I think it is an interesting metric and worth measuring. But is it really a fair metric of difficulty especially if have to measure time to death in tens of seconds. There is a lot of opportunity for player action in that time. I guess I've been playing a lot of chess lately where the subject of difficulty comes up. There difficulty is also quite subjective, but it always considers sequences going forward. How easy are they to spot. How careful do you need to be to thread the needle. That does make it difficult to look at real numbers though. But some simple metrics that would consider the player's ability to act would be 1) total HP of enemies with hot sight on the player (basically how long in time it would take to shoot them down) 2) Time to safety. How much time does it take to get somewhere where safe. 3) DPS on the player multiplied by HP. Basically how much HP loss you would take if you attempted to shoot them down. 4) Time to 1 on 1 opportunity. Basically if you can get yourself safe by going into a closet but you can't get out to attack without exposing yourself to more than one enemy then the closet hide would satisfy metric 1 but not represent an easy situation. If you wanted to look at it really chess like what you would do is get a lot of players with different skill levels. You would give every player and every map an ELO score. And then by looking at the success rate of different players on different maps you could refine those scores. The more refined your player ELO scores are the more accurate you can make your map ELO scores and visa versa. If you want to organize something I'd play.

@ChaingunChaCha

25 күн бұрын

Thanks for the thorough comment! This is not in any way a fair way to measure difficulty of a map, or even a hot-start in general. This is more of a casual look at the maps we've played many times but with the oddity of taking no action(s) that would affect the game. So that means staying still and seeing how long it takes to take damage and also die. (Doom has RNG for most actions but a fixed RNG table so it replays the same every time.) It's all a bit silly but it was born out of my challenge of trying to complete levels without taking any damage and then thinking about doing what no player does - stand still at the start. Player actions would change the outcome totally by advancing through the RNG table, thus affecting enemy actions. It would be like asking you opponent in chess to checkmate you as quickly as possible without you having your turns. And then when you do the same again but this time take your turns the opponent's actions are affected. But analysing the start of every level has brought out some funny RNG moments like in E1M2 in the previous video and Map01 in this video. Anyway back to your main point - it's not fair, as in a real game the player would be taking actions and affecting the play. If we're talking measuring the difficulty of hot-starts (and potentially maps in general) then you bring up some great measurements. I'd also add: - Enemy types (along with positions, damage output and likelihood to fire) - Health, armour and other pickups available (along with distances from the start and their value) - Weapons and ammo (along with distances from the start and their value) - RNG table sequence at start of play - Player skill and style (and how cautious they allow themselves to be before it gets silly) - Geometry of the level - I'm sure there are plenty more variables that come into it Overall though in my view it would only give a vague, general view of the difficulty as a lot of difficulty in Doom comes from organic situations like how enemies path, damage taken by particular shots (every shot is RNG damage) and many more things. Playing differently for just a second or so will change everything beyond that second. There are so many RNG checks in a second of play that it is hard to replay a level section exactly the same way. Firing your pistol just once can alter whether an enemy will hit you with their shot and also what damage it does. Doom difficulty is an interesting idea but hard to quantify. Edit: Added comparison to chess.

@JasonMitchellofcompsci

25 күн бұрын

@@ChaingunChaCha Very interesting. I code social forums, and the idea of measuring doom difficulty almost has me tempted to code in some hidden utilities to allow users to submit their successes and failures and calculate those scores. Then you could measure. Of course you would have to find people interested in submitting that. The advantage might be for them that they could use it to find maps that hit their skill level exactly, 50% odds of success on a blind run. It really has me thinking about that if only to code it for fun. In fact to make the data valid, or at least consistent, you would have to only accept blind runs. But who knows. Maybe blind runs are fun if you are guaranteed a map that exactly hits your difficulty level. So you could also use it to get a random suggested map, make an attempt, and then report. You could then play random maps without getting instantly destroyed and also without having a map that is a snooze fest.

@ChaingunChaCha

24 күн бұрын

That seems like a better measure of difficulty compared to analysing the level properties. If you are focusing on the offical levels in Doom and Doom 2 then I think that the challenge would be to find people who have not yet played them. This idea would likely work better for all of the community made maps. The Doomworld forum also runs something interesting - a monthly "Ironman" competition where users play a community wad (set of maps) and see how far they can go without dying. And then they submit their score. It's all just for fun but that does in some way assess how difficult a set of maps is.