Roll Eyes
Lips Sealed
1 page
List results:
Search options:
Use \ before commas in usernames
I remember when Ken's e1m2 ER in 00:29 came out a lot of people were in disbelieve that it could possibly be true. I was one of them. Until this year I still couldn't figure out how one could get that time.

But three months ago I had one of those occasional bursts of wanting to play Quake and managed to beat my old PB of 30.7 s with relative ease. Then I set out to lower that time until eventually... I got to 29 s Grin So it seems like it was just a matter of painstakingly shaving time here and there. I wanted to make this thread just so people know it is possible for mere mortals.

And not only that, it should be improvable. My best run had a bad start and a bad finish. Here's a comparison of two runs from me and the record run:

Reference |     My :30     |     My :29     |     Arturo     |  Combined
start     |  01,33 (1,33)  |  01,33 (1,33)  |  01,37 (1,37)  |  01,33
door      |  10,78 (9,44)  |  10,95 (9,62)  |  10,79 (9,42)  |  10,75
sk        |  14,58 (3,81)  |  14,61 (3,65)  |  14,48 (3,69)  |  14,40
sd        |  19,72 (5,14)  |  19,70 (5,10)  |  19,89 (5,41)  |  19,50
button    |  21,51 (1,79)  |  21,52 (1,82)  |  21,74 (1,85)  |  21,29
end       |  30,04 (8,53)  |  29,98 (8,46)  |  30,00 (8,26)  |  29,55

I wanted to improve my run, considering it's slower than the record if you take away the 0.04 s advantage in loading time, and considering I later managed to improve the beginning and ending of my runs. However, that craving for Quake passed before I managed to do that. Those later improvements considered, the best combined would actually be closer to 29.4. But, of course, getting everything right in a single run is improbable.

Another thing, my testing indicates that jumping the gap to the silver key would save about 0.5 s, so 28 s is a possibility... if someone pull off that trick Wink
Thread title:  
I love YaBB 1G - SP1!
I just checked this site for the first time in months and I now see there are guys beating so many records on ID maps! Wow, amazing. I have not played Quake in such a long time and don't even have it installed, yet I want to watch all those new demos! Smiley

I found this thread talking about my e1m2 ER and had a good time reading it. So someone also got :29, nice, congratulations my friend. It was definitely not easy for me, and yes, at first it seemed impossible to me as all other folks who got :30 were getting 30.9's... then someone showed me a 30.7 I believe (if my memory isn't failing me) and I decided to try and shave some fractions of a second here and there, in corners and stuff like that. Fun way to improve a demo but frustrating at times.

A guy sent me a demo once showing me a bunny directly to the silver key, without extending the bridge. I tried VERY hard to pull off that trick, with ZERO success. It actually made me think the guy either cheated, or created the demo in a different host_framerate, as I remember setting that variable very low allowed for some longer jumps. But he seemed so honest to me, that I just had to believe him back then....

So cool that guys still play this awesome game. I am not and have never been a gamer, so I don't play games and have not for years.... but if I ever decide to play one it must be, of course, Quake. I have three sons now... maybe one day I'll set up a network and show them how to play and we can play some QuakeWorld, haha. In my later years of playing Quake I enjoyed QW so much more than speedrunning, but with crappy internet connections and all servers in Europe or the USA, and me being in the caribbean, I had to be content with like a 120ms ping... and still found it fun! But would have loved to play with a ping of 13ms like the Europeans and frag the sh*t out of people in my favourite 1on1 map, dm6 Smiley

Ok that got off topic... sorry. Anyways nice to hear another person got :29. I would be interested in watching this, can a video be made of this? Maybe a prive upload to Youtube? GinKo, if you can, please let me know. I may check this site every few months or years, so give me a shout at garcialasca (, if you have some spare time. Happy running!
Here you go:

I also never managed to do the jump to the silver key (without cheating), but it's so close that still think it might be possible. It's such a long shot that e1m1 ER in 0:20 seems more likely.
Stupid question, but is there no room to first do a 360 or 180 inside the silver key room to gain some extra speed or is it not faster anymore?

When you're waiting for the barrier to open why couldn't you do some strafing to get some extra speed before entering the final corridor?

I don't run this game :DDDD but I love the runs.
On the silver key, you mean in the hole on the wall where the key is? There's not enough room there to get speed to start bunny hopping. So what we do is we hit the wall (to quickly lose the speed towards that direction), then walk out of the hole and onto the bridge that is being extended, and then start jumping from there. Or do you mean trying to get more speed before going to the button? In that case, I don't think there's enough time to do more than what you see in the demos, which is to prepare for a good jump as soon as the wall is lowered enough.

As for the barrier leading to the final room, it is possible to do a jumping sequence and then time it so you go through the gap with more speed. But timing it would be really difficult and you wouldn't gain much from it. It's better to go for a single jump so that you can time it better and not lose time by reaching the gap too late. But yeah, you could get maybe a tenth if you manage to fit a good jumping sequence there.
Okay then. Now someone just needs to render that 40 minute easy 100% by Elgu...

Are you sure you couldn't use some kind of buffer that starts from hitting the plate that triggers the barrier-opening sequence? 10 boomstick blasts and a certain positioning? Oh well.
I just supposed it would be hard, but I hadn't tried it. Now I did some testing, and it turns out it isn't much harder than the normal way.

Ideally, you would go back to the room before the silver key door and start jumping from there. But on a proper run the grunts stay in the way. So the best strategy I found was to start from right after the door and do a 3 jump sequence. The jumps are a bit awkward because there's not much room, so you can't get too much speed from them.

The bad news is, you really don't save much time with this. With the brief testing I did, I was able to save 0.08 s over my 0:29 run (in that specific part of the run), which is just on par with Arturo's run. With some better timing (and better jumps), I would expect up to 0.1 s saved relative to his run.

As for rendering the id1 EH, I'm afraid I can't help there. I struggled to render that 40 second video already.
Ha! I technically found an improvement :/ If someone's going to use that they'd better credit me!
I love YaBB 1G - SP1!
Just saw this now. Very nice, the 0:29. Thanks for sharing.

GinKo: thanks for letting me know you at least tried the direct jump to the SilverKey. Try it many times from a saved game (from around the yellow armor area) and see if you get it. Some jumps seem impossible but are doable. For example, there is a jump in the popular 1on1 map dm6, across it's very center, that is very easy to do in QuakeWorld (you gain tremendous speed there), but in regular Quake I was never able to do it. Maybe I did not try too hard, but I remember never doing it, and I saw a video by Markus doing it. Not only do you have to develop tremendous speed, but seems a key is jumping off of the VERY last possible frame.

Unless the person who showed me the e1m2 silver key-jump was cheating, then I suposse this is how he managed it. You have to jump from the very edge, almost from a place where you think you'd fall straight to the water if you land there, but there is actually a little bit of (possible invisible?) space where you touch ground and land, and get to the SK. It would require a lot of patience but at least the rest of the run is not too hard and not very monster-dependant, so it might be doable... man, the good old days. Smiley You folks have brought me some great memories.

Once again thanks a BUNCH for uploading that to Youtube!! Smiley
I love YaBB 1G - SP1!
Hi Ginko! It's me once more. I am so late to reply to this whole thread, but today I just re-read your original post up there and noticed something funny. You said when my :29 came out, and I quote, "a lot of people were in disbelieve that it could possibly be true. I was one of them".

May I ask what exactly does this mean?

I remember, years back, someone who I guess trusted me quite a bit, let me know about some rumors regarding my demos. To me, that was absolutely hilarious. I never even *thought* about cheating in any way. I guess one of the consequences of respecting the game so much back then was exactly that: it never even crossed my mind, and I do mean NEVER.

I was always very naive when it came to making demos. Only towards the end of my demo-making years I discovered one could make water transparent... this helped tremendously in maps like e1m4. I was also made aware of a "script" to lower the frames (or whatever) which enabled me to make the e1m1 jump "out of the water", which was otherwise impossible. And I used that bit of code and even mentioned it in the TXT file.

I was made aware of who was accusing me, and while I understand I p*ssed off quite a few speedrunners as a younger guy, I was never ill-intentioned and was quite surprised that people would even think I would cheat.

Anyways, if you could enlighten me as to what "a lot of people", and you, thought, or what your explination for my :29 was up until this year, I would love to read it, if only for a tiny moment of amusement. Smiley Thanks in advance! Smiley
I'm afraid you got it wrong. What I meant by that was that I was really surprised, because I thought :30 was optimal (bar some route change). I thought it couldn't be done, so I was a bit dumbfounded by the news. (keep in mind that my personal best at that time was probably :32, if not :33). And that's my recollection of what the general reaction was at the IRC channel, people were in disbelief because they didn't think it was possible*.

* that is, before watching the demo. Once you see it, you think "well, I guess it was possible". I don't recall anyone calling it cheating. I do have a vague memory though of someone going on a rant about "recent demos being suspicious" back then, but I'm not even 100% sure who it was or if it was really about this demo. I might be mixing it up with something from the saga of the guy that actually cheated. It's been 10 years after all (God, we're getting old).

Another point, which is partially what motivated be to create this thread, is that, as I wrote, "I still couldn't figure out how one could get that time". Don't get me wrong, it has nothing to do with cheating. It's that I always try to understand where does the times come from. Sometimes it's easy to identify it as the use of an extra trick, or better bhopping, or much quicker and more precise movement. I don't think I will ever be able to get e2m3_022 ER, 100m_009 ER or e3m2_013 ER, but I can tell where I lose time and what it would take to get there. But in other cases it's not obvious, I seemed to do everything just the same on e1m2 and got 30.9 or 30.8. The end ER in :17 was the same thing. Understanding what you need to do to match or beat a record is always the first step (in my approach, at least). There are still plenty of runs that I haven't figured out.

I have the impression that my reflexes aren't as good as they used to be. But I think I'm now better than ever when it comes to understanding where time differences come from. And I wanted to share my findings in this particular run because it was interesting how I couldn't figure it out for 10 years and, in the end, it was really just a case of being a tiny bit faster at every point (and the obvious playing over and over until everything went perfect all in a single run).
I love YaBB 1G - SP1!
Hi GinKo! Sorry I am a bit late, but I'm very thankful for your detailed response! I will first share a short (old!) story and then share input on the very subject you're touching with this latest reply.

First: When I broke some other records, I heard claims that people who had tried to break them (or had held the record at some point) felt that I did things they never managed to do, or they never got "lucky" as in, monsters in a certain place, etc. I think a person questioned in my e1m1 NH and a kill I got while quickly turning around and firing some quaded nails "behind" me. This type of kill you do not typically train, it just so happened that in that particular run I felt a monster behind me, tried to kill it without losing time, and that ended up being the record run... no big deal to me. Said person later said that I had "learned a thing or two from <insert last name here>". I will not mention who was being referenced to here, but apparently someone who had used either scripts or other cheats in some demos. I was very shocked as I never even considered doing such thing. What would be the point? I think in my e3m1 EH I kill a dog in a similar, lucky way, but that demo got beaten so it's a moot point. I then proceeded to show one of the SDA staff a few demos of me getting very close to the record time, or getting it but missing a kill or so, without the "lucky kill", just to show I was consistently finishing it without any "suspicious" kill.

Then in an update, I remember breaking a record in a custom map called "pb1", someone had taken both runs, I think Peter, with :34 or so, and one day when I learnt how to bunny properly I got :33 in both runs... and I discovered a nifty trick (not something I usually do) about getting an ogre grenade-jump + fiend boost to the top, which I knew would cut the run in half or thereabouts; many times I thought I had gotten to the above floor, yet it seemed I got "stuck" at the very border and then fell down; when I finally managed to get the trick, I also felt I might get "stuck at the border" so I moved my mouse nervously from side to side, trying to "strafe" my way into the top, if that even makes sense... in that update it was asked "why the strange mouse movement at that point?", which to me felt like a hint to something "suspicious", but again, it was a clean demo. I liked doing that one. Smiley Added comment about this "getting stuck on the border": this is the most I ever got trying to jump directly to the silver key in e1m2, sometimes I felt I got stuck at the very border, thinking I might actually make it, only to then land in the water...

On to your point about "knowing" where the time is saved... you're absolutely right, it is sometimes obvious, but experience had taught me that many times it is not, and VERY TINY subtleties make a difference. e1m2 is a great example... I too, like you, at one point was getting :33 or :32. How did these guys made :31 by doing the exact same things as the :32? And then :30 doing, again, what seems to the untrained-eye to be the "exact same" thing? The answer is, it only LOOKS like the same thing, but small subtleties here and there add up to a big difference. I shaved time in little corners and turns that perhaps each saved just a few hundredths of a second, but added up, they make up for a full second at the end of the run. Smiley Another map that taught me this was e4m7! From the time Markus had made :45 I was in love with this map. I once chatted a lot with a QW'er called Simon Nordberg and pushed him to make some demos, since his bunnies were very stylish to me; he matched Markus's :45 but could not beat it. QW style bunnies are stylish but eventually proved to be slower than optimally-performed bunnies in regular Quake. Then came Jozsef and got :44 .... wow, how could he do that while using the "exact same" route? (hint: it is never "exact same"). And to make things worse, then came Kay, and got :43, again doing pretty much "the same" (NOT!). I was just playing that map for fun because I liked it one day, and was getting decent times but nothing that made me believe I could break the record... but upon noticing small changes here and there and the times getting slower, I eventually got a :42! If you had told me back in the days of Markus's :45 being the record that a :42 was doable without any visible new trick... I would have not believed it, that's for sure!

As if this post wasn't long enough, another example: e1m1! The very first map. From Ilkka's :24 to Markus's :23 you can see better bunnies and placed in better places... but how did it get down to :22???? It is hard to explain. I remember when the "bunny comparisson" thing started, I played a lot of e1m1 ER as I loved it, and I also got :22 which I think is submitted there somewhere... and it is a funny feeling... at first (say, a couple of days earlier), you think you're doing everything right and you barely get a high :23 and think (how can I go faster????), yet, by continuing to do what appres to be THE SAME thing, boom, you get a :22. In those days I felt I was on fire and I remember getting three consecutive :22's one after the other! Talk about a tough feat to match! It got so easy for me to get :22 I felt I could do them whenever. Yet, when I lost practise I couldn't do them again on demand, not even close!! When you play the same map over and over you learn those subtle little things, you realize in which places you have to turn what way at what moment, where to wall-hug, where to jump from, where it is best to land, how to best manage some stairs, etc. This can only be learnt by continually trying and experimenting. By the way, I always thought a :20 could be done in e1m1 with the barrel. I never even got that trick to work, not even once. Sad

Sorry this got so long, but considering I rarely post here, lets pretend I just wrote one parragraph in a different week and it's not really such a long post! Smiley
"with the barrel" - you mean by jumping next to and blowing up the barrel in the entrance hall, or jumping off it or what?
I love YaBB 1G - SP1!
Sorry to take so long to respond.

I meant to say: if several of us got :22's without using the barrel to gain speed, and I remember getting several 22's at one point.. I think my best may have been that 22.8 in the bunny comparisson page but I was getting so many 22's that I knew with "perfect" play a much lower 22 was possible, even for me. So lets start from there: lets assume that a "close to optimized" e1m1 without the barrel is actually a low-ish 22... and this is all a hypothesis because many demos were thought to be "optimized" and, apparently "out of nowhere" a whole second was shaved by the techniques and little nuances I mentioned above.... so lets say a low 22 is possible with just bunnies... and who knows if maybe even a 21 (!?!?) ... how can the record, using the barrell for impulse, be "only" 21? Lets consider this: a) Perhaps Peter in his run was not so efficient in the part of the map before the barrel; b) perhaps he did not maintain  asspeed through the bunny sequence with the impulse as best is it could be done, and c) perhaps after that sequence is over he was not the fastest in the last part of the map;

Or maybe I am crazy and he was absolutely perfect in every stage? I always refused to believe it. I always thought with "perfect" play a :20 was possible. Like I said, if we start from the hypothesis that a low 22 or maybe even a :21.9 is possible WITHOUT the barrel usage, how can it's incorporation only save so little time? There are plenty of bunnies that can be done in continuous way after the barrel. Also, if I remember correctly, he pushed the button to release the "floor" over the slime. Is this truly necessary? I remember a video, maybe by Attila Csernyk (sp?), where he jumped over the slime without the button. Now, specially with the impulse gained from the barrel, is this really necessary?

But even if it is... there are parts of the demo where he "walks" a little bit. I insist that a :20 should be doable with perfect play and the most efficient usage of the barrell and it's following sequence. The pushing of the button should be evaluated, as not pushing that button would preserve much more speed throughout several bunnies (untill having to reach the first of the three buttons by the slopes). I would love to see one of these newcomers who are still breaking ID records take a shot at this.
You know "ginkō" is 'bank' in Japanese right? That means you can bank on him.