Posts for FatRatKnight


Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
As for my experience, now that the submissions are closed, I never did any TASing groundwork, as I simply couldn't rekindle my desire for TASing. I did state I considered myself retired from TASing, and drifting interests has firmly asserted itself in my habits. That said, I did analyze some bits of the game and threw together an enemy hitbox script. It was a lot easier to guess some positioning related stuff for my teammates to work with from there. Hopefully our discord makes it easy enough to find.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Admittedly, my charisma is rather lacking in trying for help in this game. The fact I'm considering myself retired from TASing probably isn't helping matters. Anyone who knows a thing or two about botting would really help in the first fight. I know lua, but I don't know botting, I just can't seem to tie the concept together. After that first fight, everything else can be tried manually for reasonably good results, as none of the later fights are complicated. I can probably coach anyone interested, the routing won't be complicated. Just that the first fight is... a mess.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
It's quite a bit of planning on just the first fight, I will admit. After that first fight, however, it's pretty smooth. Well, relatively speaking, anyway. We just slay our second fodder in the second fight while mowing through those kobolds, then just waltz on through with luck manipulation a little more restrictive, but never too much dice rolling needed to win. Essentially, the route is: * Plan out the first fight like a freaking champ where everything comes together in a horrid luck manipulation storm of planning. * Lose second fodder on simple fights to the Golden Hammer Inn * Get Stone Key and replace Wobby the Wizard with Knucklebones, who has more EXP * Get to 3F and learn Snilloc's Snow. Swarm and Eagle's Splendor * Sweep living groups with Snilloc's Snow. Swarm * Get Magical Cape and use that to glitch our stats * Sweep undead with Charisma boosts * Sweep Albrik fight with Turn Undead, Snilloc's Snow. Swarm, then strength boosted attack of opportunity for Albrik himself * Sweep later floors with Fireball once that's available * Sweep Hell Hounds with stabs or Lightning Bolt, the one weakness Fireball has * Seriously, we just sweep through everything in a hurry. Minimize encounters with our routing! * Defeat the final boss by having him generate with an HP overflow and a single low damage hit, and the undead bodyguards by Lv7 Cleric's Turn Undead None of the rest of this stuff is as complicated as fight number 1. We'll need Sleep to speed up some early fights (it's an overpowered status spell), but after a few levels, it's Snilloc's Snow. Swarm then Fireballs all the way, and producing a path where we fight the least number of encounters to reach key points.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Thread #16296: Eye of the Beholder GBA There exists an Eye of the Beholder thread already. The project is dead as I am no longer interested in doing TASing myself these days. However, if anyone is up to manipulating luck to get the first fight going how I envision it should, I will give my best support in hopes that it will revive. Turn Undead is stronger than it really should be. Instant killing entire groups of skeletons at a high chance? My strategy involves just charging forward through just the required stuff with a Cleric and a Wizard, and killing off two sacrificial party members to get the required party of 4 down to 2 for EXP reasons. I did find we get to Level 3 at a great time for Wizard spells (Sniloc's Snow. Swarm and Eagle's Splendor). Once we can reach certain magic items, we can then glitch our stats to the moon. If you're curious about my "first fight" plan: * Use a default Cleric and default Wizard, plus two sacrificial Wizards with low CON / HP * Manipulate Hide to work with our entire party. * Have two kobolds act while party is hidden, ideally because of Initiative rolls. This instantly ends their turns without animation. * Let one sacrificial Wizard kill a kobold, then move in order to die from two attacks of opportunity at once. (Kobolds needed to act to allow these opportunities) * The other three party members do their killings * The last kobold dies because of stupid movement AI and triggers an attack of opportunity, probably from the Cleric * The ordering of these actions are not set in stone If that "perfect first fight" can happen, I will be happy to see the results.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Looks like we've got competition now! Five teams out of six are now nameless! They're copying my winning strategy! Now everyone's saving frames on name entry! For those curious, just glance over the prior DTCs. There's a strong correlation between winning teams and short names. It's clear the meta now is to go nameless teams, such a boring strategy, no variation.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
I have been roped into joining a team by moozooh. Whatever team that includes this member, I accept. To anyone else on the team, I consider myself a retired TASer at this point. Make sure this museum piece stays in mint condition, someone else might want me for an exhibit. I can't guarantee much TASing effort, but I'll try to keep analysis and organization in top shape. And some scripting, perhaps.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Here, now I got our stage 1-2 here for you. While browsing through our Discord, I realize I have one weakness I can't overcome on my own with my habits: Handling images. Don't bother asking why, just I'm not sure how to go about dealing with them at all. Attempting to get me used to doing so will be met with stupid amounts of resistance from my own personality. Anywhere images are concerned (like, oh... our maps with routes drawn on them), they will be lost in my migration attempts to here. This is where help will be mandatory, and no amount of time on my own will fix that. In the meantime, rediscovering some tricks reading over old things is fun. What is BackJump, anyway? It's a deceleration abuse involving tricking the game into thinking we're headed in one direction, thereby having us "decelerate" to a higher speed threshold during the landing animation. Understood none of that? Good, because I'm still picking apart some of the intricacies of our tricks. At the least, I can go into details. The player's direction is stored in 0A1E. If the player's D-pad is pressed in the same direction as the player's momentum, this is updated to match the direction you're headed. For some reason, Up+Left is the only exception to this rule, in which any direction of north, northwest, west, southwest, or south is updated to be northwest if you press Up+Left. If the D-pad is neutral, 0A1E is also updated to match your momentum. When landing from a jump, the game decelerates Zeke/Julie based on 0A1E. The deceleration is not affected by speed caps, because it's supposed to decelerate you, not speed you up. From full speed south, we can set our direction to the northwest easily because of the weird exception. We can then hold Up+Right to start moving east without losing our northwest direction. With most of the momentum to the east, and a little south, you can alternate between Up+Right and Down+Left in order to trick the game into keeping your direction as northwest. Once we land from the jump, at close to our speed cap east, the landing will magically ratchet our speed up faster to the east. I suppose I should also talk about the speed cap. The game has your speed and subspeed. When asking if you're going too fast, it first checks your subspeed before checking the integer speed. If your subspeed is below the subspeed cap of your current movement, the game shortcuts the check and says you're good. If your speed cap is 2.50 px/f, and you're traveling along at 3.23 px/f, it just looks at the .23 part and decides you're under the speed cap after noticing it's below the .50 part, despite the integer being well above where it should be. The BackJump is abusing the deceleration direction to push the fractional portion up enough to get us to reach the next integer, then our subspeed is well below the subspeed cap and the game says we're good. At the start of 1-2, this is what we do to get our speed above 4 px/f. We choose this speed because the camera itself doesn't go faster than 4 pixels per frame anyway, and most of the things we want done are limited by the camera. From there, we can pay the 120 frames it takes to get into Sprint mode while outracing the camera. Then we jump as soon as Sprint is triggered, splash into some water, and trigger Perma-Sprint and dash around 1-2 at a crazy pace. Ah, I need to figure out the mechanics of Perma-Sprint now...
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Been dealing with a sickness that's keeping me down through fatigue. Also heard it's going to last around 2 to 3 weeks, so that's a pain. I'm already through a lot of the supposed time, so hopefully I'll be more energetic soon. I'm already practically guaranteed full recovery, just the annoying fatigue and coughs to go through. Here's a 1-1 upload just to make sure it's around. I put my description with the userfile. I'll be trying to put effort in getting one upload per day, if nothing else. It's not as though things are clean in the Discord, despite past attempts, so it might be a bit before I can track down our 1-3 attempts. I'm still hoping for assistance, but even without it, I'm going to make sure the knowledge base and TAS are available over the course of several weeks. The requested help will ensure all this happens sooner than the projected time.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
In case anyone's wondering since the contest, there was a burst of progress shortly after its end and... nothing. We're carrying a lot of information in the Discord channel, but that's not very visible here. So I'd like help migrating everything in there to here at TASVideos. This should help give the little TASing project a bit more life than hoping I somehow get my act together. My TASing interest has been dead for a while, and still has yet to be resurrected, so I'd rather get what we have out here than forgotten in some unknown spot. Regardless, I'm not sure how much I can put together on my own for the migration. I'm not interested in TASing, but I am interested in keeping all the info intact. So, this is a call for help on anyone still about on the project before I start attempts myself.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Reading the description on the player choice and letter-based upgrades, while the CPU cheating in letters is true, that wasn't my primary reason for selecting P4. The actual reason is that the letters ask for particular players. The letters on the track can ask for a particular slot, and each slot generates a random letter based on what some player doesn't have. The slots are as follows: (1) Based on Player 1 (2) Based on Player 2 (3) Based on Player 3 (4) Based on Player 4 (R) We read the race number, then select a player based on that (?) We ask the RNG for a player number Then it's a matter of figuring out what slots each race uses. Race 1: (1) Race 2: (?) (2) (R) (3) ... The Race letter picks P2 Race 4: (R) (4) (1) ... The Race letter picks P4 There are eight possible letters out of our PRO-AM II to get in the first four races. When I selected P4, I just wanted to abuse the letter selection at race 4 to guarantee the two generated letters are ones we didn't pick up, as the P4 slot and the Race slot both ask P4 for the letter. Which basically meant the only manipulation necessary at that specific race is to ensure I don't double up on the same letter, apart from I if my route didn't get any of that. Race 6: (1) (2) (3) (R) (?) (?) ... The Race letter picks P2 Race 7: (4) In the case of Race 6, while two letters pick the Random slot, they can't be independent of each other, as they're on the same slot. Which means those two letter boxes are tied together no matter what the RNG said. With only 7 letter pick-ups in these two races, the only way for the upgrade to happen at the end of 7 is for the Random slot to pick I, as that's the one letter you can double up on. On the randomness scale, I is bottom priority for letter generation RNG, so that was a bit of "fun" manipulating that, along with every other letter, when I was fresh from the first upgrade and the other CPUs didn't have a lot of cheated letters. I will point out that, by picking P4, Race 6 could not generate the letter P. The Random boxes are already locked into the need for the I, and the other boxes all depend on the CPU players, who each cheated their own P by now. That left RO-AM II as whatever is left for me to pick up. Once at Race 7, as that letter was asking for P4, it was just a trivial 50% chance of generating the one letter I was missing, another advantage for picking P4. The other letter it could have generated is the I, which is the fallback case for when it couldn't generate any other letter. So, for the most part, P4 was selected not on the CPU cheating their own letters (with P1 cheating at races 1 and 5), but on the maze of letter pick-ups the developers decided on. Oh, and fix the link to the forum post under your tires explanation. You included the period in your tag. As for other thoughts that aren't on the technical reasons why I chose P4 on RNG... Well, I can't think of a lot to say already. I'm actually not all that good on encouraging descriptions anyway, I claim. It is interesting to note that the bot can beat me on the comparable races, so it's clearly doing something right. I'm curious if it can optimize my early races a little better, with particular constraints of what it must pick up. The bot appears to be pretty robust to have beaten me. ... Sorry, there really isn't a lot I can think of to say. Um, have a thumbs up in writing from me? Again, I claim not to be a particularly wonderful speaker of encouraging things, but just throw me at the technical details or explanations and you'll see what I write.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
How did I get my numbers backward? At least my video time was right, even if a bit off on the frames. As far as authorship goes, I don't consider that amount of input to be a big factor. If anything, I'm pretty ambivalent on the credit. If you think enough of this TAS is based on my work, enough to deserve an extra name in the author list, go ahead. Regardless of the decision, I'm still enjoying seeing the effects of my past, and I get to see another of my runs obsoleted. I absolutely don't want to stand in the way of any progress made. Sorry I wasn't around to help during the production, but again, drifting interests has me elsewhere now.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Indeed, a message on GameFAQs got to me. My interests have drifted over time, so it took a PM there to find me. I do have one more thought in mind, beyond the one question about the terrain bumps in P12... It's about depositing fuel into the wormholes. Specifically, the amount of time the message screen is up. I never did figure out why the message locks us from clearing it for variable amounts of time. The comparison video does a great job displaying the effect of this difference on P13, on the first fuel pickup. (Good job on fiddling with the tether boost on the way there) Parts of my prior TAS is also visible in the comparison video. Well, it's pretty clear at the times the jetpod is moving in exactly the same way. That, at least, tells me the optimization I did then was pretty solid up until the two runs differ. Neat, my "really slow shot that blows up a turret" early in P13 is still there! As for the warp to P8, I'm convinced of some CPU timing problems around an interrupt that the code doesn't handle appropriately which allowed it to skip over P7, though I can't recall many details of my assembly diving. Good work obsoleting my TAS!
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
I'll give a status report: 1-1 : We're mostly following Team 8's route, but with a few minor tweaks early and somewhat larger changes later. We're planning on a perma-sprint setup late into the level off a bench while waiting on imps to handle a pair of victims. We have a short TAS that has dealt with these early minor tweaks, but that's about it. 1-2 : We use what we call the BackJump, where we can build speed (slowly) every time we land from a jump. We do this to get to 4+ speed to get the camera moving as fast as possible, then set up perma-sprint from that speed using the water to the north. This lets us collect a few more keys for free, as well as clip through an apparently hollow door to pick up an extra mortar. 1-3 : There was a brief mini-competition of sorts trying to optimize the snap out of this. I took a look at it myself and couldn't improve on what others did, so it looks fantastic. 1-4 : Team 6 had the right idea jumping out a window to the north. We have a partial TAS going through using that route, although it can stand to have a few others look. 1-B : We've disassembled enough parts of the code to understand the boss behavior! The boss acts based on HP thresholds and how many actions taken so far. We can abuse the game pause to slow down the rate the boss acts. We have two routes mapped out in the fight in case the theoretically faster one isn't viable. 2-1 : Some thought about perma-sprint setups here, but with so many doors, we'd lose it fairly early. It's still a bit before we have to hit the first door. Someone (me) tried to theorize some crazy zigzag route through many vertical walls, but it's determined we just can't go north fast enough at the horizontal speeds we'd be going. 2-2 : Apart from Team 6's lovely clips, we don't really have any particular analysis done here. One team did go with 2 victims, as victim #2 isn't that far away, and we don't need to waste time failing to save them with the boss right after. Whether that short jog is more than whatever is saved from speeding up a second victim in 2-1 is another question. 2-B : The boss is generally vulnerable the whole time, and moves with us, so no one bothered to do any special analysis. We're probably just going to center the boss on the death spot and manipulate the victims optimally. 3-1 : We did determine that something is very wrong with the pirate enemies around here, as it's capable of corrupting our inventory. There's been some analysis on how to get through effectively, but alas, that discussion is off on the general channel rather than the stage 3 channel, making it harder (for me) to track down. 3-2 : We have a full TAS here. Perma-sprint setup pretty much right from the start, where we begin the sprint north and bunny hop to the bridge to keep it. Because of camera, we'd have to optimize the setup or the keys right after to save any frames. 3-3 : We're planning an UpJump, then setting up perma-sprint while waiting on the camera on the victim. Other than that brief planning, we sort of left this alone. 3-4 : Basically, follow what Team 7 did, but faster. Nothing else comes to mind. 3-B : Well, it's another boss fight, where we don't have to worry about invincibility attacks, so I guess we just optimize the death position and victim grabs. 4-1 : Alas, I'm failing to remember a lot of what we did here. I recall a bit of talk about routing through it, but anyone else who feels better than me about it can chime in. 4-2 : We had another mini-competition to TAS this level, so we've got a nice run for this level. Turns out we can stand on the spikes, if we carefully manage our tile position correctly. I think we're pretty good on what to do for this level. 4-B : Most likely we'll take the fastest kill that gets around that annoying shield, and take the fastest victim manipulation we can. I don't know if we have good plans on centering the boss on the magic death spot, though. 5-1 : We have a TASed run here. Team 6 had the right idea about perma-sprint, but we decided it's faster to set it up right away. Perma-sprint is fast, but it can go a lot faster if you use the right input sequence. 5-B : Another boss, geez. Those jumps are really annoying, and the fight tends to lag as well. I'm guessing frequent pausing will be involved to adjust boss timings of these jumps. With the extra keys in 1-2, we're probably going to check our route to reduce the out-of-the-way keys on the way forward. As is said already, progress is a bit slow, but it does help to know what was done so far.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
I probably should say hi, at least. I mean, I did some stuff back in the day, maybe I have insights. Except it's been such a long time ago that I don't really have a lot to say. Turning Sei-Ryu into instant art was definitely not something we could do mutant-only. I remember feeling pretty happy with what I did back in the day, though now I wonder how much the added combat turns would have hurt. How was the luck manipulation on resets, though? We never did get a working bot together to handle that.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
A game of a decent length. Each level is mostly self-contained... ... but the player's inventory and saved victims carry over. Routing is mostly to the individual level, but needs to account that. Just enough physics, and fairly interesting glitches to go with it. The competition actually had some closely competitive teams with wildly different routes and techniques, and a pretty significant optimization gap when looking at a slower team, indicating the game wasn't remotely trivial. I'd say the choice for this contest was spot on. It's nice to know that I haven't lost all my cool habits being part of the top three, though respect for Mittenz for keeping up with me toward the end. Invariel had given some useful analysis as well. R30hedron feels bad about being intimidated by all the "mad scientist death robot physics" we were into at that point, but don't worry, it's easy to feel it's impossible to catch up if you haven't followed from the start. Stay around and read all the interesting things we'll keep working on, we ain't stopping ya! I'm wondering if a contest can be held on a game with 100% self-contained levels. Alien Hominid had come close, except the RNG carries over. I had managed to get sync across pretty much my whole TAS when I carefully tweaked the RNG right in an earlier stage. The idea basically being that after an initial start-up period, have the teams provide the next stage every 2 weeks or something. It would make things a bit less silent over the course of the contest. Well, it's an idea.
Post subject: Re: HOLY H*CK, RESULTS
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Samsara wrote:
If there's anything to take away from these results for DTC11: Sign up late and don't give your team a name!
Saving frames on name entry is a TASing standard, you know! Team 6, it would appear you've done very well. Take this win for yourselves and know you've beaten an old DTC champion. Someone will want my script, though I'm not too much in a mood (sleepy) to type a whole lot. I did end up using this contest to see where my interests were these days. There were some highs and lows during this time, though I've found that I don't feel fully ready to return to TASing. Hey, I've put in a fair amount of effort, my teammates will definitely say I've helped to push out what we did. I'll be around to assist on explaining what we did find, but first, I need to be less sleepy. I haven't looked at the winning run yet, though I suspect I'll find what I would call Dash Mode acceleration abuse, BackSlides, and UpJumps. There was zero attempt at boss behavior manipulation in my team's TAS, by the way.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
We teams have been pretty quiet during the closing hours of this contest. I mean, we should be posting movies of runs that forget frame advance and save states were a thing, especially where there's nowhere near enough time to analyze the sloppy gameplay to tweak the winning movie. Or talk about tricks and mechanics that don't even exist in this game. Geez, everyone! You should have put all your points into Strength for the lengthy boss fights! EDIT: If this all happened on TASVideos Discord, and not on this board, then disregard this thought. I wasn't there for all that cool stuff. Whatever the results, congratulations to the winning team. If we win, then I'd be really egotistical to congratulate myself, but I'm so going to congratulate them winning team regardless!
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Samsara wrote:
Since I was asked about this and it is growing increasingly relevant by the day: The earliest point to end input, and the point where all runs will be compared, is entering the door after the final boss.
Samsara wrote:
REAL ACTUAL TEAM NAMES: 8. R30hedron, FatRatKnight, Invariel, Mittenz - "[final boss music begins to play]"
The final boss door, you say? But what if we are the final boss? Ah, right. Some other team just needs to beat us and enter the door to win the competition. And beat all the other teams as well. But someone must beat us for us not to be in 1st. Also, while exploring Ghoul Patrol, I've discovered the high score name entry shows up when you get a Game Over, but it's completely skipped when you reach the credits. I feel cheated -- We don't even get the name entry when we defeat the final boss! How is anyone supposed to know what we'll do when this contest ends?
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Score attacks are allowed as an objective goal now? I remember putting effort into getting good scoring rules down. I've not been following much happening with the site recently, but the score attack being listed as one of the criteria came as a surprise to me. No idea if that thread would help define any rules around scoring TASes, but it's a piece of history I remember.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Now, now. Don't encourage everyone to put me into first place. I've already gotten there plenty enough for you all. Though I won't exactly make it that easy, so it'll be some sweet stuff at the top if our team misses first place this time. As for Mothrayas' prize, I will be handing it off- what? Non-transferable? Well, my portion of the prize will go unused. So extra encouragement for you all to win, as there will be one unused emote in the TASVideos Discord if I win.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
I will say, from my prior analysis, the game's RNG looks RTA-viable to manipulate. To the extent that a TAS can't do any better, provided the RTA doesn't make mistakes like stepping wrong or some nasty mis-menuing. Really, the trickiest bit is starting the game on the right second for the clock seed. Basically, I'm wondering if any RTA strategies made use of RNG manipulation. ... I'll be watching the run, I'm interested.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Here to make ya sweat with a last-minute sign-up, of course! Perhaps I should seek a somewhat less dramatic entrance the next time I disappear and come back. Actually, I'm starting to wonder if I should stick around to be on a team. Well, I am sort of throwing a wrench into your carefully laid out plans, when every team is kitted out with four people each. Whatever everyone thinks is fair, I'm also perfectly fine backing down (and making my own TAS of the DTC-chosen game in secret... er, you didn't read this thought, so it's still secret!).
Post subject: Added: Stipulation
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
SNES? I'll bite. I'm up for grabs on any team. Or maybe my past victories gave me enough ego to count as three other people. (Considering the time left, assume I will confirm any team assignment - EDIT: First come, first serve. I'm not sure I'd have the time to pick from multiple offers, so only the first suggestion should be assumed to be taken!) As for my disappearance, just chalk it up to drifting interests. We'll see how much of my TASing habits I've kept.
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
Aran Jaeger wrote:
[...] Also for TASes that go by how soon the last frame with ''non-trivial'' input appears ([...]), the ''no input state'' also being seen/interpreted as counting input for the length of a TAS would deem every in this manner optimized TAS to be unable to ever finish, effectively infinitely long on-going, and would mean all such TASes would be instantly unacceptable for TASVideos as the input file in that sense would keep having input for ever and ever as the input files (and associated videos) would never end.
Absolutely yes on this detail, on the interaction of my viewpoint of what constitutes an end of the TAS, and the exclusion of all other possible metrics apart from input length. To get my viewpoint to work at all, the length of the input file will always be considered infinite, and therefore the metric imposed will always fail as we're unable to compare infinity. That therefore means the interaction of this metric and the measurement to game end is inherently flawed. As for the rest of the post, the take away I'm getting is simply: * TASVideos is a site dedicated to perfection of video games by removing the human element. * The non-null length of input can be trivially machine-verified. * The definition of the game's end is pretty arbitrary, pretty "human", in what makes it get to the big win. * * As an aside, some runs beat the final boss as simply a part of clearing their target goal, not the final endpoint. * Without a hard machine fit-all of determining the game's end, we can't remove the human element of determining when an ending begins. * The other metric of non-null input is what's left. Therefore, my preferred method of determining when an run ends (time to game's ending) is inherently mired in human limitations of what determines the ending. I don't see a problem with that. Most of my arguments really just boil down to human interactions. After all, we're making machine-perfect runs for us mere humans to enjoy. The human can readily identify when an ending happens. Or at least a good enough one. It's not up to the machine to determine when we perceive the ending. The machine can always determine the differences of lengths of input, and when all that's left is an infinity of null input. So the only (obvious) machine method of measurement would be these lengths. In the quest to remove the human part, this is what's left. Let's move on to the TASVideos tier system. We have Stars and Moons, and then we have Vault. The Vault category is the most rigid tier, as it doesn't depend on the human notion of entertainment. If the run is Stars or Moons compliant, then we're already in human territory, so judgement can be based on what ends the run "better," for whatever definition of better is to be used. For Vault, which strives to remove the human element as much as possible, I am left with a question: Is "time to ending" valid for being largely only measurable by humans? Normally, the difference between shorter input and quicker ending is trivial. Usually, most of the TAS in question remains intact, except for the last few seconds at the end. In the case of Monopoly here, the strategies are apparently widely different. Are we in Moons (or Stars) territory? If so, then we can side-step answering the above paragraph until another submission. All that really needs to be answered is whether the run appears better than the previous. = = = Unrelated to the argument, I enjoy seeing the CPU lose before they roll their dice for the first time. I'm always disappointed when there are after-ending activities like the high score naming, or even mini-games in credits, that are entirely ignored when the TAS finishes. After all, null input is still input...
Editor, Experienced Forum User, Published Author, Skilled player (1158)
Joined: 9/27/2008
Posts: 1084
My own personal view is that null input is still input. Fact is, even if you drop the controller, you are still feeding empty input to the game. I'm decidedly for the idea of getting to the ending faster. Despite this, I still ended input early for R.C. Pro-Am, as it ended the run humorously. My April Fools' Guardian Legend TAS took default input state in a different direction, by having a button on auto-hold rather than having it affected by the movie file. If we're waiting for null input, that will never happen, but the game ending does take place anyway. Unsurprisingly, the TAS was rejected (and misses the Vault criteria in at least two ways), but I feel the statement has some merit regardless. As an aside, the Dream Team Contest #6 could have a different winner depending on "ends input early" vs. "fastest game ending." My team won, but if it was ruled for fastest ending, our setup to get the boss to kill itself on our permanent damage box would have been worthless, and we would have been in second place (based on some memory of calculating things a while back). Looking back on this, with my outlook, I do feel like we deserve the #2 spot, but it still provided a nice demonstration of a glitch in the end, so I feel positive overall. I thought a bit of my history may add a little perspective. It's mostly been author's choice from what I understand. Input length is easier to measure, as it's practically automatic with the size of the file, but time to ending doesn't care about semantics of what input is.