So, as some of you may know, I'm currently working on a test run for my upcoming low% TAS. I'm fighting Torizo at the moment.
Since the beginning is always similar in any type of run, I decided I'd show a bit of it so you could see the optimizations I was talking about all the time. Here's the first (not sure whether it will be the last) WIP of a test run:
WIP #1: Landing site → Bombs.
It's precisely half a second ahead of JXQ in ingame time, yet it contains some very small mistakes I've listed here (I didn't plan on releasing a WIP when I compiled that list, so it was timed by the full .smv; the offset is ~15350 frames). This is only a test run — the final run will certainly be faster and more clean.
New and almost new tricks and strategies showcased:
1200—1300: so called "moozooh jump". Saves at least 4 frames and looks nicer due to not having to land.
4000—4100: contrary to the popular belief, mockballing straight to the platform at a right speed is faster than unmorphing and running.
4400—4450: "catnap jump". I improved it by clearing the top row of blocks in one shot.
4450—4500: morphing serves multiple purposes here. It pushes my hitbox down, it gives me more maneurability, and finally, it lets me land in a crouched position to shoot the door before I roll up to it.
6480: hopping cancels the turnaround animation and thus lets me shoot the door much faster. The drawback is that I have to wait until Samus drops down (I miss SMR's gravity!), so it only saves 2 frames here.
6780: missile accelerates slower than a regular shot, so it scares the pirate one frame later than usual, saving a frame.
Other improvements come from better precision and jump timing. Enjoy.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
I think I will continue working on some better game than losing my time posting here.
Oh, it's hard to disagree.
Seriously, you have enough anticipated projects waiting to be worked on, rather than improving entertainment in Donkey Kong.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
I think it would be morbidly funny to list lowest entertainment value... But then, it does mock the authors. Still, the authors know that the rating is low.
I fail to see that as something the author shouldn't be ready for beforehand, especially since low entertainment rating depends on a poor game choice in nearly all cases. That shouldn't prevent statistics from having place, it would help the others notice lowest rated movies faster and possibly try to improve them regardless of who the current author is (see also: Jaws).
Chamale wrote:
Still, the authors know that the rating is low. If we make it show only non-obseleted movies, then it might motivate authors to improve their movies.
Do you see anything obsolete on that list?
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
How many times has a movie been obsoleted by a movie that is longer, but more entertaining? I assume this would fall into a similar category, and if it has been done before, then you certainly have a point that speed isn't all we care about. But speed certainly seems to be the one criteria that is considered above all others.
It has happened before. The major criteria, as I see them, were that the successor movie was better both technically and entertainent-wise. Going on a harder difficulty would prevent such a movie from being faster than the previous in actual length.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Yes vote. Unlike the others, I liked the music (I like oldschool chiptunes in general), and like others, I disliked the laughpauses and menu action. But other than that, it was short and fast enough for me.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
What I didn't like about the autoscrolling part: too much offscreen action, especially in the first half. That's humorous when you hear a hit and see a body falling down, but I still prefer to see the full action.
What I did like: continuous walljumping off the ship! And needless to say, the fighting sequence on the top deck was brilliant.
I believe one could create a pretty cool platformer on GH engine even without all the guns. Movement possibilities are so underused in this game…
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Statistics can't be positive or negative by definition (not to mention that the page already includes some of the "negative" criteria, anyway), so I say go for it. The more information, the merrier. :)
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
If I would share all of my work now, it would get copied in no time, and since you apparently have more time to TAS than me, you would finish your runs much faster than I would, and that way there wouldn't be any reason for me to continue my run at all. I want to save you all from me cancelling my run and thus miss a epic on this site (has nothing to do with being selfish or anything, just a fact) so I just have to do it like that in order to keep my motivation alive.
Well, I'm sorry to hurt your motivation further, but we're slowly, bit by bit, little by little, discovering more and more timesavers you've likely used in your run, and even those you likely haven't. We already have the way to improve JXQ's Torizo fight by 12 frames, and I know it can be improved by a dozen more. So if we continue (and I know I will) methodically nailing every possible improvement, that can not only make your run less unique and surprising, but even obsolete before it will even have a chance to be finished. :\
Deep Loner wrote:
Not to mention that some of us are getting pretty sick and tired of all the bullshit being thrown around...
I can say I'm pretty entertained. :D
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
I was pretty amazed you were the first to discover that glitch and use it in the movie. Living up to your name, aren't ya? :P
Yes vote. The very existance of this submission is an auto-yes from me. :D
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Also moozooh, you didn't answer my question why you are afraid to work on Ceres having just the best opportunity in the test run. What is the reason to not analyze it for improvements like you do in Zebes? If you decided to ignore me all of a sudden, oh well. I'm just curious, you know.
I have said it to you multiple times that I left Ceres for later for a number of valid (for me) reasons.
1. I'm doing a test run to gather all the possible improvements and implement them in the full run. But, with Ceres I will fail to get any further help in optimizing it, anyway, so I will have to do it entirely on my own. It doesn't scare me, but it requires a lot of time and can easily kill my motivation. If I will have a complete run by then, it will serve as an additional motivator, not counting greater experience I will have by that point.
2. If Snes9x v1.51 gets released by the moment I start my full run, I will definitely make my run on it. Having a perfect Ceres would be a waste of time in that case, because the emulators have different enough timing to cause desyncs.
3. I will optimize Ceres as I have already optimized Zebes. I repeat, it didn't scare me, it just was a great pain in the ass I didn't want to deal with at that moment.
Also, I don't distrust your numbers, I have always trusted in your accomplishments. But you know my ability to spot improvements when it comes to actual TASing, so if I consider a certain trick of mine improvable, I might have reasons for doing so. Whether I succeed in improving it is another question, but again, you know my dedication.
Looking forward to your WIP.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Is it right to shove it in someones face that you improved a jump or a minor section by four frames in a relativley complex and long game?
Alright, let me give you a brief summary.
During the last half a year, while Saturn was working on his version of any%, there have been two Super Metroid movies submitted. Both contained tricks not yet known by Saturn, both contained mistakes known by Saturn.
As a result, Saturn decided to start his movie over using the tricks from the published runs, but still didn't mention the improvements he knew in any way, other than "I managed to saved n frames by this point".
Of course, he did uncover some of the techniques he found, but still this was the recurrent theme of our relationships regarding Super Metroid (it was different with SMR).
So, regarding your question, consider the fact that I'm willing to show my WIP to any TASer that is going to or already have made a Super Metroid TAS, so the improvements I'm talking about are "physical", observable. Saturn's are only verbal, since he hasn't released anything aside the numbers.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
You do realise that someone like me who has practically zero knowledge of SM isn't going to realize the significance of a four frame improvement on one jump or minor section.
How is that relevant? Also, since you couldn't know the tricks in questions, let me inform you that they look different, and the difference is visually noticeable. Also, if someone won't realize the frame-scale difference, does that mean everyone should produce suboptimal runs because of that?
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Well, the real achievements are: at least 5 frames from better optimization (up to blue Brinstar to Crateria elevator, and I've saved 8+ more on JXQ's run after that by the moment of taking bombs, which I'm not sure you have), 7 frames from catnap jump, 4 frames from moozooh jump, not counting at least another 5 frames of confirmed opportunities I missed and left for the full run… Those constitute to 16—21+ frames of improvement in the real achievements exclusively, on Zebes alone, by the moment I stopped. Also, regarding "fantasy" values, I should say that the values I listed denote the parts I know I have played sloppily, or was so close to the goal that I had no doubts it was achievable, or just made a mistake. Most of them actually mean an improvement, but you won't get the actual values until I start the final run. I'm pretty sure I'll be able to improve upon your WIP further, whether you will provide me with more data or not, but it's inarguable that sharing it will surely save my time (and your, too) and make the run better.
But, I have troubles understanding your goals. For one thing, you don't want to hinder my achievements because you think sharing more data would lead to me trying to match your results, rather than search for my own ways (which is at least odd, since I use other references, anyway, it's just that they are less optimal). On the other hand, you claim you'll beat me in the end, which kinda contradicts with the previous statement.
So in the end, what is it that you want, for all of our (and Hero's, and JXQ's, and whoever comes after them) movies to be better, or for yours to still be the best? I honestly don't get it. Up to this evening, I wanted to share my WIP with you to exchange some ideas, but I had to reconsider it, because up to now it was mostly one-sided kind of exchange. It costs much to get even frame numbers out of you, so why should I feed you with my ideas when you're so reluctant to share yours? I'd rather share them with Hero and JXQ, since they do appreciate them, are putting them to good use and helping me in other ways available to them.
We have one similar aspect to our goals: to make an unimprovable movie, but guess what's the difference between the realizations of our goals.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
That's a site bug. It's not like Warp constructed the text by hand, it's just a call to the site code to generate a top 100 list.
I know he didn't, but it shouldn't be too hard to make the code generate only one string per ###M entry in this list.
[EDIT]
So I take it there's nothing to be done about that?
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
You might want to remove duplicate entries (like the ones caused by two AVIs published for one run). Otherwise, it's good, thanks.
[EDIT]
Also, "most voted movies" sounds a bit wrong to me. I'm not sure what to suggest instead, not with my knowledge of English.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Keep in mind, Saturn, that it's hard to surprise those who are actually closely working on SM TASes. I, for once, won't be surprised easily, and you know it, and I'm more than sure that collaborating openly would benefit you as well (especially since I've already improved upon your WIP).
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
So a solution that would allow everyone to listen to anything however they want would be with 3 tracks: one with A and B, another with C and D, and a 4.0 track.
Actually, 4.0 track alone would be just enough, since you can select and disable channels or downmix them to stereo, anyway.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
