Speedrunning Guide: Looking for input and thoughts

Omnigamer

Edit history:

Omnigamer: 2013-11-25 11:44:22 am

Omnigamer: 2013-11-25 11:41:16 am

Omnigamer: 2013-11-21 12:04:35 pm

Omnigamer: 2013-11-21 12:03:37 pm

Omnigamer: 2013-11-21 12:01:22 pm

2013-11-21 09:58:55 am

All the things

Something I've heard from stream viewers, beginner runners, and even seasoned runners is that they don't know where to begin or how to approach a speedrun that they'd like to try. It doesn't matter whether it's an often-run game or a game that has no known run; there is a fundamental process to organizing and executing a speedrun that many people do not know how to apply. The issue is that there's no single established resource that formalizes this approach. This is something that I'd like to change.

I've thought about writing some variety of "guide" for speedrunning for some time now, but have been putting it off for various real-life reasons (among laziness/procrastination). With the way this young hobby is growing, the demand for something that gives people enough general information and tips to get started will explode before too long. There are pieces of this available on the KB, but the articles themselves are scattered and not cohesive.

I'm not thinking of making a "Beginner's Guide" but something more comprehensive. It shouldn't focus on declaring a process and having new players follow it. Rather, it should detail the "stages" of a speedrun with enough depth that even experienced runners can go back to it for tips on how to look for new things or increase their efficiency. It should be accessible to runners of all skill levels and experience. The end goal is a more informed speedrunning base with abilities to push runs further and increase quality.

I don't pretend that my knowledge and experience alone is enough to make this happen. As such, I'd like to ask the community for their own input and experiences to help in laying out productive methods for learning, routing, and optimizing runs. I have a rough outline of everything I'd like to cover at the bottom of the post; I can explain pieces of what I'm targeting in more detail, but I'm hoping to not write entire sections of the Guide just to explain my direction. I'm definitely open to criticism about which sections should be removed/merged/added, so feel free to discuss anything you feel might be missing or inadequately placed. The final product should be a comprehensive document that is great when read in its entirety, but each section should be able to stand on its own.

Another section that I'd like to include is one dealing with the philosophy of speedrunning. In particular, I intend to set up reasoning for what a speedrun inherently is, as well as discuss many pieces that are taken for granted by the community. These things include the perspective on glitches, multiple categories, RTA vs segmented, and real-time vs game time. This is not meant to set up rules or anything like that, but rather to bring up reasoning for why various things are used or accepted in the current state of the community. I am often asked by non-gaming friends or colleagues why I speedrun at all, or why we "cheat" (use glitches). This section should provide some answers to those questions. It's also incredibly subjective; I am definitely open to hearing discussions on these items.

This is not something I expect to have finished within a few days or weeks; the first draft could easily take 3-4 months. I opened this thread not only to get input for the various sections, but also to discuss the interest and value of such a guide going forward. With some effort I think we can have a great set of documents with plenty to support current and future runners, and even non-runners.

1. Introduction and overview of speedrunning in its current form

2. Speedrunning Philosophy
+What is a speedrun?
-Shortest path solution
-Personal trial
-Competitive medium
+Points of Controversy
-Use of glitches
-Real time vs game time, other timing methods
-Categories

3. The Speedrun Process
+Three phases: Investigate, route, execute
-Not one way track; feed back into each other
-Collaboration
-Documenting
+Picking a game
-Fun and time
-Tedium factor
-Obscurity

4. Investigation: How does stuff work?
+Adjusting to gameplay
-Initial playthroughs
-Mechanics exploration
+Checking resources
-Technical Information & guides
-Past runs/TASes
-The game as a resource: accumulating data
+Understand the game
-What are the limits of what you can do?
-What does randomness affect?
-Why do events happen as they do?

5. Routing: Building a game plan
+Point-to-point planning
-Enumerate goals, requirements
-Consider backup plans
-Audio/visual cues
+Pathing
-Minimal event sequences
-Pattern recognition and exploitation
+Testing
-Empirical comparisons, assumptions
-Consistency checks
+Improving
-Ask "What if..." questions
-Experiment
-Change assumptions, risks

6. Execution: Finishing the run
+Practice techniques
-Chunking
-Muscle memory
-???
+Optimization
-Movement
-Luck trade-offs
-Consistency improvements
+Review teh urn
-Technical errors
-Luck vs execution
-What can still be improved?

7. Miscellany: Tools, tricks, and other resources
+Emulators & advanced investigative techniques
+Common tricks/terms (links, not full explanations)
+Recording equipment & related hardware

Thread title:

Something I've heard from stream viewers, beginner runners, and even seasoned runners is that they don't know where to begin or how to approach a speedrun that they'd like to try. It doesn't matter whether it's an often-run game or a game that has no known run; there is a fundamental process to organizing and executing a speedrun that many people do not know how to apply. The issue is that there's no single established resource that formalizes this approach. This is something that I'd like to change.

I've thought about writing some variety of "guide" for speedrunning for some time now, but have been putting it off for various real-life reasons (among laziness/procrastination). With the way this young hobby is growing, the demand for something that gives people enough general information and tips to get started will explode before too long. There are pieces of this available on the KB, but the articles themselves are scattered and not cohesive.

I'm not thinking of making a "Beginner's Guide" but something more comprehensive. It shouldn't focus on declaring a process and having new players follow it. Rather, it should detail the "stages" of a speedrun with enough depth that even experienced runners can go back to it for tips on how to look for new things or increase their efficiency. It should be accessible to runners of all skill levels and experience. The end goal is a more informed speedrunning base with abilities to push runs further and increase quality.

I don't pretend that my knowledge and experience alone is enough to make this happen. As such, I'd like to ask the community for their own input and experiences to help in laying out productive methods for learning, routing, and optimizing runs. I have a rough outline of everything I'd like to cover at the bottom of the post; I can explain pieces of what I'm targeting in more detail, but I'm hoping to not write entire sections of the Guide just to explain my direction. I'm definitely open to criticism about which sections should be removed/merged/added, so feel free to discuss anything you feel might be missing or inadequately placed. The final product should be a comprehensive document that is great when read in its entirety, but each section should be able to stand on its own.

Another section that I'd like to include is one dealing with the [i]philosophy[/i] of speedrunning. In particular, I intend to set up reasoning for what a speedrun inherently is, as well as discuss many pieces that are taken for granted by the community. These things include the perspective on glitches, multiple categories, RTA vs segmented, and real-time vs game time. This is not meant to set up rules or anything like that, but rather to bring up reasoning for why various things are used or accepted in the current state of the community. I am often asked by non-gaming friends or colleagues why I speedrun at all, or why we "cheat" (use glitches). This section should provide some answers to those questions. It's also incredibly subjective; I am definitely open to hearing discussions on these items.

This is not something I expect to have finished within a few days or weeks; the first draft could easily take 3-4 months. I opened this thread not only to get input for the various sections, but also to discuss the interest and value of such a guide going forward. With some effort I think we can have a great set of documents with plenty to support current and future runners, and even non-runners.

1. Introduction and overview of speedrunning in its current form

2. Speedrunning Philosophy
  +What [i]is[/i] a speedrun?
      -Shortest path solution
      -Personal trial
      -Competitive medium
  +Points of Controversy
      -Use of glitches
      -Real time vs game time, other timing methods
      -Categories

3. The Speedrun Process
  +Three phases: Investigate, route, execute
      -Not one way track; feed back into each other
      -Collaboration
      -Documenting
  +Picking a game
      -Fun and time
      -Tedium factor
      -Obscurity

4. Investigation: How does stuff work?
  +Adjusting to gameplay
      -Initial playthroughs
      -Mechanics exploration
  +Checking resources
      -Technical Information & guides
      -Past runs/TASes
      -The game as a resource: accumulating data
  +Understand the game
      -What are the limits of what you can do?
      -What does randomness affect?
      -Why do events happen as they do?

5. Routing: Building a game plan
  +Point-to-point planning
      -Enumerate goals, requirements
      -Consider backup plans
      -Audio/visual cues
  +Pathing
      -Minimal event sequences
      -Pattern recognition and exploitation
  +Testing
      -Empirical comparisons, assumptions
      -Consistency checks
  +Improving
      -Ask "What if..." questions
      -Experiment
      -Change assumptions, risks

6. Execution: Finishing the run
  +Practice techniques
      -Chunking
      -Muscle memory
      -???
  +Optimization
      -Movement
      -Luck trade-offs
      -Consistency improvements
  +Review teh urn
      -Technical errors
      -Luck vs execution
      -What can still be improved?

7. Miscellany: Tools, tricks, and other resources
  +Emulators & advanced investigative techniques
  +Common tricks/terms (links, not full explanations)
  +Recording equipment & related hardware

doicm

2013-11-21 10:05:01 am

I know that there was another thread that talked about the science behind speedrunning...could be useful.

Not everyone has the same approach when it comes to speedrunning. For example, I spend most of my time planning a route, and I maybe do 2 or 3 runs where I can get a relatively decent time. If I were just to repeat attempts over and over (which I've done), I get frustrated easily and it ends up being counter-productive. There are runners that have the opposite approach that I do, where repeating attempts is more beneficial to them than planning.

Altogether though, I'd definitely be interested in this guide. I'd offer to help with compiling it, but I don't think I'm terribly qualified. Good luck!

Omnigamer

Edit history:

Omnigamer: 2013-11-21 11:12:56 am

2013-11-21 11:11:09 am

All the things

That point is definitely something I want to drive home throughout the resource: there's no "absolute" way to go through the process. People can experiment and find where their own strengths and desires are, and leave it at that. To that extent though, there should be something that supports those goals, which is why the individual sections need to be strong on their own. This is also something I want to address in the philosophy section: a runner should be able to perform runs in the manner that most befits their own goals. We do this for fun, after all.

There is a divide for some things though, particularly between popular games and obscure ones. Especially obscure games may only have one person to actually try to run it, so it's important that the runner is familiar with what things to try out and consider on the way to a finished run. The popular games may already have plenty of information about the route etc, but many people are largely unfamiliar with investigating ways to further improve it. The end product should be able to satisfy both varieties.

That point is definitely something I want to drive home throughout the resource: there's no "absolute" way to go through the process. People can experiment and find where their own strengths and desires are, and leave it at that. To that extent though, there should be something that supports those goals, which is why the individual sections need to be strong on their own. This is also something I want to address in the philosophy section: a runner should be able to perform runs in the manner that most befits their own goals. We do this for fun, after all.

There is a divide for some things though, particularly between popular games and obscure ones. Especially obscure games may only have one person to actually try to run it, so it's important that the runner is familiar with what things to try out and consider on the way to a finished run. The popular games may already have plenty of information about the route etc, but many people are largely unfamiliar with investigating ways to further improve it. The end product should be able to satisfy both varieties.

flicky

2013-11-21 01:07:07 pm

$15 per rant/allegory

I have so much stuff for this omni.

I'm serious, a lot of stuff.

I'm happy to compare notes.

Naegleria

(user is banned)

2013-11-21 01:31:57 pm

+1

Warepire

Edit history:

Warepire: 2013-11-21 02:31:07 pm

2013-11-21 02:29:23 pm

Heavy Metal Powered

I thought I'd share some of my experience running No One Lives Forever on PC as it is one of those less talked about games and as it was my first project, so it might have some beginner views in it as well. Perhaps there are some bits from it that can be useful when talking about different aspects while building the guide.

The most important thing about speedrunning games cannot be repeated enough, it has to be a game you can consider playing over and over and over..... Because you will play through the game MANY times during the various stages of producing the run. And in case of segmented runs, you will play through the same section of the game MANY times.

Running this game became a project that I worked on on-and-off for the process of 5 years. Speedruns should not be 5 year projects, but I am a little slow, and the interest in the game was quite cold, so I had to do a lot of the work alone, the project would still not be done if it wasn't for ymh coming along, which afterwards turned my solo project into a shared effort, no matter how obscure the game is, finding just one other person interested in the game helps IMMENSELY, when you have someone to help you test things and discuss ideas with it can open up new ways of thinking about the game in terms of inner mechanics. It also helps a lot when making the route as you may see things the other person missed, and vice versa.

I found it important to learn the game mechanics beyond casual play, this means things like "What happens if I jump up against this slope?" or "Can that slope/edge catch my fall and prevent me from dying if I'd jump off this ledge to reach the bottom of this area?" etc. The idea is to get a feel for how the mechanics in the game which aren't used a lot in casual play work. Knowing how they work can open up new interesting possibilities.

At this point I was also simultaneously trying to complete the game 100% (which means collecting all the intelligence items), just to give me breaks from the testing or to see if it would let me encounter situations I hadn't thought about. Thanks to an intelligence item being unobtainable and I googled around for help, lead me to a forum called The Glue Factory (it doesn't exist anymore, but fragments are preserved in the way back machine of Archive.org). Reading this forum looking for help, lead me to discover that things being discussed in game tech support can be very useful reads if you are speedrunning a game. I learned things that helped me find a place to start when looking for glitches later on.
Today this should also be extended to looking for youtube videos (specially videos tagged glitch, but also let's plays) or checking Twitch for people playing the game.

Cheat codes are your best friend! Yes, I said it. You should obviously not use them when doing attempts, but when exploring the game the help they provide is insane. They can let you focus on exactly the thing you want to find out and not care about enemies, timers, or preventing a lot of wasted time restarting because you think "Oh, I should test with this too!" before you realize "BLAST! I don't have that, now I need to start over so I'll have it."

Some glitches were already discovered in this game before I fully started running it. The ones that hadn't been discovered yet ymh and I discovered (often by accident, even though I had some ideas of possible weaknesses in the game mechanics) when playing through each mission over and over testing things, and discussing it, then testing new things in order to find the most optimal way to go about the objectives. With every new thing discovered we went back and checked if our new knowledge could lead to improvements in every mission. A few things that should apply to many games are: If the game has explosions, check if they have an effect on your speed if you stand "too close". If the game has quick saves, check their integrity by quick saving at probably unexpected moments (for example when falling to your death or in the middle of a mandatory conversation). Check how the game handles you getting stuck between various objects.

When making routes, it's a good idea to make a test-run of the route, just to see if it is even remotely feasible, these tests can be really sloppy, as long as they provide the necessary info. You will notice if you need more of something than you can ever pick up, or if you cannot possibly make it through with enough health for the next section etc.

The final thing I had to come to terms was: Decide when you are done exploring and begin making serious attempts. It is of course important to learn enough about the game before making attempts at the final product, but you can be stuck in the exploring stage for a long time otherwise, and the pay out will not be worth the time spent achieving it. It's OK to have possible improvements if you discover something too late while making a segmented run for example.

I thought I'd share some of my experience running No One Lives Forever on PC as it is one of those less talked about games and as it was my first project, so it might have some beginner views in it as well. Perhaps there are some bits from it that can be useful when talking about different aspects while building the guide.

The most important thing about speedrunning games cannot be repeated enough, it has to be a game you can consider playing over and over and over..... Because you will play through the game MANY times during the various stages of producing the run. And in case of segmented runs, you will play through the same section of the game MANY times.

Running this game became a project that I worked on on-and-off for the process of 5 years. Speedruns should not be 5 year projects, but I am a little slow, and the interest in the game was quite cold, so I had to do a lot of the work alone, the project would still not be done if it wasn't for ymh coming along, which afterwards turned my solo project into a shared effort, no matter how obscure the game is, finding just one other person interested in the game helps IMMENSELY, when you have someone to help you test things and discuss ideas with it can open up new ways of thinking about the game in terms of inner mechanics. It also helps a lot when making the route as you may see things the other person missed, and vice versa.

I found it important to learn the game mechanics beyond casual play, this means things like "What happens if I jump up against this slope?" or "Can that slope/edge catch my fall and prevent me from dying if I'd jump off this ledge to reach the bottom of this area?" etc. The idea is to get a feel for how the mechanics in the game which aren't used a lot in casual play work. Knowing how they work can open up new interesting possibilities.

At this point I was also simultaneously trying to complete the game 100% (which means collecting all the intelligence items), just to give me breaks from the testing or to see if it would let me encounter situations I hadn't thought about. Thanks to an intelligence item being unobtainable and I googled around for help, lead me to a forum called The Glue Factory (it doesn't exist anymore, but fragments are preserved in the way back machine of Archive.org). Reading this forum looking for help, lead me to discover that things being discussed in game tech support can be very useful reads if you are speedrunning a game. I learned things that helped me find a place to start when looking for glitches later on.
Today this should also be extended to looking for youtube videos (specially videos tagged glitch, but also let's plays) or checking Twitch for people playing the game.

Cheat codes are your best friend! Yes, I said it. You should obviously not use them when doing attempts, but when exploring the game the help they provide is insane. They can let you focus on exactly the thing you want to find out and not care about enemies, timers, or preventing a lot of wasted time restarting because you think "Oh, I should test with this too!" before you realize "BLAST! I don't have that, now I need to start over so I'll have it."

Some glitches were already discovered in this game before I fully started running it. The ones that hadn't been discovered yet ymh and I discovered (often by accident, even though I had some ideas of possible weaknesses in the game mechanics) when playing through each mission over and over testing things, and discussing it, then testing new things in order to find the most optimal way to go about the objectives. With every new thing discovered we went back and checked if our new knowledge could lead to improvements in every mission. A few things that should apply to many games are: If the game has explosions, check if they have an effect on your speed if you stand "too close". If the game has quick saves, check their integrity by quick saving at probably unexpected moments (for example when falling to your death or in the middle of a mandatory conversation). Check how the game handles you getting stuck between various objects.

When making routes, it's a good idea to make a test-run of the route, just to see if it is even remotely feasible, these tests can be really sloppy, as long as they provide the necessary info. You will notice if you need more of something than you can ever pick up, or if you cannot possibly make it through with enough health for the next section etc.

The final thing I had to come to terms was: Decide when you are done exploring and begin making serious attempts. It is of course important to learn enough about the game before making attempts at the final product, but you can be stuck in the exploring stage for a long time otherwise, and the pay out will not be worth the time spent achieving it. It's OK to have possible improvements if you discover something too late while making a segmented run for example.

Patashu

2013-11-21 03:21:30 pm

Any good speedrunning guide should reference the neurologically correct guide to practice topic that was posted here a few months back Smiley

johannhowitzer

2013-11-21 04:12:10 pm

Yes, there should definitely be some coverage of the right and wrong ways to practice.

There's definitely a blend of "big picture" thinking and "detail" thinking in speedrunning. You need to be able to do both well, and many people (perhaps most) are weak in one of these. I personally am a detail person and have trouble backing up and seeing the forest as a whole. This is where collaboration comes in. Also, some games lean more toward details or more toward big picture. Very non-linear games that are low-execution and high-planning favor the big picture, while more linear, execution-based games can be very detail-favoring. Of course this happens on many levels, and branching paths and decisions can still pop up under the surface of a game that appears linear to the casual player. Now why on earth do I favor high-planning games so much when I'm a detail guy? GUH.

This sounds like a mammoth undertaking and something that could really add to the community. I'll be watching this when I remember to, for sure.

Yes, there should definitely be some coverage of the right and wrong ways to practice.

There's definitely a blend of "big picture" thinking and "detail" thinking in speedrunning.  You need to be able to do both well, and many people (perhaps most) are weak in one of these.  I personally am a detail person and have trouble backing up and seeing the forest as a whole.  This is where collaboration comes in.  Also, some games lean more toward details or more toward big picture.  Very non-linear games that are low-execution and high-planning favor the big picture, while more linear, execution-based games can be very detail-favoring.  Of course this happens on many levels, and branching paths and decisions can still pop up under the surface of a game that appears linear to the casual player.  Now why on earth do I favor high-planning games so much when I'm a detail guy?  GUH.

This sounds like a mammoth undertaking and something that could really add to the community.  I'll be watching this when I remember to, for sure.

johannhowitzer

2013-11-21 04:12:55 pm

Viking it up

Woops, cookies. ^ Yeah, that was me.

mike89

2013-11-21 06:10:12 pm

SEGA Junkie

Following this with interest. I could probably contribute some things if you're interested, too.

kirbymastah

2013-11-21 07:38:43 pm

<(^_^)>

^ Same.

I can provide some insight on timing differences, and provide examples for certain games and the pros/cons for RTA vs IGT (i.e. metroid fusion/zm).

I've also been learning many new games lately, and I can contribute with how I personally learn a game and why it works for me.

Patashu

2013-11-21 07:55:09 pm

If there's real time vs in game time stuff, I think there should also be frame rate differences / platform differences / lag differences between console, official emulators and PC emulators, etc. Basically 'how do I compare runs of the same game on two different platforms/emulators, which is the real better time'. Maybe better as an appendix though because the actual speedrunning doesn't need to care about this Smiley

And yeah, it's good to emphasize that everyone learns in different ways. Some people learn better having notes to refer to, others learn better internalizing/making mnemonics/purely memorizing.

Omnigamer

2013-11-21 10:15:22 pm

All the things

I don't think I'll write the section to completely flesh out the controversy of RTA vs IGT. I mainly just want to give the topic exposure so people know that it is an issue and there are reasons for and against each side. Bringing up an entire debate is not the intention (there are threads exactly for this), but it's probably worthwhile to include an appendix of notable games with unique timing and some reasoning for it.

The Sid

2013-11-22 03:50:56 pm

raising the stakes on being bad!

I'll try to add some stuff later when I have a bit more time, but this topic really piques my interest.
as someone newer to speed running ( I do actually run a few games not well but i run em') I'm not sure what I have to add, but I think a some what of an outside perspective might be useful in some way.
I think game choice and discovering which games you might enjoy to run might be some thing I wanna talk about :v

zewing

2013-11-22 04:15:26 pm

just( •_•)>⌐■-■ ..... (⌐■_■)wing it

I like this idea a lot and I was intending to put something like this on my twitch page eventually. I've been lucky enough to route a game and being the first runner for it and I went through a process that was simple to me but when you put it into a format is quite difficult. One thing to mention is community decides timing. Spyro uses SDA timing, Pokemon using IGT, almost everything else uses RTA (standard "first input to loss of character control")

BaronHaynes

Edit history:

BaronHaynes: 2013-11-23 11:20:32 pm

2013-11-23 11:08:54 pm

Strange days, incredible days

I also really like this idea, I enjoy reading about the process other runners go through and comparing my experience to that.

The outline you've put together focuses on process and methods -- I'd suggest also including a brief section on how temperament and attitude affects speedrunning, when it's a good idea to take a break or adjust your thinking, how to keep from getting discouraged, how to keep it from feeling like a grind, discuss healthy vs. less healthy reasons for running a game (because you enjoy it vs. wanting fame/attention etc.).

The most common question I see in stream chats is "I'm interested in trying this, what game should I run?" And the simple answer is always "pick something you love to play", but everyone follows up with more detailed advice like "pick something you don't mind seeing the beginning of a lot", "think about how much RNG or downtime it has", "make sure it's not something you'll ruin for yourself by overplaying it", "make sure you're okay with competition if it's a popular game, or okay with not much feedback if it's not". A 'Picking a Game' section could explore a lot of this stuff while also being a good intro to someone who's curious but might be overwhelmed if it's too technical right out of the gate.

I also really like this idea, I enjoy reading about the process other runners go through and comparing my experience to that.

The outline you've put together focuses on process and methods -- I'd suggest also including a brief section on how temperament and attitude affects speedrunning, when it's a good idea to take a break or adjust your thinking, how to keep from getting discouraged, how to keep it from feeling like a grind, discuss healthy vs. less healthy reasons for running a game (because you enjoy it vs. wanting fame/attention etc.).

The most common question I see in stream chats is "I'm interested in trying this, what game should I run?" And the simple answer is always "pick something you love to play", but everyone follows up with more detailed advice like "pick something you don't mind seeing the beginning of a lot", "think about how much RNG or downtime it has", "make sure it's not something you'll ruin for yourself by overplaying it", "make sure you're okay with competition if it's a popular game, or okay with not much feedback if it's not". A 'Picking a Game' section could explore a lot of this stuff while also being a good intro to someone who's curious but might be overwhelmed if it's too technical right out of the gate.

Omnigamer

2013-11-25 11:36:50 am

All the things

I can definitely agree with your second point; I will include it as the initial subsection to the Investigation phase.

On the first point, it's harder to turn into its own section given the rest of the structure. I can say that all of those things will be taken into consideration while writing out pieces, but they will be sprinkled into the most relevant sections rather than having a single section for all of them.

I also think I will include a new subsection in Routing to deal with Pathing, or determining the best small steps to take when proceeding from one goal to the next. This would cover strategies for optimization of tasks rather than an entire goal. For example, "beat the boss" is actually a process of hit 1 -> hit 2 -> hit 3 -> dead. Figuring out what to do for each task of the goal is a bit more specific and tricky than general routing, although portions of the process are similar. Organizing the section will be tricky though since I would probably have to incorporate specific examples.

Finally, something I didn't mention in the initial post is that I want to include "Case Studies" of sorts at the end of each technical section. The goal would be to have several clips from a variety of games that feature a mix of optimized and non-optimized play. Some questions and a few explanations will accompany the videos, but the point is to have the readers apply some critical thinking to what they're watching and think how they might improve the clip given what they had read in the section. I would definitely be interested in hearing about some examples that could support this.

I can definitely agree with your second point; I will include it as the initial subsection to the Investigation phase.

On the first point, it's harder to turn into its own section given the rest of the structure. I can say that all of those things will be taken into consideration while writing out pieces, but they will be sprinkled into the most relevant sections rather than having a single section for all of them.

I also think I will include a new subsection in Routing to deal with Pathing, or determining the best small steps to take when proceeding from one goal to the next. This would cover strategies for optimization of tasks rather than an entire goal. For example, "beat the boss" is actually a process of hit 1 -> hit 2 -> hit 3 -> dead. Figuring out what to do for each task of the goal is a bit more specific and tricky than general routing, although portions of the process are similar. Organizing the section will be tricky though since I would probably have to incorporate specific examples.

Finally, something I didn't mention in the initial post is that I want to include "Case Studies" of sorts at the end of each technical section. The goal would be to have several clips from a variety of games that feature a mix of optimized and non-optimized play. Some questions and a few explanations will accompany the videos, but the point is to have the readers apply some critical thinking to what they're watching and think how they might improve the clip given what they had read in the section. I would definitely be interested in hearing about some examples that could support this.

BaronHaynes

2013-11-25 03:11:56 pm

Strange days, incredible days

Quote from Omnigamer:

On the first point, it's harder to turn into its own section given the rest of the structure. I can say that all of those things will be taken into consideration while writing out pieces, but they will be sprinkled into the most relevant sections rather than having a single section for all of them.

That's a good idea. I think at some point every runner loses perspective on why they enjoy it as a hobby and gets caught up in the frustration/competitive aspects; the guide can reflect this by just acknowledging that speedrunning is supposed to be fun (and if it's not fun, you can simply change how you're doing it).

Quote from Omnigamer:

Finally, something I didn't mention in the initial post is that I want to include "Case Studies" of sorts at the end of each technical section. The goal would be to have several clips from a variety of games that feature a mix of optimized and non-optimized play. Some questions and a few explanations will accompany the videos, but the point is to have the readers apply some critical thinking to what they're watching and think how they might improve the clip given what they had read in the section. I would definitely be interested in hearing about some examples that could support this.

Alucard any% in Castlevania: Symphony of the Night is a great example of how much a game can evolve over time. It'd be a good example of broad changes on the macro level.

Micro level stuff like demonstrating a series of optimizations in a particular trick has a lot of possibilities. It might be a good idea to show how pathfinding and routing are important no matter the genre of game, and have a few examples of different ones, e.g. platformer, shooter, RPG, puzzle game.

[quote=Omnigamer]On the first point, it's harder to turn into its own section given the rest of the structure. I can say that all of those things will be taken into consideration while writing out pieces, but they will be sprinkled into the most relevant sections rather than having a single section for all of them.[/quote]That's a good idea. I think at some point every runner loses perspective on why they enjoy it as a hobby and gets caught up in the frustration/competitive aspects; the guide can reflect this by just acknowledging that speedrunning is supposed to be fun (and if it's not fun, you can simply change how you're doing it).

[quote=Omnigamer]Finally, something I didn't mention in the initial post is that I want to include "Case Studies" of sorts at the end of each technical section. The goal would be to have several clips from a variety of games that feature a mix of optimized and non-optimized play. Some questions and a few explanations will accompany the videos, but the point is to have the readers apply some critical thinking to what they're watching and think how they might improve the clip given what they had read in the section. I would definitely be interested in hearing about some examples that could support this.[/quote]Alucard any% in Castlevania: Symphony of the Night is a great example of how much a game can evolve over time. It'd be a good example of broad changes on the macro level.

Micro level stuff like demonstrating a series of optimizations in a particular trick has a lot of possibilities. It might be a good idea to show how pathfinding and routing are important no matter the genre of game, and have a few examples of different ones, e.g. platformer, shooter, RPG, puzzle game.

Omnigamer

2013-11-27 01:45:27 pm

All the things

I plan to start writing one or a couple of the technical sections over Thanksgiving weekend. Not sure which one yet. I will put preliminary drafts on a blog or something for people to review. In the meantime, thank you everybody for the suggestions and keep it up. If anybody is interested in editing or contributing a section/subsection, post here or give me a PM.

Yagamoth

Edit history:

Yagamoth: 2013-12-02 02:43:22 am

2013-12-02 02:41:17 am

INTJ

I suggest another section somewhere: Getting started
- Do a casual run with the simplest speedrun route (If there is no speedrun route, make a rough outline of it)
- Huh?

To explain this: For me personally, getting started was the hardest portion when I wanted to run Secret of Mana. It took me literally 4-5 years until I actually did my first run. I'm not sure how special my case is though. The reason why it took me so long to get started is, because there was so much information, research and so many options available that were necessary to take into consideration, I didn't have the courage/motivation and didn't know where to start. Even though there was already a good run (in hindsight: an extremely good run), I knew there was still so much left to explore, try, test and especially optimize. The damage formulas were known, all stats, basically everything you need to route the game into detail. Secret of Mana was a monster (project).

In the end, the event that got me started was another player doing a considerably more simple run using a non-SDA-conform 1 player 2 controller method. I found another skip in cooperation with him, suddenly the run was very doable. Tried my first casual playthrough with the speedrun route, and suddenly the monster was not as bad as I thought it would be.

I suggest another section somewhere: Getting started
- Do a casual run with the simplest speedrun route (If there is no speedrun route, make a rough outline of it)
- ???

To explain this: For me personally, getting started was the hardest portion when I wanted to run Secret of Mana. It took me literally 4-5 years until I actually did my first run. I'm not sure how special my case is though. The reason why it took me so long to get started is, because there was so much information, research and so many options available that were necessary to take into consideration, I didn't have the courage/motivation and didn't know where to start. Even though there was already a good run (in hindsight: an extremely good run), I knew there was still so much left to explore, try, test and especially optimize. The damage formulas were known, all stats, basically everything you need to route the game into detail. Secret of Mana was a monster (project).

In the end, the event that got me started was another player doing a considerably more simple run using a non-SDA-conform 1 player 2 controller method. I found another skip in cooperation with him, suddenly the run was very doable. Tried my first casual playthrough with the speedrun route, and suddenly the monster was not as bad as I thought it would be.

Deuceler

2013-12-02 01:39:49 pm

SPEEDruns not SAFEruns

I think this is a really cool idea, but at the same time I'm conflicted about it. I think the reason I love speed running and the community so much is because it's something I discovered through curiosity and exploration - much the same way I've routed/ran games. It's like I stumbled upon it on accident, although it's always been right here, and learning how to stream/improve my stream was/will always a really fun process, and extremely rewarding when I felt like I got it right. I may be alone in this, but if someone was just handed all the tools and shown exactly what to do they may not enjoy/appreciate it as much as most of us do. It becomes more about viewers/twitchbux then it does actually running games because you love them more then your own family, and can't wait to shave off a couple seconds.

I'm not saying in anyway you shouldn't put something together, I've often thought about putting up Xsplit tutorials and cap card tutorials to help those who are curious - but if someone isn't willing to put in the effort to learn how a simple set up works, will they be willing to die on the same level over and over? Maybe I'm just being pessimistic now haha

Anyway, I say go for it - but you can see where my conflict arises from

I think this is a really cool idea, but at the same time I'm conflicted about it. I think the reason I love speed running and the community so much is because it's something I discovered through curiosity and exploration - much the same way I've routed/ran games. It's like I stumbled upon it on accident, although it's always been right here, and learning how to stream/improve my stream was/will always a really fun process, and extremely rewarding when I felt like I got it right.  I may be alone in this, but if someone was just handed all the tools and shown exactly what to do they may not enjoy/appreciate it as much as most of us do. It becomes more about viewers/twitchbux then it does actually running games because you love them more then your own family, and can't wait to shave off a couple seconds.

I'm not saying in anyway you shouldn't put something together, I've often thought about putting up Xsplit tutorials and cap card tutorials to help those who are curious - but if someone isn't willing to put in the effort to learn how a simple set up works, will they be willing to die on the same level over and over? Maybe I'm just being pessimistic now haha

Anyway, I say go for it - but you can see where my conflict arises from

Hsanrb

2013-12-02 05:06:24 pm

If theres one point I'd include, its the "When to reset vs Keep going?" I think too many speedrunners' make the reset button their new best friend, when their PB is too far off from the time they are aiming for to actually make it work. The journey to the finished product (whether your submitting to SDA, or just to your own youtube channel) is best enjoyed over the long haul then 2 or 3 runs. I usually use the 1-2% rule, in that if your PB is greater than 102% for games ~1hr of your goal (WR or a time you would be satisfied with) that its not worth reseting. The % can change based on length like 10m might be 5%, 10hr games might be "Just don't after 30% completed, you've already invested the time where reseting means you are unable to finish"

It is possible to burn out because the last time you saw the credits, was September because you overeset. If your running MMX and have like a 3hr stream block, I'd do a warmup run, play a bunch of attempts, then last attempt is finish regardless what goes wrong. Short on time, just turn it on and finish. Sometimes the credits is a reward best worth rejoicing over, not just when the final time goes down

I have echoed Deuceler's viewers in other threads on this forum so I do agree somewhat. It is possible when people look up a game that there is too much information out there as to what works, and what is done. If your planning to learn Super Metroid, I WOULD NEVER EVER START WITH THE CURRENT ROUTE ITERATION! After AGDQ13, I saw so many people jump on Super Metroid, and they gave up because the route they saw is not a beginner, intermediate, and even advanced speedrunning guide... it was like Professional quality I wouldn't even touch and I've done a few hundred runs. Essentially just because information/runs are available doesn't mean you should leap head first into the massive database of your game. You'll just end up swimming with the current, possibly getting washed away and turned off from speedrunning the game altogether, rather than having background knowledge of when the current is taking you the wrong way to the fastest way.

If theres one point I'd include, its the "When to reset vs Keep going?" I think too many speedrunners' make the reset button their new best friend, when their PB is too far off from the time they are aiming for to actually make it work. The journey to the finished product (whether your submitting to SDA, or just to your own youtube channel) is best enjoyed over the long haul then 2 or 3 runs. I usually use the 1-2% rule, in that if your PB is greater than 102% for games ~1hr of your goal (WR or a time you would be satisfied with) that its not worth reseting. The % can change based on length like 10m might be 5%, 10hr games might be "Just don't after 30% completed, you've already invested the time where reseting means you are unable to finish"

It is possible to burn out because the last time you saw the credits, was September because you overeset. If your running MMX and have like a 3hr stream block, I'd do a warmup run, play a bunch of attempts, then last attempt is finish regardless what goes wrong. Short on time, just turn it on and finish. Sometimes the credits is a reward best worth rejoicing over, not just when the final time goes down

I have echoed Deuceler's viewers in other threads on this forum so I do agree somewhat. It is possible when people look up a game that there is too much information out there as to what works, and what is done. If your planning to learn Super Metroid, I WOULD NEVER EVER START WITH THE CURRENT ROUTE ITERATION!  After AGDQ13, I saw so many people jump on Super Metroid, and they gave up because the route they saw is not a beginner, intermediate, and even advanced speedrunning guide... it was like Professional quality I wouldn't even touch and I've done a few hundred runs. Essentially just because information/runs are available doesn't mean you should leap head first into the massive database of your game. You'll just end up swimming with the current, possibly getting washed away and turned off from speedrunning the game altogether, rather than having background knowledge of when the current is taking you the wrong way to the fastest way.

Patashu

2013-12-02 07:10:29 pm

Quote from Deuceler:

I may be alone in this, but if someone was just handed all the tools and shown exactly what to do they may not enjoy/appreciate it as much as most of us do. It becomes more about viewers/twitchbux then it does actually running games because you love them more then your own family, and can't wait to shave off a couple seconds.

I don't think this will be a problem. People who are or can make themselves good at speedrunning won't be worse at speedrunning due to the existence of a guide telling them how to get started. People who are bad or will remain bad will quickly self-select themselves out of streaming (or find something else that makes them worth watching - and I don't see any problem with that!).
It's basically an equivalent argument to 'There shouldn't be any tutorials for Photoshop, if you really wanted to learn how to manipulate images like an expert then you wouldn't need your hand held to do it'

I have no name

2013-12-02 08:28:22 pm

Obscure games ftw

The few pieces of advice I give new speedrunners:
1. If you feel up for it, route the game yourself. You might find something that everyone else overlooked. Additionally, the skills you develop while doing so will help when it comes time for optimizing, and at least for me it's easier to remember routes I came up with (so even if you match the route that already exists- you already know it!)
2. Don't worry about if there's already a near perfect run: run for fun! It's not all about the record, it's about improving your own time. I got discouraged from running Ape Escape a while ago, and now I've had a few runs on WR pace (which, sadly, the RNG ruined. Which leads me to my next point...)
3. Don't worry about RNG. It can be very tempting to keep resetting until you get flawless luck-but it's useless if you screw up after from lack of practice. This ties into how often you should reset-only if it's impossible to PB, do some resetless runs/races too, to get valuable endgame practice.
4. Don't be afraid to ask anyone how help on how to do something, or WHY something is done. Ask yourself WHY you do the things you do-a lot of little things can be found that way.
5. If it stops being fun, stop running that game for a while. It might be more enjoyable after a break, and if you ran other games in that time you'll probably come back with a really good run with general improvements. If you want to keep doing runs, pick up another game.
6. Don't run games just because they're popular or well established, and don't run obscure games just for a free WR. Run the games you want to run because you want to run them, not because other people do or for a record.
7. When possible, drill individual levels. This is the most time efficient way to get better at the game. I got a very good run of Ape Escape, then did all the time attacks a bunch and when I went back to any%, I got a minute ahead of my splits- by saving some time I didn't even know I had, just by optimizing movement further: since the time attacks REQUIRED this level of skill and confidence in movement.

Looking at the suggested sections, for consistency improvements the way I do it is I get a feel for how frequently I get the strat and how much time it saves over a easy method that I get 100% of the time. So if something takes 10 seconds the slow way, 5 seconds the fast way and 13 if you fail the fast way, and I get it 75% of the time, then if 5*0.75+13*0.25<10 it will, on average, save you time. In that example 3.75+3.25=7, so on average it would save 3 seconds: well worth attempting in a run.

The few pieces of advice I give new speedrunners:
1. If you feel up for it, route the game yourself.  You might find something that everyone else overlooked.  Additionally, the skills you develop while doing so will help when it comes time for optimizing, and at least for me it's easier to remember routes I came up with (so even if you match the route that already exists- you already know it!)
2. Don't worry about if there's already a near perfect run: run for fun!  It's not all about the record, it's about improving your own time.  I got discouraged from running Ape Escape a while ago, and now I've had a few runs on WR pace (which, sadly, the RNG ruined.  Which leads me to my next point...)
3. Don't worry about RNG.  It can be very tempting to keep resetting until you get flawless luck-but it's useless if you screw up after from lack of practice.  This ties into how often you should reset-only if it's impossible to PB, do some resetless runs/races too, to get valuable endgame practice.
4. Don't be afraid to ask anyone how help on how to do something, or WHY something is done.  Ask yourself WHY you do the things you do-a lot of little things can be found that way.
5. If it stops being fun, stop running that game for a while.  It might be more enjoyable after a break, and if you ran other games in that time you'll probably come back with a really good run with general improvements.  If you want to keep doing runs, pick up another game.
6. Don't run games just because they're popular or well established, and don't run obscure games just for a free WR.  Run the games you want to run because you want to run them, not because other people do or for a record.
7. When possible, drill individual levels.  This is the most time efficient way to get better at the game.  I got a very good run of Ape Escape, then did all the time attacks a bunch and when I went back to any%, I got a minute ahead of my splits- by saving some time I didn't even know I had, just by optimizing movement further: since the time attacks REQUIRED this level of skill and confidence in movement.

Looking at the suggested sections, for consistency improvements the way I do it is I get a feel for how frequently I get the strat and how much time it saves over a easy method that I get 100% of the time.  So if something takes 10 seconds the slow way, 5 seconds the fast way and 13 if you fail the fast way, and I get it 75% of the time, then if 5*0.75+13*0.25<10 it will, on average, save you time.  In that example 3.75+3.25=7, so on average it would save 3 seconds: well worth attempting in a run.

Omnigamer

2015-02-18 08:47:42 am

All the things

Reviving this a bit since I picked up some renewed interest in this project recently, possibly to prepare it for another medium. One of the sections eventually did get churned out on the KB, so you can find it here: What is a speedrun?

Basically, I'm still looking for any input and feedback as I restart this process. My mental outline of it has changed a bit, and I'll try to edit it into the first post as I formalize it. Some new sections I'd like to include cover various lingo used within the community, timing methods and other controversies, and a "Tools of the Trade" section about various programs, websites, communities, and otherwise for getting started.