Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
DrD2k9 wrote:
If not maxing out the distance on the last round is 3 frames slower than this submission, then don't change it and make a longer overall run. We obviously wouldn't want a longer any% run.
Sorry, I meant to say that the movie would only be 3 frames shorter, not slower, if the 416-yard longest drive goal was taken away. My apologize for any confusion brought on by that.
DrD2k9 wrote:
The game also "very clearly and unambiguously" displays each round's drive distance on the Score Board screen, so I don't feel that only the round score being shown on the driving screen is a very strong argument as to why distance shouldn't be considered.
Yeah, I will admit, the score board does "very clearly and unambiguously" display each round's drive distance on the score board screen, in addition to win count. But that doesn't necessarily mean drive distance ought to be considered in the scoring metric. For instance, baseball score boards track all kinds of metrics besides runs which include hits, errors, balls, strikes, and outs, and I think we can all agree none of those latter parameters should contribute to a final score tally in any sense. So I don't feel that drive distance merely being displayed on the scoreboard is a very strong argument that it should be considered.
DrD2k9 wrote:
... I'd still feel that [drive distances] would be a better metric than simply rounds won. Mainly because, maxing out the drive distances leaves nothing else to maximize for the game; where only winning all rounds (with less than maximum distances) does leave something else that can be maxed out.
I guess let's first ask the movie rules what the "maximum score" goal actually is:
Score Attack: Aiming for maximum score while otherwise completing the game as fast as possible.
Well that's literally all it says. It gives no mention that you must include any sort of ... secondary measurements of in-game performance in with your maximum score, which is really what the yardage is. It's just a tool the game uses to assign the points. Lumping yardage into the score equation simply because it is "something else that can be maxed out" is beyond the scope of what the movie rules say "maximum score" actually is. Take NES Tetris "maximum score" for instance. We can all agree it sure did max out the score but let's not forget there's also a whole bunch of other metrics on screen as well. Should the authors have maxed out the line counter as well? The level counter? There is certainly "something else that can be maxed out" in many, many maximum score submissions on this site. And my argument is just because there's something else you could max out doesn't mean you must.
Let's also not forget about Olympic-style games like NES Track & Field "max score, playaround", where many events give you three attempts, saving only the best score/time. It appears to be common practice in Olympic-style TASes to max out on the first attempt, and simply foul out or get a low score in attempts 2 and 3 to save time. Why is that unanimously seen as acceptable TAS practice? I mean clearly, the score on attempts 2 and 3 could have been maxed out as well and be perceived as an overall higher performance on the scoreboard. But I know what everyone's counter-argument is. At the end of the day, attempts 2 and 3 don't matter because your final score at the end of the event is equal to your best. And that's what I'm saying here about Long Drive Golf. Whether I win the round scoring 416 yards or 342 yards, it doesn't matter because the final score awarded at the end of the round will be 1 regardless.
Lastly, page 17 of the the plug and play unit's instructions manual explains Long Drive Golf. Of particular relevance to how the game defines "score" is point 6:
The computer keeps track of each player's score and the distance of each drive on a separate scoreboard screen. Following each round, the player that hit the ball the furthest will be awarded 1 point. At the end of 11 rounds, the player with the highest score wins.
The first sentence goes out of its way to imply the "distance" is a separate thing from the player's score (i.e., the win count).
Regarding the last sentence, if the player with the highest score wins, that must mean total yardage is not considered score ... that's cause you could play a game where you get a higher total yardage than your opponent but still lose.
And I know somebody out there might say "instruction manuals shouldn't count as evidence of anything whatsoever because they're a separate entity from the game", or whatever. But instructions manuals have been used to support judgements in the past. This submission in particular would have been accepted had the instructions manual told the player how to perform the cheat which the TAS used.
nymx wrote:
If you still don't agree, we can certainly discuss this among our staff.
Please do so, I believe I've presented all the arguments I have to the table.
TL;DR I maxed out the win count because it constitutes the score. I did not max out yardage because it is not the score. Therefore, the "maximum score" goal is met.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
nymx wrote:
Can this submission be reworked to salvage your movie?
Yes, very easily. I can remove all discussions from the notes regarding maximizing drive yardage and define score solely based on # of wins. The movie would just require a minor tweak at the end to not max out the 416-yard longest drive, and would only be 3 frames slower shorter. I agree with most of DrD2k9's points, especially that I admittedly went for a weird amalgamation of scoring metrics. However, I would argue # of wins is a better metric for basing "maximum score" on than total yardage. This is because the game very clearly and unambiguously labels your # of wins as "SCORE" on the driving screen. And that was my main intent with this "maximum score" submission. I think if a future submission wanted to max out total yardage the branch should be called "maximum yardage" instead.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
I don't think anyone is tackling improving those TASes right now, though there are many known improvements that have been discovered but not yet reported in a centralized manner. Maybe I should do that one day in this thread. Anyway, Favini's main focus seems to be on TASing the ROM hack Kid Chameleon 2: Heady Metal's Revenge, and a TAS of the full game was being worked on for a while, but I believe the project had to be put on hold for a bit due to the immense amount of work involved. I don't blame them, optimizing movement in this game can be a real pain.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
The Ooze
I believe this is the third time it's been requested in this thread but I can't believe there isn't a TAS of THE OOZE yet!!
The game is so freaking hard and totally deserving of the TAS treatment.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
Game & Watch Galley 3
Ever since seeing Darkman's TAS of Game & Watch Gallery 2, I've wanted to see one of Game & Watch Galley 3 as that was the game I grew up with.
Bugs Bunny in Crazy Castle 4
Similarly, Alyosha and Brandon's TAS of Bugs Bunny: Crazy Castle 3 renewed my interest in seeing Bugs Bunny in Crazy Castle 4, as that was also the one I grew up with. I was working on one for a while but have since stopped completely. While I don't ever plan on returning to it, I thought I should link my progress should anyone want to pick up where I left off. I managed to TAS all the way to world 4-5: Download the WIP here
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
I stopped working on this project years ago and while I still don't plan on returning to it, I thought I should link my progress should anyone want to pick up where I left off. I managed to TAS all the way to world 4-5: Download the WIP here
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
I was wondering what, specifically, the latest movie rules mean by "adding gameplay" when pertaining to the use of in-game codes.
In-game codes that add gameplay are allowed for a separate branch, as long as such codes are used optimally. Codes that modify or disable in-game mechanics are not allowed, unless they unlock an in-game item or a skill that does that.
In particular, I want to talk about things like "final boss" and "credits warp" codes, as while almost all instances would remove significant amounts of gameplay over their full-game TAS counterparts, they often times do bring some new gameplay to the table. Here are some examples of what I'm talking about:
I submitted a movie to the site about 7 years ago for the Kid Chameleon "plethora warp", which requires a specific combination of inputs at a specific place in the 2nd level to the warp to the final boss. This was rejected simply for using a debug/cheat code. Clearly however, in-game codes are now allowed, but does this qualify as adding gameplay? While the cheat cuts out much of the game from the current publication, 52% of the gameplay in the video itself is new, as all gameplay after collecting the Red Stealth helmet in the second level is different gameplay you would never see in a full-game TAS.
I found a 1-frame improvement since making that submission so here is the updated video:
Link to video
Additionally, there is the Vectorman "credits warp", which is a code you input at a specific spot in the 1st level to warp right to the credits. Again, technically 64% of the gameplay in the video itself (not including the title screen) is stuff not seen in the published TAS, although significantly more gameplay is removed than is added overall.
Link to video
So my main question is what exactly is meant by the movie rules when the in-game codes must "add gameplay"? If a particular code also removes gameplay, how much must be added to offset what as removed? Do the examples I listed even count as adding gameplay or is this just modified gameplay?
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
Bagel Brothers is done, and another 10 seconds of time save has been found for both TASes! Incredible! Though the video enters the level as Berzerker, you would actually enter as Cyclone in the full-game TAS. Not to worry though as there is a hidden Berzerker helmet near the start.
Favini also describes a time-saving technique I've never thought of nor have I seen anyone else discuss (edit: RandomPieIV has talked about it before). If you align the camera in a certain way when touching the flag (i.e., aligning the screen edges to the block grid), you can reduce the number of blocks on-screen and therefore the total time it takes for the screen to transition to the holodeck style. Even the types of blocks on screen make a difference—blank and background tiles appear to be removed at a rate of 7 per frame, while solid tiles convert to gray at a rate of 3 per frame. Favini saved 6 frames in this video just by taking all this into account. I calculate that at the two extremes (all-blocks misaligned vs no-blocks aligned) the holodeck transition can take anywhere from 40 to 105 frames.
Link to video
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
Freddi Fish and Luther's Maze Madness
This is the one Windows game I would love to see TASed that wouldn't be impossibly difficult to make with current TASing tools. The game can be run through ScummVM, making it possible to be TASed through the Linux platform. I've even already did a speedrun of it as well. This run does not use the game's built-in cheats called "Junior Helpers" which give unlimited bubbles and/or unlimited lives and, while they aren't hidden features, I think a TAS that doesn't use them would be far more impressive. Check out the video and description for more info if interested.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
Posting my list here since I am not good with Linux and am unable to TAS these myself. I have tested all of these games and they are fully-functional in the latest Ruffle version.
Super Mario Moto • My brother and I both thought this game was hilarious as kids. It's Mario on a freaking motorcycle riding through a level while doing outrageous flips throughout. Pretty hard game, too. I know changing the motorcycle pitch can affect the speed gained upon landing, so it may make for a flashy TAS.
Dutchman's Dash • Was a Spongebob flash game on Nick Games back in the day. I didn't get very far in it because of the game's strange momentum and hit detection at times. Would make for a very impressive TAS, as the existing speedrun is nowhere near clean but still impressive despite a couple of deaths.
Mario Starcatcher 2 • I did a speedrun of this many years ago. It is a very basic platformer and I think would be simple and quick to TAS.
Supra Mayro Bross • I did a speedrun of this one too and I hate to say it, but this is my most-viewed youtube video at almost 100K views now... I'm not kidding. Yes, the TAS would be trivial and silly, but it may be popular with viewers because heehee haha funny meme.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
Get ready guys, some improvements to the currently published TASes are definitely on the horizon. A YouTube user by the name Favini01 has discovered and TASed some pretty impressive stuff recently, the latest being that airwalking into a boss head as Berzerker while in the charge state absolutely decimates the thing. Check out their quick work on Shishkaboss below, which I would estimate saves about 15 seconds over the current maximum levels TAS. I suggest you subscribe to this guy as I have confirmed they will soon be TASing similar for Bagel Brothers. I'm anxious to see if the time save there is just as impressive!
Link to video
They also found that Skycutter can in fact gain speed from bouncing off enemies if you face backwards when the bounce happens. Even though the video below is from a ROM hack, I've tested it and the speed gain does in fact happen in the base game as well... although I don't think the base game has any situations that could yield a speed gain this... extreme xD
Link to video
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
Alright everyone breaking news, I just cooked up a giant 6-second improvement to this today! I know publishing is already underway so if it is too late to replace the file, just let me know and I'll make an entirely new submission.
Link: https://tasvideos.org/UserFiles/Info/638493564966079454
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
Mikewillplays wrote:
Question: Wouldn't it, in theory be faster to erase double inputs (two yellows in a row for example)? I know that it would be extremely unreliable, but it's still an interesting thought
The only way to change the input sequence is to press start on the title screen on a different frame, which severely limits your ability to manipulate such a favorable sequence. I'm sure somewhere way down the line there's a start frame which yields a sequence of 31 inputs which aren't doubled anywhere, but you'd have to wait on the title screen so long that the total movie length ends up being longer. Plus, I already calculated that the sequence played in this movie does indeed finish quickest when you factor in wait time on the title screen.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
Okay I have a 52-frame improvement to replace this movie with at: http://tasvideos.org/userfiles/info/69339170071981656
RNG was unfortunately slightly worse in all four stages this time around, and I forgot to mention this earlier that there is a strange framerule-like system that the game uses which limits the number of suitable RNG patterns to test. For example, in the case of the fourth stage, this was the only RNG pattern which made the fastest framerule.
[quote nymx]...but I'm pretty sure that a game like this doesn't make use of subpixels.[/quote]
It appears you are correct! I was using a RAM address which I thought tracked X subpixels but upon closer inspection, it does not. Visually, it is actually possible to see that the manipulation I make allows me to jump from the very edge pixel of the platform, whereas without the manipulation the rightmost I can jump is two pixels to the left of that.
I have already updated the TAS notes and temporary encode to reflect all of the above.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
[quote xxezrabxxx]In my opinion if there are multiple characters to pick from and they are all different in some way, they should be branched individually[/quote]
I think it would be more accurate to have separate branches for the 7 adventures than the 10 characters cause the adventures vary much more significantly (i.e. length, monsters and items, features, etc.) than the characters. In fact, I believe 9 of the characters' times would all be within about 2 seconds of each other. I arrived at that number because that is the total length of time the Waif had to stop for encounters and weapon manipulations, something a character with stronger weapons to start with could minimize. The other character, the Novice, is the only one that could save a more significant 16 seconds in Level 3 by using the double speed potion.
Going off this line of thought, I already made a TAS of Novice difficulty (not to be confused with the Novice character), and I did not submit because I thought it would be rejected for not being hardest difficulty. Bottom line, could the different adventures be separate branches? If so, I'll submit it!
[quote InfamousKnight]Oh wow, this one is rather long compared to the others. I watched it without the volume as I found that to be annoying. Nice run![/quote]
Thank you! It's funny how 7 minutes for an Intellivision TAS is considered long :P
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
[quote GJTASer2018] Even the longest currently accepted submissions don't use any kind of speed up in the way you're proposing.[/quote]
A skill level 6 submission of this would be remarkably different from those because it would feature static gameplay for LONG periods of time. It looks like those are all RPGs and one racing game where things are constantly happening, so a speed up would surely ruin those experiences. Not in the case of INTV Chess!
[quote Patashu]You might be able to find a faster/less enemy thinking win than what Stockfish can offer, since Stockfish plays as though its opponent is as good as it, while the poor Intellevision in reality is no match.[/quote]
I could see that. Stockfish probably made an unnecessary king move on move 24 and INTV made some really questionable pawn moves and the queen move on move 26.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
[quote creaothceann]What if the emulator is in Turbo mode?[/quote]
I was thinking more in terms of the encode that goes on YouTube. The fastest my PC can run BizHawk unthrottled is about 600% which is still over a 1h10m watch for me.
Although this actually makes me think: I've seen encodes published at slower speeds for TASes where the game is too fast. Has the site ever put out a high-speed encode for games that are too slow? If not, is say, a 4000% speed encode so that this would fit into roughly 10 minutes something that could be arranged? If so, I just might start putting a Level 6 TAS together.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
[quote Spikestuff]I absolutely dislike the fact that you used the easiest difficulty in chess instead of demolishing and manipulating a harder difficulty.[/quote]
Maybe I can try and convince you otherwise. Since an argument can be made that pressing the "Force Move" button means you are not playing at the highest skill level (page 12 of the instructions manual even states this), and as Patashu said skill level 7 would not be a vaultable category, it seems the type of movie you would like to see is the fastest checkmate on the skill level 6 computer without pressing the "Force Move" button. So I ran some tests:
Using only moves recommended by Stockfish 12 at 35 depth, the computer was checkmated after 31 moves and thinking for 7 hours, 12 minutes, and 54 grueling seconds.
Using only Leela-recommended moves, 7.5 hours were spent on the first 30 moves. On move 31, the INTV computer has so far not moved for 24 hours. I'm not sure if they will ever make a move... perhaps the computer doesn't know how to interpret that position.
So now, knowing how long it would take to demolish a harder difficulty, I think it's clear why the easiest difficulty was chosen: entertainment. Nobody wants to sit around and watch this game for 7 hours and 13 minutes (or encode that movie!). Believe me, I would've had the same opinion about picking the harder difficulty if the computer wasn't so damn slow.
Experienced Forum User, Published Author, Experienced player
(759)
Joined: 7/18/2016
Posts: 91
Location: United States
I've actually found a potential improvement to this. I uploaded an edited movie file which shows that the menuing at the beginning could be 21 frames faster by using the second controller to squeeze in more commands. Unfortunately, it desyncs after the second throw so much of it would need to be redone.
I would volunteer to help finish it myself but I am unfamiliar with the game's mechanics, nuances, and such. Hope this comes in handy!
