New E1M4 demos
Thank you for taking the time to create and upload these. I know how counterintuitive it can be to record in this style, and we still have far too many level descriptions with no demos at all.
Having said that, I must register my strong opinion that your work is inappropriate for a walkthrough article. A neophyte player will simply not be able to follow a demonstration which moves so quickly, both in terms of strategy and in terms of resource management decisions (trading health/armor for speed and getting it back later). The maze area, for example, should be done slowly enough that the novice can flip back and forth to the automap and see which passages go where, at least on a first viewing.
More problematic than the speed itself, however, are the tactics. I think we can assume that anyone who really needs these demos (as opposed to just looking at our map and following the suggested route) is not very good. That person's main concern will be health, not efficiency or ingenuity. In your UV demo alone, I count at least 6 frontal assaults (2 others are understandable, but dubious) which could easily go very wrong for the novice player, due to the vagaries of slalom running and/or his inexperience with circlestrafing and snap shooting. We can do better than that. This is why my own recordings here are purposely as linear and conservative as is palatable while still showing off every important area. If the reader later decides that he is good enough to try speedrunning, great; there are uncounted thousands of fast demos out there for him to learn from (every walkthrough article includes an FTP link to doom2.net, for example, and other repositories are listed in Speedrun). Until then, I would guess that he needs my way of clearing the oval slime room in e1m3wik4.lmp far more than he needs your way of clearing the blue door room in e1m4wik4.lmp.
Some further, more specific examples:
- You play with the mouse, which will not be the case for people running Doom95 and Windows XP (i.e. 98 percent of the readers). For people who haven't played a lot of first-person shooters, this can be extremely disorienting.
- The elevator trick in the rocket launcher room is very risky because of the exacting footwork needed. A novice player should probably not try it because he is very likely to end up with no supercharger and a long slime passage to wade back through. At the very least, it should be done slowly enough that the viewer can zoom in with the automap and see where to stand (no, I am not making that up; it's in the instruction manual).
- Due to the number of monsters in the blue armor trap on HMP/UV, I can't see a good alternative (again, for a novice player, not for you) to getting the rocket launcher first. Certainly there is no need to save the rockets for the yellow key area instead, with so many bullets available.
- You never shoot barrels as a defensive posture (i.e. to avoid being hit from behind when it is destroyed by an enemy). If COMPET-N submissions do this in Plutonia, then we should be doing it here.
I am sorry that my own demos are the only examples available for comparison; I wish there were many others. Obviously they are not completely perfect even by the "pedagogical" criteria I describe above, but I believe that they suit our intended audience much better than yours do. The opening few seconds of e1m2wik4.lmp and e1m3wik4.lmp, for example, were deliberately designed to expose the player to as little gunfire as possible before he got the shotgun. In the latter case, 15 of 20 attempts resulted in 20% damage or less (there is 70% health in the room). I see no such consideration in your recordings.
Our wiki does not have any speedrunning tutorials yet. Perhaps you should start some, and create examples in various styles. For instance, in e1m4wik4.lmp, you get 100% items (40 seconds maybe?) and have 11 rockets left at the end. 3:10 would have been a world record in 1996...
You might disagree with all this, of course. Unless we somehow get feedback from very inexperienced players, we will never know for sure. If you do disagree, however, I would be extremely curious to know why. Ryan W 10:03, 16 February 2006 (UTC)