In this movie we watch Spiderman act more like Superman, dodging bullets, not losing any health, and flying (well, swinging anyway) through the air, not to mention turning many infamous villains into whimpering children. This movie obsoletes the previous movie by Neofix by slightly over 1 minute.
I used Gens version 9Z Build Version 20070209, with Allow U/D+L/R, although I didn’t find any ways to abuse this. No warps or passwords. I first and foremost aimed for maximum speed, with taking no damage as a secondary goal.
This is actually version 2 of this run; I completed the first version in around 4000 rerecords, but I found an improvement on the first level. Hex-editing for this game is a total nightmare due to dead frames causing desynchs, but even though it cost me another 1500 rerecords, I managed to save another 10 seconds spread throughout the levels, so it was worth it. Improvements are many and varied. A lot of time saved came from beating the bosses faster due to Spiderman’s kick being more powerful than his webbing, which was used extensively in the previous run. More optimal web swinging, improved tactics, and lag reduction at the end of levels also played a part. Here’s a breakdown of my level split times compared to Neofix:
Level Finish times
Neofix Run
My Run
Total time saved
Level Time saved
Level 1
4885
4449
436
436
Level 2
8236
7465
771
335
Level 3
11761
10268
1493
722
Level 4
14604
12898
1706
213
Level 5
18500
15718
2782
1076
Level 6
22678
19519
3159
377
Level 7
26502
22873
3629
470
The most frustrating part about this game was it’s ungodly like ability to desynch during hex-editing. It always chose the most inopportune times to add in a dead frame, especially during the Lizard and Venom fight. I can imagine another run of this game maybe taking off 5 or 6 more seconds due to velocity control (ie through a RAM viewer), but it better be done right the first time around. Anyway, I hope you like this as a much needed improvement to one of the oldest (and least optimized) runs on the site. Happy viewing!
Truncated: Accepted as an improvement to the current run.
mmbossman: Canceled due to Neofix creating a faster, damage taking run.
Nice wall jumping! Very quick, very smooth just like a real spider. This will be accepted without a doubt. Neofix's run was nice but it needed to be actualized. Did they have frame advance and all in 2004?
Anyways, i vote yes.
The game was boring but the run was well planned and played... that, and the fact that it improves a published movie by quite a large margin warrants a YES vote from me :)
It's a clear improvement but there's no excuse to avoid watching RAM with Upthorn's improvements to Gens. You really can't tell what you'll find until you look. I strongly recommend some searching and testing. If that sounds discouraging, remember that it usually takes multiple runs through a game to adequately understand and exploit it for a high-quality tas.
I'm refraining from voting until this game is better examined.
Joined: 11/18/2006
Posts: 2426
Location: Back where I belong
IdeaMagnate wrote:
It's a clear improvement but there's no excuse to avoid watching RAM with Upthorn's improvements to Gens. You really can't tell what you'll find until you look. I strongly recommend some searching and testing. If that sounds discouraging, remember that it usually takes multiple runs through a game to adequately understand and exploit it for a high-quality tas.
I'm refraining from voting until this game is better examined.
My only excuse for not using the RAM viewer is that I didn't want to. I went through two versions of this run, and saved 14% over the old run. For me, trying to search out abstract and numerous parameters that may or may not help me isn't fun, and if I don't have fun making the movie, then I won't do them. I enjoy working frame by frame to make the best movie I can, but dissecting each individual game I do down to it's engine is no longer enjoyable for me, it's a chore. I respect that you want to see a totally optimized run. But IMO, by seeing a run obsoleted by this much, while not perfect, is still better than seeing the old run on the publication page.
EDIT: This is not to say that I'll never find a need to use the RAM viewer. I'm thinking of improving another run which has a good variety of weapons, so a RAM viewer would be a necessity to find which combination would work best with which bosses. However for this game it is more of an annoyance, since Spiderman's kick is the most powerful (three hits for most enemies), and there's enough back and forth motion that keeping a constant speed is less needed.
Joined: 11/18/2006
Posts: 2426
Location: Back where I belong
The bosses in this game are strange. Some flash while they have invulernable periods (Doc Ock, Lizard, Kingpin). The Green Goblin changes his animation but does not flash. And that giant gorilla (not monkey), doesn't do anything to indicate when you can hit him again. All the bosses have variable invulnerablity times which seem to be connected to how close you are to the boss (and possibly if you're in the air too). However, with the gorilla there is a tradeoff between how fast you can hit him (probably a little faster if you stand right next to him the whole time), and how much ground you can cover towards Sandman's sandbox while fighting the gorilla. I do believe 6 hits is the minimum required to take him down, but I'll do some testing to make sure that webbing or punching doesn't take him down in less than 6 hits. Hope this answers your question.
EDIT: 10+ hits are required when using webbing and punches.
My only excuse for not using the RAM viewer is that I didn't want to. I went through two versions of this run, and saved 14% over the old run.
I'd prefer to know that all the best tools available were used to find the best possible techniques and strategies. A relative improvement over a known-suboptimal movie makes a movie maker feel good (myself included), but better doesn't imply optimal.
mmbossman wrote:
For me, trying to search out abstract and numerous parameters that may or may not help me isn't fun, and if I don't have fun making the movie, then I won't do them.
I think you're overestimating the time it takes to find values in RAM. Once you know what you're doing it shouldn't take more than 5-10 minutes to find simple values like character position, HP, invulnerability counter, etc. Still, if it's not fun for you I'm content to have a difference of opinion. A published possibly suboptimal tas won't cause too many problems in the grand scheme of things. If it's improvable and worth the effort, it'll be improved.
Joined: 11/18/2006
Posts: 2426
Location: Back where I belong
IdeaMagnate wrote:
Still, if it's not fun for you I'm content to have a difference of opinion. A published possibly suboptimal tas won't cause too many problems in the grand scheme of things. If it's improvable and worth the effort, it'll be improved.
Exactly, and I have no qualms about someone else trying to improve on my work. However since the previous run has been around since late 2004, I doubt our fellow TASers will be lining up to try their hand at it. Glad we can agree to disagree:)
True, although taking damage.... ehh, not so sure about it. But it is certainly much faster, with better routes, so I can't argue too much. Congrats, and you may see me taking another crack at this.
