1 2 3
Posted: 2/28/2009 9:53 PM
Post subject: Which game is the hardest to TAS?
Quote
Warp
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
While re-watching the Gradius3 TAS I started to wonder which of the published TASes could be considered the hardest one to make. I know from experience that the Gradius3 TAS with the goal of maximal kills is rather hard. There are times when the screen is so crowded with destroyable objects that trying to kill them all is almost impossible (and, in fact, there are a few places where it apparently is impossible to kill them all because the game mechanics simply can't be stretched enough to do that). However, I doubt Gradius3 is the hardest game overall to TAS (which has been published). Of course "hardest" is a rather subjective measurement, and can in fact depend on what you count as "hard". The amount of rerecords per movie length ratio doesn't necessarily correlate with difficulty of TASing, as it may simply be a symptom of the author aiming for frame perfection. It might also be a symptom of one or a few places in the game which require tons of rerecords to get a perfect result, but overall the game might still be relatively easy to TAS (at least for an experienced TASer). There may also be different categories of "difficulty", such as: 1) Playing the game through as fast as possible is relatively easy, but an additional goal makes it very hard, due to its nature, as it stretches the game mechanics to the limit. I think the Gradius3 TAS is a perfect example of this. 2) The game is so difficult even with tool-assistance, that getting the fastest possible completion requires enormous amounts of work and rerecords. The total amount of rerecords might in this case directly correlate with the difficulty of making the TAS. 3) Getting the fastest completion time might not require humongous amounts of rerecords per se, but requires so much luck manipulation that it requires very deep understanding of how the game mechanics have been implemented (such as how the RNG works and what affects it). Often requires memory watching tools and even disassembling and studying the machine code of the game, or creating bots to grind through the hardest luck-manipulation parts. The TAS itself might not be enormously difficult to do once all the relevant information has been gathered, but it's this information gathering and possibly creation of game-specific tools which requires tons of work, and which makes creating the TAS very hard. 4) The game is simply so long that it takes a very long time to create a TAS completion of it. I suppose any of those RPG TASes which took over a year to make could count. The difficulty in this case correlates more to the perseverance needed to create the TAS. Maybe one heuristic for "very hard to TAS" would be if a newbie TASer would have little hope of competing with the current best submission, without extensive experience in TASing. I think it might be fun to give some kind of award or at least mention to the "hardest game to TAS" published. :)
Posted: 2/28/2009 9:54 PM
Quote
Comicalflop
Editor, Experienced player (734)
Joined: 6/13/2006
Posts: 3300
Location: Massachussetts, USA
Speedy TAS of 2009
Yoshi Island Wins. Topic done.
☣ Homepage ☣ Retired
Posted: 2/28/2009 9:57 PM
Quote
ShinyDoofy
Former player
Joined: 12/5/2007
Posts: 716
I've only played it on console ages ago and haven't cared about speed or the like - so how come this is your ultimate winner?
Posted: 2/28/2009 9:58 PM
Quote
Tompa
Any
Editor, Expert player (2215)
Joined: 8/15/2005
Posts: 1941
Location: Mullsjö, Sweden
NES TAS of 2018 SNES TASer of 2017 SNES TAS of 2017 NES TAS of 2014 GBA TASer of 2011 GBA TAS of 2011 SNES TASer of 2010
Games that randomly desynch are hard to TAS. The Super Star Wars games are three examples. There is also a lot of manipulation to be done. I can imagine that some 3D games are hard to TAS, but I have no experience from that.
Posted: 2/28/2009 9:59 PM
Quote
NrgSpoon
Joined: 6/14/2004
Posts: 646
Well, if and when Starfox 64 gets finished, you'll be able to give the award to that.
I like my "thank you"s in monetary form.
Posted: 2/28/2009 9:59 PM
(Edited: 2/28/2009 10:05 PM)
Quote
Comicalflop
Editor, Experienced player (734)
Joined: 6/13/2006
Posts: 3300
Location: Massachussetts, USA
Speedy TAS of 2009
I was hoping, in the ironical sense, to not have to post a huge essay in a topic started by a huge essay. -Only 3 submissions; 2 of them are the super short warped runs, but the any% is super long, and has yet to be obsoleted. -The 100% run everyone wants to see, yet still remains incomplete. -Every frame of ever second of every minute you must watch tons of memory addresses. -This. -Speed increments, but you can;t freeze it above normal like in SMW. -Acceleration is a bitch. -The shooting system of eggs is immensely complex. Aiming is tricky, managing eggs is tricky, making them ricochet perfectly is tricky. -staying at maximum speed is tricky. -The game is immensely complex. new routes are being found constantly. -massive rerecording behavior needed; even then, constant redoes are necessary. Lemme know if you need more. there's still plenty of reasons; I could go on and on. Starfox 64 doesn't deserve it because a run would have easily been done already if meepers and I weren't so lazy. I would like to mention that I think this topic sucks ass. The term 'hardest' is about as subjective as entertaining; there is not going to be a universal decision on what game is hardest. Also, I do believe that the site should not be about whining or bragging about how tough a game is to TAS.
☣ Homepage ☣ Retired
Posted: 2/28/2009 10:00 PM
Quote
moozooh
Senior Moderator
Joined: 8/4/2005
Posts: 5777
Location: Away
Do you mean "hardest to perfect" when you're saying "hardest to TAS"? There are many games like that, and most of them have something to do with thinking. Arkanoid comes to mind, simply because it's near-impossible to tell what positions and angles will result in an optimal solution. Scripts, bruteforcing, and memory addresses aren't of much help here. Yoshi's Island, on the other hand, can benefit from them.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Posted: 2/28/2009 10:11 PM
Quote
Warp
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Comicalflop wrote:
Lemme know if you need more. there's still plenty of reasons; I could go on and on.
Thanks for the explanation. This actually made me appreciate that run a lot more. I think I'm going to read that tricks page and watch the TAS again. :)
moozooh wrote:
Arkanoid comes to mind, simply because it's near-impossible to tell what positions and angles will result in an optimal solution. Scripts, bruteforcing, and memory addresses aren't of much help here.
Yes, Arkanoid is indeed one type of game which is really difficult to perfect due to its completely non-linear nature, requiring a lot more tactical thinking than bruteforce tool-assistance. It really excels in that category.
Posted: 2/28/2009 10:33 PM
Quote
Kyrsimys
Former player
Joined: 4/16/2004
Posts: 1286
Location: Finland
Having tried out YI, I can say I agree with Comicalflop.
Posted: 2/28/2009 11:08 PM
Quote
Mitjitsu
He/Him
Banned User
Joined: 4/24/2006
Posts: 2997
TAS of 2007 N64 TAS of 2007
Difficult game to TAS without making errors which would be obvious to a veteran TASer - Sonic games Difficult games to improve on - Mario games Difficult games to truly get frame perfect - Zelda + Metroid games Mainly due to the sheer randomness and lag, as well as the likihood of having glitches and sequence breaks that save noticable amounts of time.
Posted: 2/28/2009 11:09 PM
Post subject: Re: Which game is the hardest to TAS?
Quote
DarkKobold
Skilled player (1652)
Joined: 11/15/2004
Posts: 2202
Location: Killjoy
Arcade TASer of 2010 PCE TASer of 2010 Arcade TAS of 2010 Lucky TAS of 2009 Lucky TAS of 2007
Warp wrote:
3) Getting the fastest completion time might not require humongous amounts of rerecords per se, but requires so much luck manipulation that it requires very deep understanding of how the game mechanics have been implemented (such as how the RNG works and what affects it). Often requires memory watching tools and even disassembling and studying the machine code of the game, or creating bots to grind through the hardest luck-manipulation parts. The TAS itself might not be enormously difficult to do once all the relevant information has been gathered, but it's this information gathering and possibly creation of game-specific tools which requires tons of work, and which makes creating the TAS very hard.
Um, in actuality, all that stuff makes TASing easier, from personal experience. What you are creating you at least 'know' is optimal (to a certain extent), which reduces quite a bit of the frustration factor. You stated the initial start-up of a project like this is more 'difficult.' I heartily disagree. It does require a skill-set that most TASers don't come pre-equipped with. However, the start-up in a project as you mentioned is actually is the most rewarding aspect, since the results are very apparent once you've determined the necessary information to create the tools. I can't stand the guess and check of creating most TASes. "I guess I should throw a bomb on this frame." "I guess a delay of 10 frames may make the 2 minute boss fight faster." "I guess this route may speed things up." The problem with those is that you have to do each thing perfectly before you will know if its faster. Having a program to plug a bunch of values into? Cakewalk.
Sage advice from a friend of Jim: So put your tinfoil hat back in the closet, open your eyes to the truth, and realize that the government is in fact causing austismal cancer with it's 9/11 fluoride vaccinations of your water supply.
Posted: 2/28/2009 11:09 PM
Quote
Chamale
He/Him
Player (182)
Joined: 10/20/2006
Posts: 1355
Location: Canada
As far as hardest to perfect goes, plenty of games come to mind. Yoshi's Island of course, as well as Pokémon or other RPGs that require a lot of luck-manipulation and bug exploiting. Many 3D games and FPS games must be very hard to TAS, as well as games with many different possibilities for the route. Some other games have technical issues which make them harder to TAS, such as most N64 games, games which tend to desynch, and Baseball Simulator 1.000 (crashes at random when you try to load savestates).
Posted: 2/28/2009 11:36 PM
Quote
Warp
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Chamale wrote:
Some other games have technical issues which make them harder to TAS, such as most N64 games, games which tend to desynch, and Baseball Simulator 1.000 (crashes at random when you try to load savestates).
I was not really talking about "hard to TAS because the emulator is lacking", but "hard to TAS because of how the game has been designed/implemented", even if the emulator and other tools work perfectly.
Posted: 3/1/2009 12:11 AM
Quote
DarkKobold
Skilled player (1652)
Joined: 11/15/2004
Posts: 2202
Location: Killjoy
Arcade TASer of 2010 PCE TASer of 2010 Arcade TAS of 2010 Lucky TAS of 2009 Lucky TAS of 2007
Warp wrote:
I was not really talking about "hard to TAS because the emulator is lacking", but "hard to TAS because of how the game has been designed/implemented", even if the emulator and other tools work perfectly.
If that was the actual question, then the answer is pretty simple - Any game with an RNG that is purely sequential, and requires delay of n frames per sequential movement of the RNG becomes an NP complexity problem very quickly.
Sage advice from a friend of Jim: So put your tinfoil hat back in the closet, open your eyes to the truth, and realize that the government is in fact causing austismal cancer with it's 9/11 fluoride vaccinations of your water supply.
Posted: 3/1/2009 12:19 AM
Quote
Warp
Banned User
Joined: 3/10/2004
Posts: 7698
Location: Finland
Doesn't that assume that you don't know how the RNG works and consequently you can't calculate in advance what it will produce for a given series of actions? If the RNG has been disassembled and every affecting factor studied, and its behavior monitored with a memory watcher, there's nothing NP-complex about it. (Ok, perhaps it could *still* be NP, maybe, but it would probably require for the RNG to be cryptologically strong, based on the yet-unsolved-in-less-than-NP factorization of huge numbers. I doubt any console games have such high-quality RNGs.)
Posted: 3/1/2009 12:54 AM
Quote
Soulrivers
Former player
Joined: 12/27/2006
Posts: 532
Location: Göteborg, Sweden
OoS. :(
My published movies [03:45:05] <Naohiro19> Soulrivers: ... [03:45:19] <Soulrivers> ? [03:46:35] <Naohiro19> <Soulrivers> No! <Naohiro19> So? <Soulrivers> Yes! [03:46:48] <Naohiro19> joke
Posted: 3/1/2009 1:29 AM
Quote
Cpadolf
Skilled player (1444)
Joined: 7/15/2007
Posts: 1468
Location: Sweden
SNES TASer of 2009 Rookie of 2008 SNES TASer of 2008
Soulrivers wrote:
OoS
Why? (I am guessing that you are talking about Oracle of Seasons).
Agare Bagare Kopparslagare
Posted: 3/1/2009 7:50 AM
Quote
Kles
Joined: 7/28/2005
Posts: 339
I've yet to see a TAS even attempted on Stunt Race FX/Wild Trax. 3D chip AND racing game? Preposterous!
Posted: 3/1/2009 8:09 AM
Quote
Bisqwit
Editor, Active player (297)
Joined: 3/8/2004
Posts: 7469
Location: Arzareth
Any game that has long-lasting and difficult-to-control randomness is hard.
Posted: 3/1/2009 8:10 AM
Quote
Dromiceius
Joined: 10/3/2005
Posts: 1332
I'll chip in 7th Saga (U). The game became unbalanced during localization- deliberately or not- such that a lot of grinding is necessary to have any chance of reaching the end. This is true even in a TAS, because it's not possible to manipulate luck to the same degree as in most RPGs, where one can often coast to the end on EXP gained from boss fights.
Posted: 3/1/2009 8:50 AM
Quote
Kirkq
Player (210)
Joined: 7/7/2006
Posts: 798
Location: US
7th Saga seems pretty linear, honestly. The battle entry and fighting has to be brute forced as well as the stats upon leveling, but at the end of the day we can just level up one extra time if we don't quite make the stat cut. The routing is fairly straightforward once you work on it for a while. Anyone have any input on Secret of Mana in this regard?
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 3/1/2009 10:58 AM
Quote
Soulrivers
Former player
Joined: 12/27/2006
Posts: 532
Location: Göteborg, Sweden
Cpadolf wrote:
Soulrivers wrote:
OoS
Why? (I am guessing that you are talking about Oracle of Seasons).
Because of the luck manipulation. A sword slash in one screen can change stuff fifteen screens away, or more, and none of that resets when you enter buildings or dungeons. So you're basically stuck with optimizing until that fifteenth screen, only to find out it didn't turn out the way you wanted it to.
My published movies [03:45:05] <Naohiro19> Soulrivers: ... [03:45:19] <Soulrivers> ? [03:46:35] <Naohiro19> <Soulrivers> No! <Naohiro19> So? <Soulrivers> Yes! [03:46:48] <Naohiro19> joke
Posted: 3/1/2009 12:48 PM
Quote
moozooh
Senior Moderator
Joined: 8/4/2005
Posts: 5777
Location: Away
See also: Aria of Sorrow, all-souls run.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
Posted: 3/1/2009 1:28 PM
Quote
Yrr
Joined: 8/10/2006
Posts: 289
Location: Germany, Bayern
Actually, all Aria-of-Sorrow-Soma runs are quite annoying... the game mechanics is pretty complex, it is impossbile to tell if you have done things perfectly. TASing this game takes forever. I started a run of this game at the time I had joined this forum and I havn't finished it yet =/
Posted: 3/1/2009 1:39 PM
Quote
moozooh
Senior Moderator
Joined: 8/4/2005
Posts: 5777
Location: Away
Well, Zurreco had started an any% even before you joined, and quit after something like 200k rerecords due to luck manipulation failure, marketed under hard mode obsoletion precedent [1]. :) Even more disheartening was the fact that it got obsoleted by using a single new trick… twice.
Warp wrote:
Edit: I think I understand now: It's my avatar, isn't it? It makes me look angry.
1 2 3