more - Easy to get after 1 playthrough of the game, you just need to know what you are supposed to do. 1:00:00-1:30:00 - Easy to get after some basic training in each level, and trying to minimize your deaths. No speedrun tricks/routes are even required 30:00-1:00:00 - You need to start incorporating speedrun tricks and routes to your runs, and practice the levels much more seriously less than 30:00 - you need to know and execute almost all known tricks, while doing everything very quickly, with very few mistakes/deaths. "Only" 41 people are in this level.

So yeah... Celeste is not an easy precision-platform game on your first casual playthrough; but after getting game knowledge and experience, speedrunning it CAN be easy, depends on what your goals are.

• Lots of games on the site have only 1 or 2 runners. That's just how it is, difficulty is irrelevant.

Go to the Resources page in the menu at the top of the site.

1. Define "overrated in terms of difficulty"

2. Difficulty is subjective. One runner can struggle with learning/running one game, but having lot of ease with a second game; while for another runner it can be the opposite.

3. Achieving "satisfying times" is subjective. One runner might run a game just for the fun of it, without much consideration for the end result; another runner might want to run a game up to a certain time-range that represents a certain skill level; and another runner might not be satisfied with anything less than the world record.

4. You are confusing DIFFICULTY with POPULARITY, which have very little in common. If you decide that a certain game is hard for several reasons, does it really matter if 1000 people are running it, or 100, or 10, or even just 1? Who knows, there might be a game on the site that requires more technical skill than top runs of Celeste or Portal, but nobody knows about it other than the runner who requested that game in the first place (and again, "more technical skill" is also a subjective term). You said that hundreds of people are capable of getting good times in the game, and so it makes the game look more "easy" for you. But if a game is popular, and lot of players are trying to achieve fast times, then of course lot of players WILL achieve fast times.

Since you asked, my vote for the hardest game I know will be "Getting Over It". In general, I am and always was bad at fast/precise reactions with the mouse. In this game, it took me about 27 hours of play time before I gave up, and I also never managed to get past the cliffs above the orange. At the end, I got good at the very early part of the game because I had to repeat that many many times, but it's nothing compared to what you can see the speedrunners are doing.

If I will limit myself only to games that I did run so far, and actually finished, and actually enjoyed playing - The hardest game for me is "Robot Wants Kitty". Runs can be very short, but to get top times you gotta be super precise with your movements and shooting, and even one mistake usually means resetting the game.

Perhaps you will have better results with that, if you ask in the Astroneer forums or in their discord channel. https://www.speedrun.com/astroneer

From what @R0main said about deleting games from the site - game moderators can delete their game in the first 14 days, after that only full-mods/admins can do it (maybe not if there are verified runs in the leaderboards, dunno).

Anyway, the button for deleting games is still visible; in the top right corner of the "edit game" page:

I don't know what will happen if I press on that and I obviously don't want to try. This button should be removed entirely from the page, if the criteria for deleting the game is not valid anymore.

(Not sure if that belongs to site bugs)

Yes, remove the "default" property for the "java edition" value. Leaderboards are always filtered by default values for any variable, when you open or reload the game page.

That's not how this works. Only game moderators can delete their games if they wish. Also, there are many highscore based games around, all of them were accepted to the site before submission rules against highscores came in place (and they don't apply retroactively).

Welcome to the site! As shadowflare said, random suggestions rarely work without knowing anything about your preferences. A general advice is to start speedrunning games that you like and enjoy playing casually, and start from there to other games.

@dha then you can request moderation for those games, after submitting runs in those games. I'm sure you will not be interested in moderating ANY random web game, especially when you don't and never will run it.

I like how this thread gradually gets more posts, even when the original poster has deleted himself from the site.