sprivenspriven

so i took some time to look at the 3 main ways to play the game and discovered some interesting things. IGT inaccuracy is one thing but a 10 second difference between the anniversary collection, hardware and gambatte is pretty crazy.

i think this pretty much proves that the anniversary collection runs faster/has less lag. if you wanna do your own testing i can provide the video's i used in this just hit me up.

Oh_DeeROh_DeeR, DancarnateDancarnate and 3 others like this. 
  Rayquaza911Rayquaza911
(edited: )

When they first submitted the 19:19 time, I only tested the autoscroller section, which I thought was the laggiest. I only did the vertical autoscroller part. I managed to tie the time on Gambatte (that was a time of 2m23s) using the same strategies, which was 10s improvement than my old run.

After their new insane time I figured I try a bit more, I only did the first 2 sections pretty much identical I messed up on small part, and the SGB2 timer was actually faster for about 25 seconds, until lag management became the problem. https:/​/​streamable.​com/​1whpzt So yeah I believe regardless of platform getting a good time is doable (as I tied the WR auto scroller section done on emu).

So want to get a good time? Destroy everything as soon as possible, you whip a candle say 5 frames later, you will be behind, enemies stay a few frames longer on screen? You will be behind. When I first played this game I didn't know candles contributed to lag, you know how many there are in the game.

From my opinion, timer isn't the issue, gameplay I did not notice much if any difference, and I don't think the AC version would be faster than the SGB if you played through the whole game. They did make a new TAS done on emu and there are more places where objects can be destroyed so this time can still go lower. Until someone wants to do an AC TAS we wouldn't know how much faster it can be.

I have no time to prove how much fast the AC version is for the rest of the game, someone else can do that but I'm certain the time is beatable if someone actually wants to do all the record strats, just from my old run I can already save 45 seconds in certain areas. But if all you're going by is a timer, then this discussion is pretty much moot lol

 
  LaxxusLaxxus
(edited: )

Nice that you mentio his TAS. Cause his current PB is 20 frames ahead of his TAS after the first two rooms. not bad on the execution side.
1314 vs 1293 from standing at the beginning until entering the third screen.
Not gonna comment the other nonsense anymore

edit: and the whole first stage is only a few frames slower than the TAS, with mistakes like missing the eyeball and having to do standing whip. respect for that execution.

 
  Rayquaza911Rayquaza911

Yeah, he didn't destroy every candle in the beginning if you noticed, hence why I said the TAS is improvable in certain areas, but ok.

 
  LaxxusLaxxus
(edited: )

alright, more denial, didnt expect anything else.
His vertical Stage 3 autoscroller is almost 6 seconds faster than his TAS. From the first frame a spike is visible at the bottom until the frame the candle at top appears:
PB 2:18.133
TAS: 2:23.900

yep, his TAS must really suck. How much proof do you want?

 
  KomradeKomrade

Sounds like the anniversary collection version is the version to run then. At least that makes sense to me. Run the fastest version, I thought that's what we did around here. I guess splitting such a tiny leaderboard is a better idea, though you probably want to do it for emu and super game boy, cause there's differences there as well, but ultimately I'd say let it rock.

BurbBurb likes this. 
  Rayquaza911Rayquaza911
(edited: )

Denial, didn't expect anything else? Your attitude needs improving. (thought you were done commenting on nonsense?)

I'm happy to give answers to questions to those who actually want it. The TAS was made last August, improvements have been made since then, and to keep repeating myself to get the point across, sprites that stay longer on screen contribute to the lag. Almost 6 second difference in RTA but only 3 IGT. The answer is the falling platforms, they stay longer on screen in the TAS. Some worms are on screen longer too. I'm done explaining differences between the TAS and WR, keep comparing different gameplay styles by frames if you want, I've made my point.

Comparing the WR and the current TAS with outdated strats I guess is pointless, my point of the TAS was that its time on that can be even lower. Play the game yourself with WR strats instead.

CrankykongCrankykong likes this. 
  Rayquaza911Rayquaza911
(edited: )

@kmrdkmrd missed your post but yeah makes no sense splitting the game for every platform this game exists on, run with what you have, or the fastest. The SGB is proven to run games ~2.4% faster, but its time is converted.

Based on the video above, the AC version seems ~1.6% faster in terms of timer only (if that's conclusive to everyone, I did show in my game footage the SGB2 timer was actually faster for the beginning). I don't know what exact framerate the AC emulator runs on.

So what does anyone suggest? Do you all want to increase all AC versions times by ~1.6%? That would put their current WR time at 18:30, which is still the record.

PS: Also deleting your run on the board as protest or w/e just because someone is in disagreement with you, is not the way to go.

 
  KomradeKomrade

I'm just saying, I thought running fastest version was a normal speedrunning thing, at least that's how most leaderboards used to operate. I'm not sure when that changed. It's not like we're talking about a large leaderboard here, either. The game has had 6 runners in 6 years, 2 of which are active, only 1 of which is running the AC version. If there was a time to divvy up a board, it's not now, imo.

 
  Rayquaza911Rayquaza911
(edited: )

From what I've seen it only changes when someones WR is lost. The DK NES game split emulator and console because they found it "unfair" that emu could get some RNG more consistently, even though it was doable on console too (it wasn't that it was faster, it was just less reset heavy). They later put them back together.

 
  KomradeKomrade

I'm sure that has something to do with it, but any reason is good enough to actually explore a new version. If it's found to be faster, I figure that'd be a good thing. I mean how many "retro" leaderboards have multiple versions of a game running against each other? Seems the AC version was dismissed before anyone has done any proper testing, which I'm glad to see is actually being done some what now.

 
  Rayquaza911Rayquaza911
(edited: )

Yes, and the same goes for Belmont's Revenge, I'm sure this will come again if someone gets the record on that version as well.

 
  KomradeKomrade

Well I'm sure this incident will cause someone to look into belmont's revenge at least. Probably before and exchange of WRs. CV1 community out right rejected the AC version so I'm not exactly motivated to do the testing myself lol.

 
  Rayquaza911Rayquaza911

Hmm well it's approved now reading the hardware rules so I guess they changed.

 
  KomradeKomrade

Oh, well, the AC version is accepted, but only if you run the 1987 NES release of cv1, not the 1993 FC release of cv1 (the japanese version). Problem is, the NES version is PRG0 which will crash from sprite overload. So I've tried being an advocate for running the 93 FC release from the AC, but it's still rejected. Of course the only reason it's rejected is because of a gentleman's agreement between cantaloupeme and myself for a race back in 2017 lol.

 
  MazzinMazzin
(edited: )

ok guys, so i finally found my way into this conversation too^^
first thing i wanna point out is that my TAS run was basically my first step into some sort of competitive approach to this game and by the time i made that TAS i was barely even aware of the lag issue, only because my previous TAS didn't even come close to the fastest time then it made me aware that there has to be something else other than gameplay that i had to improve...
long story short: my TAS run is far from optimal! it even got rejected from TASvideos leaderboard despite being the WR atm. so what im saying is, that you should not use that TAS to compare anything regarding lag optimization etc.
and speaking of my current best run on AC version, im almost certain that this version runs slightly faster so to me it would make sense to put a factor on it so that the time can be compared properly with other versions. the only question is, what exactly is that factor? i would be completely fine with a recalculated time, i would just be glad if there was an accepted agreement on what factor is used.
(also i would have to change all my posts and video titles etc to the new timing, plus the running timer within the video would show something different... hmm)

PS: it's great to have a forum here finally^^

PPS: where is Laxxus' run?! why is it gone?

 
  xenkaroshixenkaroshi

@MazzinMazzin he removed it himself in protest to the inability by the mod to hear the community. Personally I did have plans to run this, I was practicing it at one point some weeks ago. The thought that AC would be mashed in on the same board with old OG GB runners is a slap to their faces. "Run the fastest version" is something I absolutely can get behind, but at the cost of telling F U to everyone prior? AC shouldn't be banned by any means as its an official release, however I don't deem it justified to completely ignore everyone whom ran the game prior its release.

It is a small community, however 2 of the leading runners have adressed their concern and I don't believe there's any doubt to say the AC version runs faster. If someone can find a way to correctly convert it to a fair result to compete with GB, that's one thing. Right now we just aren't there yet, as such: The easier solution is to separate it and/or just just make an AC game page, we already have the Konami collection https:/​/​www.​speedrun.​com/​konamigbv1

 
  KomradeKomrade

The Konami collection boards arguably shouldn't even exist. Lest the argument become that every single version of any game gets its own dedicated leaderboard. As for the "Run the fastest version" thing, it isn't saying F U to everyone that ran prior, it's just the natural progression of competition. Unfortunately, old records get beaten all the time. If someone has their run bested, they're free to try and claim it back. No one is discrediting older runs. If the same logic were applied to strats/skips/glitches (something that is far more common and time saving than differences in lag across versions), then you would be arguing that beating an old record because of a new skip is telling F U to everyone prior. I'm not talking about huge skips that warrant category separation either (eg any% vs no ACE, etc), but rather just little things that can save time across an entire run.

I just think leaderboards of old games should embrace change and recognize that ten years down the road newer versions or ports could be what newer runners are picking up.

 
  xenkaroshixenkaroshi

You are correct, the konami collection board should NOT exist. However my point is that it does because people took issue with that version of it. Same as they do with the AC one. If we wanna avoid creating a new game board dedicated to AC, I suggest handling it better this time by still listening and respecting the community regarding the differences of the releases. Nobody says its impossible to slam the two together in the future IF we separate them now, more tests needs to be made either way.

Regardless, it is clear its faster much like the Konami Collection version was. The only question now is how the community will handle it this time around. A new game page was certainly a way to handle it last time..

 
  MazzinMazzin

@xenkaroshixenkaroshi
i would love to see you compete!

btw we could just add an AC notification to the leaderboard, just like all the other emulator options get displayed there.
to be honest, i purposedly used the J version to somewhat set it apart from the rest of the board since i couldn't pick an AC marker when submitting my prior WR. please someone add this option!

xenkaroshixenkaroshi likes this.