What makes a fair map for speedrunning?

Sometimes when you're working on a speedrun it seems that everything depends only on you, so when you fail, you know that it was your fault and you can say where exactly you screwed up. Other times you feel just like a robot doing exactly the same thing over and over again, and just waiting for luck. Can you give me any tips on how to make my maps closer to the former? Talk about a map that you thought was really fair.

Or is it impossible to make a fair map? I mean, you always can try playing even more aggressively than before, taking more risks, etc. Thoughts?

I never thought of having a perspective of a 'fair map for speedrunning' as I feel practicing the map and knowings its faults and errors all comes down to you in the end.

I don't usually seriously speedrun, let alone do so for the purpose of recording my attempts, but I do try to make almost all of my maps with an eye for speedrunners.

I think anything that puts the speedrunner at the mercy of Doom's RNG is unfair. So if the player has a pistol and there is a Sarge blocking a door, there should be a way to juke around him so that the player doesn't have to pray that two pistol shots will be enough to get by the Sarge before the close the distance.

Not in regards to fairness, but I like to make some of the lower wall details impassable (like wall sections between jutting girders) so that the player doesn't get needlessly snagged on the scenery.

Any given map is bound to require luck at some point if you're aiming for the fastest possible time. The only way you can escape that robotic repetition of the same actions is by consciously sacrificing speed for reliability now and then. That's nothing to be ashamed of, especially in longer runs and table-fillers. If someone wants to beat your record then at least they'll have their work cut out for them.

I logged onto Doomworld.com today with a related question, which is why I'm hijacking this thread: would you perform a west-to-east guideless gap glide to save a maximum of fifty seconds, five minutes into a difficult ten-minute run? This isn't hypothetical, by the way.

Luck is a factor in every speedrun, and more luck doesn't equal less skill, as part of the skill factor involves dealing with the luck factor.

That being said, I still understand what you are trying to say, but it is extremely objective considering my above statement. This will probably turn from users listing "fair" maps to users listing maps they like to speedrun. It may be a better to ask "what makes a good speedrunning map" or "what maps did you speadrun that you felt were unfair."

As to what makes a good speedrunning map, again extememly subjective but I really like AV05, M205, and HR06. They are all small levels (thus no frustration from dying 5 minutes into a likely 10 minute run), relatively easy to complete when not speedrunning, but very challenging to optimize in any category (this adds replayability), and there's no monsters that just require you to boringly shoot at it repetitivly with weak weapon over and over again till it dies (e.g. chain-gun vs spidermastermind).

Flawless teleporters and not letting monsters fall off the playable area is a given.

Now for highly subjective stuff:
- diagonal and or multiple W activation linedefs (basically making it hard enough so breaking them is extremely unlikely outside of TAS)
- making sure no gap is 32+ pixels wide
- no AV/SR40/SR50/wallrun/thingrun jumps required to progress normally through the level or reach tagged secrets
- minimal sequence breaks if any

Definitely personal preference, people who love breaking maps with tricks might disagree. I feel if you make using a particular trick at a particular spot the unequivocally best option, you remove every other option and the run becomes about trite repetition of that one trick over and over until you get it right or get lucky.

In maps where HUD use is unintended, making sure monsters can't wander off and get stuck in awkward places could help.

Alleviating the random factor as much as you can decreases the need for luck. It's better to fight eight imps with 200 hp than four imps with 100 hp, because their damage is more likely to be close to the average. The super shotgun, rocket launcher and BFG have less damage variation than the shotgun, chaingun, pistol and fist, and so tend to be more reliable weapons.

Fighting bigger amounts of monsters makes their movement more predictable. Past a certain number, a Doom horde is more like a wave than a bunch of individual critters.

As Archy mentioned, quick levels and no timesinks certainly make sense.

I'd also go with no ridiculous spike in difficulty, there's little more frustrating than to be able to run through a level at top speed with eyes closed save for that one part where you've got a 10% chance of survival.

Phml said:

The super shotgun, rocket launcher and BFG have less damage variation than the shotgun, chaingun, pistol and fist, and so tend to be more reliable weapons.

I don't think chaingun has that much of variation. It shoots many bullets which means it quickly deals the average damage. BFG direct hit deals 100-800 or was it 200-800 so it can be annoying sometimes. Rocket launcher is fast shooting weapon so there's not much variation in big fights, but when killing few single demons it's just luck fest.
Berserk fist is just ludicrous with the damage variation :D

You do have a point. I just want to clarify I wasn't arguing the chaingun had much variation, just that it had more variation. I'm probably wrong regardless.

While I disagree with Phml about limiting tricks, the rest of his advice is sound.

Designs that prevent lost monsters, emphasis on SSG/BFG/RL use for max runs, no jumping/movement puzzles or inescapable pits of death. No difficulty spikes.

Properly designed monster teleporters FFS, it's 2013, people. At least 2 destinations and no long horizontal/vertical trigger lines.

Thank you all for your thoughts, I especially liked this idea, simple but makes perfect sense:

Phml said:

It's better to fight eight imps with 200 hp than four imps with 100 hp, because their damage is more likely to be close to the average.

Belial said:

Properly designed monster teleporters FFS, it's 2013, people. At least 2 destinations and no long horizontal/vertical trigger lines.

Wait... I get the multiple destinations, but what is the preferred spawn room setup? I'm still using the box with a fake door and teleport line at the end setup.

I'm guessing the long trigger line is where you have multiple monsters in one closet. The wide trigger line allows more than one monster to cross at the same time, in which case only one will instantly warp wheras the other will mill around in the closet until it re-crosses the trigger line.

As far as I'm aware the smaller the closet, fewer the monsters, narrower the trigger line and more numerous the teleport destinations the more efficient the teleport will be.

 ______________
|              \ \ \ \
|               \ \ \ \
|               / / / /
|______________/ / / /

Ribbiks said:

\ \ \ \
 \ \ \ \
 / / / /
/ / / /

for life

Are those the teleport lines? Looks very efficient!

just don't forget to point them towards the map, heh.

Those are still unreliable if too big. See sunder map 10 last encounter.

They could still be broken up into smaller, separate units, like purist mentioned. How big is too big, like over 256 wide?

I would say any width big enough to accomodate one of the monsters passing another would be an efficiency loss.