Jump to content
Search In
  • More options...
Find results that contain...
Find results in...
termrork

New doomedsda notation suggestion

Recommended Posts

The last weeks a small idea growed to something which might be interesting:

A new notation for doomedsda for segmented speedruns

what we have now: segmented runs are also denoted as TAS, no matter the demo is segmented or build or anything else edited.

pro: nothing has to be changed

con: one has to check the readme file wether it is a TAS or SEG run

idea: segmented runs: SEG, all demos edited in any other different way than simply playing the demo in parts with 100% speed

pro: a chellenge would arise, why?: in a SEG run a speedrunner can do runs which are not possible in non-tas (at all or in that time). I think there are more people who would do SEG runs than TAS runs, since it is less time consuming and in my opinion more fun.

con: not sure wether it is easy to make a new notation for the website...


example a few weeks ago:
-first I uploaded a SEG run on avj map32
-Azuruish uploaded a TAS run faster than my SEG run
-I know from one person who is willing to break my SEG run with another SEG run, but not willing to break the TAS run, since this is barely possible with a SEG run without a huge route improvement

so, this person knows Azuruish's run is a TAS run and mine is a SEG run..
another person just looking on doomedsda has no idea about this, and if he is like me, then he will not check the slower TAS run for any reason


so... if there is interest from people doing SEG runs, this might be fun for those, if not I guess it should not be realized


P.S.: recently I saw a few people complaining about TAS runs because they were not inch perfect, what they HAVE to be since they are edited... what the hell? If everybody has this opinion, then I think there is no room for SEG

Share this post


Link to post

I guess it's a little problematic. TAS demos exist since ... i don't know... A lot of it on DSDA and i doubt that someone will check each for marking it specially.
Also that separating on SEG, TAS it's a kinda weird as i used segments too. 123 segments and hundred replays each segment with speed lower than 25 to make it perfect.

termrork said:

another person just looking on doomedsda has no idea about this, and if he is like me, then he will not check the slower TAS run for any reason

Obviously!

termrork said:

what the hell? If everybody has this opinion, then I think there is no room for SEG

Yep.

Share this post


Link to post
termrork said:

P.S.: recently I saw a few people complaining about TAS runs because they were not inch perfect, what they HAVE to be since they are edited... what the hell? If everybody has this opinion, then I think there is no room for SEG

That's because one of the main objectives of a TAS is to see what the fastest possible time to complete a level in a select category is. If you don't manipulate monster damage correctly for example then you're wasting your time and not using the tools to their full extent. It's like sitting on the loo and only farting but not taking a full shit.

I guess it's a little problematic. TAS demos exist since ... i don't know...


TAS I believe existed within the early 2000's or at least that's when it was first widely discovered due to an infamous SMB 3 run.

Share this post


Link to post
Springy said:

That's because one of the main objectives of a TAS is to see what the fastest possible time to complete a level in a select category is.

Ugh, it's this argument again... Somehow I don't recall god coming down from the heavens and telling people what TAS demos should or should not be like. People will use TAS tools in ways that they find fun or interesting or whatever, and I'm sorry but I think it's not your place to say that they are wasting their time.

Anyway, regarding termrork's suggestion. When I asked Andy Olivera if it would be possible to display centiseconds for demo times, he proposed that people write them in the filenames (like "lv01-513.zip") if they want to put an emphasis on these centiseconds. Maybe we could do something similar with "full TAS" vs segmented runs. Like, put "SEG" at the end of the filename.

Share this post


Link to post
Memfis said:

Ugh, it's this argument again... Somehow I don't recall god coming down from the heavens and telling people what TAS demos should or should not be like. People will use TAS tools in ways that they find fun or interesting or whatever, and I'm sorry but I think it's not your place to say that they are wasting their time.

While I'm not opposed to distinguishing proper TAS and segmented runs, your argument is very much invalid. People can use the tools in whatever ways they want, but it's also going to be people judging them. There certainly wasn't any god coming down and telling us we have to accept whatever lazy crap some bored millenial produced to get his time-wasting hobbies validated by his equally miserable peers. You'll just have to cope with criticism, because it's not going to go away. I should also point out that Azuruish's demos would STILL be bad-to-mediocre TASes, not good SEGs, heh.

Share this post


Link to post

I wonder why only TAS demos get such criticism and snarky remarks. For instance, I don't remember anyone making fun of Daiyu Xiaoxiang for sharing his early demos, which were "mediocre" too. It's like there is some sort of "elite TAS club" which dictates what's worthy and what isn't and gets annoyed when others don't play by its rules. I don't like this... But I think I'll shut up now because I just keep repeating the same thing pretty much.

Share this post


Link to post

well, it seems so that the people think TAS has to be as perfect as the creator can do is the opinion of the majority... then segmented runs might never be accepted by the majority, I did not know that at all.

For me it is completely different: no matter what a demo is, TAS, SEG or normal, in my opinion the most important thing of ANY demo is wether it is fun to watch or not. The ranking for fun for me is:
normal > few seg run > a lot seg run > bad TAS > TAS

when I see a run I am also always wondering if I could do the same too, if not I would honor the creator a lot. If it is a normal run A LOT more since for a TAS run I am always thinking, if I put enough time in it, I might build a faster demo (you cannot really say that for normal runs, since the skill/time spend curve is comepletely different).

at least in other games I know I am not the only person who think so, on youtube there are concepts of people speedrunning life in front of an audience. there is also a doom2 video of it where the player saved and loaded a few times through the run (around 5), which is imo very interesting, because there you can see what he can do when he has only a few chances. The time thuogh is of course miserable even compared to normal runs, but that does not make it bad... at least for me

also a lot of TAS people do not even try to make their demos look pretty, since that is not what counts, only the time counds... and this way ugly (for me) huge and fast turns, ridiculous (for me) shooting monsters which are far away and so on, I guess you got my point and for now I hope I do not have to write what my opinion is

@Springy
@memfis
@dew

well I think you cannot deny you did not get memfis point... it is a point no matter you agree or not. I completely agree with him.

the thing that one would have to check older demos is a huge argument of course, but I am not sure, since when segmented runs are possible? I guess that TAS was possible before SEG. Furthermore I am not aware that there are really SEG runs out there with 100% game speed.

Share this post


Link to post
termrork said:

well, it seems so that the people think TAS has to be as perfect as the creator can do is the opinion of the majority... then segmented runs might never be accepted by the majority, I did not know that at all.


I don't think it has never been about optimizing, it has always been about entertaining, though SOMETIMES those two go quite close each other in speedrunning. I have always wanted to know, personally, how much the game can be pushed as speed, not just in doom, but in Mario, Mega Man etc. Sure, there are other things to look at than just basic speed for example different routes/tactics/glitches and smoothness.

termrork said:

another person just looking on doomedsda has no idea about this, and if he is like me, then he will not check the slower TAS run for any reason

termrork said:

For me it is completely different: no matter what a demo is, TAS, SEG or normal, in my opinion the most important thing of ANY demo is wether it is fun to watch or not. The ranking for fun for me is:
normal > few seg run > a lot seg run > bad TAS > TAS


Hm? First you say you don't watch the slower ones (shame on you!), and on the second quote it's all about 'few seg run', where the end result will be on the slower side of TAS 'categories'.

And how is it possible that a BAD TAS is BETTER than just TAS? What? Or am I looking your ranking upside down? Why is there even a ranking when you just said it's all about fun/entertainment? I'm really confused at the moment.

Share this post


Link to post
Looper said:

Hm? First you say you don't watch the slower ones (shame on you!), and on the second quote it's all about 'few seg run', where the end result will be on the slower side of TAS 'categories'.


Yes shame on me, but that how I do it and I bet there are other people who do it like that. Yes you are absolutely right, when I am not watching slower runs how could I know that there are SEG runs or not... well I do not, but I also have not seen a single one, although I have watched almost all TAS runs of iwads (maybe there were some but they were not declarated as such).

Looper said:

And how is it possible that a BAD TAS is BETTER than just TAS? What? Or am I looking your ranking upside down? Why is there even a ranking when you just said it's all about fun/entertainment? I'm really confused at the moment.


My ranking was not about a good run or a bad run, just what I enjoy more to watch. For that I like to watch a BAD TAS run e.g. which is segmented in slow speed than a TAS run. The ranking is about fun, and of course my personal one, other people for sure have different ones... I guess you could also make a ranking what you enjoy most to watch. Of course on big average. I hope it makes sense now for you, sorry that I confused you... again.

edit: I did not want to convice people that this is whole thing is a good idea. I just wanted to present it, and if it is doable AND a lot of people like it, one could implement it. At first place I did not want to present it anyway, but was convinced to do so by someone who liked the idea.

Share this post


Link to post
Looper said:

And how is it possible that a BAD TAS is BETTER than just TAS?

Maybe because too much optimization results in too inhuman-looking demos. I remember Nikolay "Heretic" Kamnev once said that he prefers Andy Olivera's older TAS demos from before Andy switched to built runs. He wrote something like "Andy's demos used to be lively, now they are purely mechanical". There is a point where TAS demos turn from believable to "OK, now this is definitely TAS", and when this change happens something important is lost. I definitely can understand where termrork is coming from (or perhaps he has some other reasons).

Share this post


Link to post

One of the big culprits in the built TASes is all the moving around from place to place without looking at the destination. That's probably the most unnatural thing.

Share this post


Link to post
termrork said:

The ranking for fun for me is:
normal > few seg run > a lot seg run > bad TAS > TAS

I prefer this:
Non-TAS -> Lame TAS -> Average TAS -> Pro TAS

Non-TAS - not all of prboom tools and xdre were used.
Lame TAS - Only segmenting TAS. Sometimes when i'm watching such demos i'm asking myself: "is it TAS?"
Average TAS - my category/quality. Not just bad but not good enough as expected. Here used all tools but XDRE used only for 'instant glide' and for searching precisely coordinates.
Pro TAS - XDRE build demo. Player in this quality is ready to spend a few weeks just for saving 10 secs. Requires best of the best time and quality.

But again. Tools is just tools and it doesn't make all work for speedrunner. All depends from skill. I happy that with TAS i can do unreal thing (SODfinal NM for example).
And... Discussing about quality of TAS it's such a endless theme as what is better: Android or IOS or Console and PC.

Share this post


Link to post

A lot of wrong assumptions flying in this thread. Lumping all TASes into one category is wrong, of course. There just wasn't any real effort to establish them "officially". If I had to stake them out, I'd probably go with:

  • segmented runs
  • played TASes
  • built TASes
Someone can try to pressure you into "improving" and using more tools, because it provides more space for optimization, but that's just their opinion. Like people suggesting UV speeds to maxers and whatnot. I also do agree with memfis on the point that overly technical demos feel a little soulless and many might find them less enjoyable to watch. The more tools, the less they are performances and more like technical achievements. It's outright wrong to use words like "bad" or "lame" to described these categories. Just because a run uses a smaller set of tools doesn't make it inferior to a fully built one except for the final time. On the contrary, all of the categories I mentioned can receive "bad" or "lame" entries by lazy or unskilled players.

Another huge mistake that's starting to be a growing trend is treating DSDA as a hall of fame or a book of records. It's a demo archive. It lumps together anything thrown its way. The way it was designed originally is definitely outdated now that we "know better", or rather know how we'd like to sort all the entries, but it's too late for that. Not to mention that marking all TAS demos the same way is the very least problem the archive has. Even having them in the same table as regular demos would be superwrong for a real record table! Not to mention having (G)Zdoom or coop demos mixed in as well!

DSDA tables cannot be taken at face value, they need interpretation. If you want a book of records, either start editing the valid times into the wiki or start a site that parses DSDA tables and interprets them in a proper way. Andy has way too much work for one human being to maintain that ever-growing megadatabase, so restructuring it and then filtering the demos one by one into the new brackets is not really an option. As for the TAS categories... it's on my to do list to write a formal introduction into the matter. It's becoming obvious we need to set a standard, or at least explain the commonly used terms to even be on the same page and have a coherent discussion.

Share this post


Link to post

@Azuruish

yes, please do not use the words good or bad for any category. the same way we could discuss wether blue or red is a better color. There is no way to be objevtive and even the comparabitlity of normal runs and tas runs in terms of good or bad goes to zero, like with colors. For Android or IOS or Console and PC I would go Android and PC, but there in my opinion you can compare these, since they try to give you the same experience, an OS for a mobile and a platform for gaming experience. Each of them has advantages and disadvantages. To rate them you have to be subjective again of course, but how could you do this with demos? Imo you can't, they do not even try to give you the same experience.

@dew

not sure what you mean with wrong assumptions? I always pointed out that the idea only works if it is implemented easily. You are right, if you read my text closely one could think that I only presented this idea because I am too lazy to look into the txt files of the demos. But this brings me to something which bothers me the most in the context of the whole discussion:
People tend to write less and less into the txt files, from which you sometimes cannot even distinguish what kind of TAS run it is (usually you can see it from the movements of course, but distinguishing good played TASes and built TASes can be very difficult).

So another way to solve this problem one could make a sample txt in which it is indicated that you really should tell how the demo is made (e.g. with example sentenses like "TAS methods used: SEG, played TAS, build TAS" and one simply has to delete out the methods which were not used).

If everybody would use this txt file then, I would be happy and there would be not that need of a new notation. If the reason people become minimal in writing txt because there are lazy, then one could write a small subhomepage with a few text boxes and check boxes which then generates the txt file automatically for download (this way one could also write programs directly analyzing the txt files and make for example a book of records from demos you have on your HD).

Share this post


Link to post

Adding some info about the tools used would be greatly appreciated, of course, even if they're only indicated by one of the categories I mentioned. "Segmented run" is a well established term used by the SDA and "built TAS" has a strong tradition in this forum. I called the middle category "played" to indicate the input was genuinely recorded ingame, not manipulated in an editor. If anyone knows of a better or standardized term used in other communities, do let us know.

Share this post


Link to post

I am not saying that a separate category for SEG is bad what my opinion was going towards was built TAS demos more specifically, ones that use XDRE 2 that are half arsed. As Dew rightfully pointed out be prepared to be judged by viewers. The fact is TAS viewers expect high quality if something is focused as a means of getting the best speed then people expect every second shaved. If it's a TAP like a lot of the MK movies are then obviously they want to see the most out of the game. I do agree though that entertainment must come first.

Share this post


Link to post

Sorting and relabeling all the TAS demos at DoomedSDA could be more trouble than anybody wants to go to. If new segmented runs are posted with SEG in the forum description then it might be simple enough for Andy to label new updates as "TAS" or "SEG" going forward. He'd have to speak to that.

It's interesting that "SEG" runs haven't had more uptake in the Doom community, when they're basic to the rest of the speedrunning scene. That's probably thanks to Doom's ability to record native demos, and the fact that efficient screen recording is a recent invention (compared to Doom's lifespan). There was no way to stitch a segmented run from saved games and video recording in the early days, and the first tools that would have made it possible to record a segmented run on a single demo file also made it possible to build demos more thoroughly. Segmented runs were completely skipped in that progression. The pressure to normalize them now comes largely from without - we see that SDA has them and we want them too.

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×