What Does Everyone Consider an "Official" Record?

As it seems few people play Vanilla Doom anymore, I was wondering what people consider an "official" record, assuming there is anything official?

I use GZDoom to play Doom, specifically with enemy count enabled in the map screen, as well as the timer and 'always run' enabled. I don't think any of this gives too much of an advantage, but it certainly helps for longer levels where knowing whether you are making good time at certain milestones helps in determining whether you should keep going.

This is all coming about because I broke a UV Max record for a particularly long level (I didn't record it unfortunately, though I took some screen captures at the tally screen). I'm sure I can do it again (and recording it as well), but was wondering if I should switch over to PrBoom+, which I also have but never use. It has a more vanilla aspect to it which would make it more legit I suppose.



Also, a huge thanks to everyone who helped me in my other thread inquiring about a G/ZDoom source port where lost souls aren't counted. This was the reason I finally got to see "100%" kills at the end of this level.

As far as "official" records go, the compet-n database and rules are your best place to go.

However, the only legit runs are ones made with vanilla doom (Doom.exe / Doom2.exe - the original DOS executables) or "competition Doom" (for a separate ranking).

Other source ports "don't count".

Most people here will recommend Prboom+ in case you're trying to record some max-demos, i recommend it too since it's the best choice for this purpose. Furthermore, demos recorded with GZDoom are generally a bit proscribed i guess.

To Patrol:
Perhaps, but I see some awesome runs on youtube that are basically vanilla, where the players actually broke the records. I remember randomly watching a video for Plutonia's Map 28: Sewers on youtube (a level I like to speedrun), and the person actually broke the compet-n record. When the tally screen came up, I said "ok, he just broke the damn record!" I'm pretty sure it was with a source port.

Same for rizera's awesome run of Doom 2's Monster Condo, and a run of Plutonia's Odyssey of Noises, I think also by Rizera, both record smashing. It seems a little 'protectionist' for the compet-n people to protect their interests when their records get smashed, even if source ports don't confer any advantage.

Tourniquet:
Yeah, I know my GZDoom run would never be considered legit (which is the reason I didn't record it), but was thinking of the PrBoom+ thing, and the legitimacy of it which I consider very real.

movement/physics between ports are not necessarily identical (particularly true for the zdoom family, which differs quite noticeably from vanilla "handling"). prb+ is great for playback/recording, but it also has features that are often advantageous (advanced huds, timers), and even full-blown TAS functionality (slow-motion, segmenting). So to me it makes perfect sense that if you want fair competition you would need to restrict players to ports where mechanics are identical and the game can't be cheated (as easily). There's significant demo activity outside of CN, so even if a demo isn't an "official" record, it's still likely you'll get decent attention/competition/what-have-you for a good run recorded in any port. that being said, I'd still recco prb+ over (g)zdoom for recording for obvious reasons (demo compatibility).

for some maps things like timers and monster counts confer large advantages. E.g. any time you don't have to backtrack for stragglers because you already know they're dead is a significant time boost compared to players forced to wander around and checking the map for them. It's debatable that because optimizing such luck-based things is tremendously not fun that it's silly to force players to go without such tools. But thems the breaks.

You can record any way you like, but:

• (G)Zdoom demos will be ignored/considered a miscellaneous oddity, because they change the game in a few non-subtle ways. The counter would be one of them, but the main critical difference is the reworked collision code. There is no "whiffing", because your target is on the edge of two blockmap tiles. As a result you can consistently kill a mastermind with one BFG shot, which is always a coin toss in vanilla Doom. Zdoom also fixes hitbox sizes, which makes it easier to hit larger monsters - especially with melee weapons. Basically, Zdoom gives you a wide range of small advantages, so it's not considered comparable to the standard. Also on the practical side, Zdoom keeps breaking backwards demo compatability, so a year down the line no one will be able to play your run back without archive spelunking for the correct binary.
  
• For all non-Zdoom wads prboom-plus is the default port most people use to both record and watch demos. You can easily control its hybrid behaviour with "complevels" that specify the sort of standard you're going for - most notably Ultimate Doom (3), Doom 2 (2), Final Doom (4) and Boom (9). Eternity Engine is capable of similar control, but with much less precision when it comes to Doom engine's more obscure quirks.
  
• For vanilla-compatible wads you can also use Chocolate Doom or CNDoom. Some runners prefer these, because they're stripped of almost any modern features and resemble vanilla the closest. It won't make any difference for the perceived "legitimacy" of the demos, except for Compet-N wads.
  
• Compet-N wads are all the IWADs, both Memento Moris, Requiem, Alien Vendetta, Hell Revealed and Classic Episode. While you may record demos with prboom, official records only count when you use CNDoom (under the reborn CN rules) or the vanilla exe (under the original CN rules). Demos from "less hardcore" ports will only land you a higher table spot at DSDA, but won't be admitted to the official sites (CN sites, doomwiki tables). The reasons include, but are not limited to: traditionalism, cheating facilitation in pr+ or extra HUD information.
  

dew said:

Compet-N wads are all the IWADs, both Memento Moris, Requiem, Alien Vendetta, Hell Revealed and Classic Episode.

And also Scythe, I believe (you can select it on compet-n site when searching for speedrun data).

RobinSena said:

To Patrol:
Perhaps, but I see some awesome runs on youtube that are basically vanilla, where the players actually broke the records. I remember randomly watching a video for Plutonia's Map 28: Sewers on youtube (a level I like to speedrun), and the person actually broke the compet-n record. When the tally screen came up, I said "ok, he just broke the damn record!" I'm pretty sure it was with a source port.

Same for rizera's awesome run of Doom 2's Monster Condo, and a run of Plutonia's Odyssey of Noises, I think also by Rizera, both record smashing. It seems a little 'protectionist' for the compet-n people to protect their interests when their records get smashed, even if source ports don't confer any advantage.

Tourniquet:
Yeah, I know my GZDoom run would never be considered legit (which is the reason I didn't record it), but was thinking of the PrBoom+ thing, and the legitimacy of it which I consider very real.

I'm just saying what the rules are. Others have already stated why source ports give advantages, but to reverse the argument - if those people managed to break records and those source ports allegedly don't give any advantages, why not simply run one of the "legit" executables and repeat the feat?

Patrol1985 said:

if those people managed to break records and those source ports allegedly don't give any advantages, why not simply run one of the "legit" executables and repeat the feat?

Two words - mouse settings. Back then when I tried original executable, mouse not turning precisely as it used to in ports was a huge drawback for me. It's also user-unfriendly hassle to work with it, so let's not be disregarding the insane requirements for setup of recording lmps with latest dosbox. Also, another notable thingie - each attempt of recording with dosbox twists starting position in random direction. Sorry, I'm not putting up with that shit and I'm definitely not encouranging anyone in doing so. Ports like prb+ emulate doom behavior and demos recorded with it play back on original executable, if that's not good enough then I'd advise on something like getting bent.

Patrol1985 said:

And also Scythe, I believe (you can select it on compet-n site when searching for speedrun data).

Scythe was added recently, and none of the records on Compet-N are actually updated for it; most runners lately have not been using vanilla for Compet-N.

Also, worth mentioning that in order to really get the "legit" experience running Doom, you would technically have to use doom.exe on an actual DOS machine or Windows 95 or something like that. DOSBox isn't a perfect emulator, and some runners have issues with how it runs, mainly what I heard is lagging and pressing three buttons at the same time (necessary for strafe50). Quickstarts are different too; I remember from an old thread that due to how fast Doom loads in DOSBox compared to older machines, starting at a particular angle and moving in DOSBox is harder because it requires more precise timing. PrBoom+ is just convenient, easy to run, and has a lot of compatibility options for different executables/ports.

Nobody blocks anyone from creating their own ranking site which will allow PrBoom+. Until then, compet-n records can only be broken by compet-n rules, unless you convince whoever's in charge to bend the rules... which is highly unlikely.

I've decided to take dew's advice and use Chocolate Doom - its controls are tight and responsive. I found PrBoom+ to be somewhat lagging in the control factor. I just have to learn to record demos; but to do it with .lmp file demos or a straight record with OBS or some other screen capture program? Leaning toward the latter.

To follow up on some other opinions - I can understand patrol's stance of keeping things pure by using the .exe's, but times have also changed. I considered j4rio's record of Mount pain to be the 'official' record over Yashar's compet-n entry when I endeavored for the record, and I don't think I'm the only one who considers source port runs with PrBoom+ and Chocolate doom legit - and all this comig from a guy who plays keyboard only, so I definitely understand the purity of the rules thing. Nothing like mastering Go 2 It on a keyboard.

In the end, it seems the PrBoom+ and Chocolate Doom people are in the majority.

All a matter of perspective. I consider compet-n records to be "world records" and I have no problem with other people having a different opinion. The original Doom doesn't allow to remove the y-axis of the mouse movement. I know it's a shitty criterion, because it's essentially "the ability to fight the interface, not the game", but still controlling this y-axis is a skill one doesn't have to have when playing modern ports.

Like I said: consider a world record whatever you want to consider a world record. It's not like there were ever some OFFICIAL speedrunning championships held by id Software.

Patrol1985 said:

Nobody blocks anyone from creating their own ranking site which will allow PrBoom+. Until then, compet-n records can only be broken by compet-n rules, unless you convince whoever's in charge to bend the rules... which is highly unlikely.

I'm not saying that Compet-N rules should be changed, all I'm saying is when I personally check for a record on a map I don't really look at the port much unless it's not compatible with whatever port the map was intended for (like GZDoom for vanilla maps). Regarding your last post, if by y-axis you mean vertical mouse movement, then that is removable legally under Compet-N using the novert utility (and many mice allow it to be removed in hardware).

RobinSena said:

I've decided to take dew's advice and use Chocolate Doom - its controls are tight and responsive. I found PrBoom+ to be somewhat lagging in the control factor. I just have to learn to record demos; but to do it with .lmp file demos or a straight record with OBS or some other screen capture program? Leaning toward the latter.

OBS recordings won't be legit, as an .lmp files are a general standard for Doom speedruns, and OBS recordings make it harder to detect cheating in some instances (OBS recordings will not be accepted to DSDA either). Also, demo recording is different from normal play because of lower turn resolution. There's really no reason not to record demos while running anyway, it doesn't take much effort.

I'm fine with separating vanilla/port demos, but imo the enhanced HUD + timer issue isn't even worth mentioning. There's a timer in vanilla, it's the music. Maybe certain maps might have a mentionable advantage with the enhanced HUD, but what maps are they? Has anyone ever put that theory to the test? As far as I know, I'm the only one to actually test the theory after someone complained I used pr+ on av26 and av29 (which I didn't have a problem doing, unlike ZeroMaster who refused to do a live stream after being asked). btw, I found no mentionable difference between pr+ and chocodoom.

RobinSena said:

This is all coming about because I broke a UV Max record for a particularly long level

What level?

All of my -respawn entries that ended with over 100% killed would have been faster with a HUD, and my perfect 100% runs would have been achieved with fewer trials.

Heh, respawn. As it was already said, you know a category sucks if using HUD yields drastically better results. Don't hate the playa, hate the game.

TimeOfDeath said:

I'm fine with separating vanilla/port demos, but imo the enhanced HUD + timer issue isn't even worth mentioning. There's a timer in vanilla, it's the music. Maybe certain maps might have a mentionable advantage with the enhanced HUD, but what maps are they? Has anyone ever put that theory to the test? As far as I know, I'm the only one to actually test the theory after someone complained I used pr+ on av26 and av29 (which I didn't have a problem doing, unlike ZeroMaster who refused to do a live stream after being asked). btw, I found no mentionable difference between pr+ and chocodoom.
....
What level?

It was TNT's Map 27: Mount Pain - I got a time of 9:19. While I am extremely proud of this time (regardless of the source port, it was a solid run - just about flawless), it's come to my attention that using GZDoom to do this gave a particularly strong advantage at the end. I executed an Arch-Vile jump in the last room to get from one side of the room to the other - something that I was not able to replicate in Chocolate Doom or PrBoom+. Evidently, ZDoom-type source ports absolutely LANUCH you when an Arch-Vile lights you up (Despite venturing into speed running, I still consider myself a casual player, which is why the more technical aspects of Doom are still unknown to me). With Chocolate Doom and PrBoom+, I couldn't even get close to the other side of the room with this method, which probably saved about 9 seconds or so in my 9:19 run. As a kid, I played the SNES and PlayStation versions - neither of which had the Arch-vile - so playing against them was a new experience when I picked up Doom again.

**EDIT** Ok, so I was able to execute the jump in PrBoom+ by running backwards, but it's much more difficult to pull off. **END EDIT**

The other side to this is that I had a run the night before where I didn't do the Arch-Vile jump, but probably would have broken the records anyway had I not died taking the lift up to the top (the normal way of doing this last room). I feel confident I can eventually break the records using Vanilla or Chocolate Doom, but since I'm on the fence about becoming a dedicated speed runner, I'm debating whether to keep putting in more time, which is the reason I haven't ventured away from the id Software released Doom games - haven't played a fan made/fan released WAD yet.

All in all, just another lesson in keeping things as Vanilla as possible for records, which is why I respect Patrol's stance, though I feel Chocolate Doom and PrBoom+ are fine so long as they're run Vanilla-style.

I mostly just care if the end product is a good watch in the end after knocking down the time. I do have issues with gzdoom though because the behavior is vastly different from prboom+ etc. I tend to ignore gzdoom speedruns and I believe this was the reason cyberdemon swapped.

Arch-vile jumps is a biggie for gz.

RobinSena said:

I've decided to take dew's advice and use Chocolate Doom - its controls are tight and responsive. I found PrBoom+ to be somewhat lagging in the control factor. I just have to learn to record demos; but to do it with .lmp file demos or a straight record with OBS or some other screen capture program? Leaning toward the latter.
...
In the end, it seems the PrBoom+ and Chocolate Doom people are in the majority.

Dew also mentioned CNDoom, which I would recommend over chocolate doom, either way give both a try and use the one you like. CNDoom is easier to set up demo recording as well.

Also I can't recall a single demo in chocolate doom being posted in the last year that I've been here.

TimeOfDeath said:

As far as I know, I'm the only one to actually test the theory after someone complained I used pr+ on av26 and av29 (which I didn't have a problem doing, unlike ZeroMaster who refused to do a live stream after being asked).

Refused is a strong word, iirc I said I was too lazy to start streaming. If anyone seriously accused me of cheating I would have no problem with it.


Vanilla, CNDoom, Chocolate doom and prboom are all fine imo, I have yet to see anyone claim that a prboom demo is less valid\offical (respawn being the exception). Edit: to clarify, I have yet to see anyone say a new record is not valid\offical because it wasn't recorded on a win 95 computer or what ever.

When i hear Official record, i think of a Record made through Vanilla original Doom2.exe .

I just want you guys to know that via Dosbox in Doom2.exe you can make it so you don't walk forward with the mouse and you can set your sensitivity about as high as you want it.

Hell, I use Doom95 with the demo and mouse patches applied to record vanilla .LMPs, since it doesn't seem to have the mouse desync bug present when recording in Doom2.exe. All demos recorded with D95 it play back Doom2.exe perfectly even despite this, so I think it would be considered legit. This is speculation though, and it's only recommended if you're on Win7 or below.

4shockblast said:

Regarding your last post, if by y-axis you mean vertical mouse movement, then that is removable legally under Compet-N using the novert utility (and many mice allow it to be removed in hardware).

Yes, I meant Y-axis and no, I did not know compet-n rules allowed removing that axis via external software :o That's actually quite a revelation - I always thought the speedrunners were simply VERY precise with their mouse, up to a point where they had complete control over the Y-axis. Thanks for the info!