Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
I'm the original creator of the Professor Oak Challenge. This is a very impressive technical achievement, and a significant improvement over your first TAS of this category. In my opinion, this goal choice is not ideal for a TAS because of the large amount of level grinding in the early game. The "Co-op Diploma" category would be faster and more entertaining, with a clearer goal choice to the casual viewer. I think Gabraltar has looked into doing that for Ruby/Sapphire, but I don't know if he's actually started on it. Anyway, I still found this TAS very interesting. I especially enjoyed the use of Linoone as a main battler, and the optimized Berry Blender to evolve Feebas. I support putting this in the Playground.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
I loved this game when I was a kid, very cool to see it as a TAS! I think that considering how fast the player can sleep, putting $70,000 in the bank and using the interest rate is faster than betting it in the casino. If the interest rate is 2%, you'll double your money in only 35 days. I'm curious what the route would be for the game mode with only 100 days to play, so that you can't just sleep for years to level up your stats.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
Twisted_Eye wrote:
Definitely my youthful years' most played Rampage so I'm biased, but I definitely liked the run. The effort to maximize efficiency is clear! It only seems repetitive because the game is so dang long. But getting the run under an hour really is impressive, good job!
Thanks! I played it a lot as a kid, and I always wondered what a TAS would look like. I'm very glad I was able to complete this.
GJTASer2018 wrote:
I certainly like the idea, but I'm wondering how much time having extra players would really save, especially in the later levels where you can take care of things really quickly with powerups using just one player. Not to mention health management and stuff would be a lot more complicated; a three player run might do better as a playaround instead of aiming purely for speed.
Aiming for maximum possible score instead of speed might be another category worth attempting. I believe you can also do the original arcade game now in MAMEHawk, but it would be a really long endeavor - the game goes through a 128 city loop five times before resetting back to Day 1.
I haven't tested what multiple players would look like. The big question is lag - with River City Ransom, a two-player TAS is actually slower than one player because of the increased lag, although RCR is a very laggy game and Rampage World Tour is not. Assuming lag isn't a problem, a three-player TAS would probably save something like 5-10 minutes, which is a lot of effort for a relatively small improvement.
I don't think a high score TAS would actually be entertaining. It would be a lot slower because of the need to pick up as many items as possible.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
gabraltar wrote:
Yes I have seen that.
Unlike in Yellow, Caterpie can actually get past the first gym on its own (with smokescreen cyndaquil for bird keeper abe and harden + x attacks for falkner). It's obviously slow as hell but idt grinding is a good start here esp with the very low lvl of the wild pokes.
I remember doing the calcs and caterpie would get to lvl 10 and evolve right after the first grunt in slowpoke well
RangerSquid was able to grind with the lvl 9 pidgeottos which gave massive amounts of exp but crystal doesn't have anything like that
There is one low-hanging fruit of EXP worth going for: Hiker Daniel's Onix is 253 EXP. You could catch a level 4 Wooper, since the run needs a Surf employee anyway, damage the Onix, and then KO it with Metapod.
Instead of the glacially slow strategy of using multiple Hardens to survive Pidgeotto's Gust, you could use Cyndaquil to defeat Pidgey, let Pidgeotto KO it, and then finish the battle with Caterpie. It would be beneficial to evolution-cancel Metapod, or go from level 6 to level 8 in one battle, to save time when Butterfree learns its fourth move.
Caterpie's EXP after each battle:
Caught - 64 (level 4)
Mikey - 134 (level 5)
Abe - 189
Rod - 311 (level 6) or 353 (level 7)
Falkner - 568 (level 8) or 610 (level 8)
Daniel - 821 (level 9) or 863 (level 9)
Grunt - 1039 or more (level 10)
If Cyndaquil helps in the Rod battle, there's no need to evolution cancel, but I'm guessing it's faster to have Caterpie do that whole battle so that it has an extra level against Falkner.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
gabraltar wrote:
FractalFusion wrote:
gabraltar wrote:
Caterpie% for Crystal
I'd like to see the strategy for that. Because I don't think that is happening.
It's mostly a meme idea atm and I haven't done a lot of routing for it. Caterpie would obv be very slow but far from impossible.
Some of the early hurdles can be beaten with scuffed strats and a lot of rng (bird keeper rod can be overcome with smokescreen cynda, falkner can be overcome with harden spam + 6 x attacks). Beyond that, you'd likely need to go more out of your way for decent moves (Return and Hidden Power, most notably).
Have you seen RangerSquid's Butterfree% TAS of Yellow version? Confusion and heavy luck manipulation allows Butterfree to defeat anything in the game. The tricky part will be getting a Caterpie to reach level 10 as fast as possible.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
It's a tough week for Home Alone 2 fans, as one of the film's actors has recently arrested on federal charges.
This TAS is a great technical improvement on every level, but I'm voting meh for entertainment. It's unfortunate that the game's movement mechanics requires that constant oscillation to get maximum speed, as it looks terrible.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
Grabbing the second Friend Ball isn't that much of a time expense. Silver could catch a level 2 Zubat before the first gym and have it gain friendship by keeping it in the party, and eventually giving it 12,000 EXP to evolve into Golbat and then Crobat. That's how I got Crobat in my 25-hour glitchless catch 'em all speedrun, but it would take much longer than going the Friend Ball route.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
Finished watching, voting yes. This is a masterpiece of route planning. I loved seeing the Eevee and Togepi eggs hatch on the exact step they were needed. Seeing Bulbasaur with a battle with Bide was an early highlight. For a publication image, I'd suggest 27:50 with three different starters being used in battle. Or 4:35:50 with two evolutions happening, and a Pidgey confronting Snorlax.
Are the overworld Suicune encounters (other than the first) actually necessary to catch it or Ho-oh? I remember being able to find Suicune despite not running into it on Route 42 or Route 35.
FractalFusion wrote:
That's a lot of coordination between the games.
Would doing 4 games be faster than doing just 3?
I doubt it. Trading between games is the major bottleneck, and Yellow version is finished much earlier than the generation 2 games.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
I've been following your progress on this TAS for years, it's an extremely impressive piece of work. Congratulations on submitting it; I plan to vote after watching, but I think it should be an obvious Yes.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
Playing against a chess computer and minimizing the thinking time is a fascinating challenge. This is an easy Yes vote, and I'm curious to see if anyone is able to make an improvement in the future.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
Arc wrote:
That is a sharp idea. However, the movie rules do say that the final input is a stylistic choice. It's not easy to decide which way is better. Your method is shorter input, but then the final scoreboard appears 188 frames later. I think that kicking the field goal so early is rather unnatural in real-life football, and it leads only to an underwhelming failed kickoff return, and so I think that ending would leave a viewer feeling less entertained than a game-winning score as time expires. More feedback is welcome here.
That makes sense. I think the fastest way to see the final scoreboard would be to fumble the ball into the end zone right before time expires, let the AI recover it, and have an AI teammate tackle them in the end zone. That would reduce the total number of plays to four. Sometimes a teammate gets thrown into the end zone after doing a block, so he should be able to recover and make a tackle.
Edit: Or, if you want to end it with a field goal, it could be kicked from as far as possible so that you can end input sooner and still finish with the same overall time.
Edit 2: Never mind, I just tried that, the clock doesn't count down while a field goal kick is in the air.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
This is a great run, I love that dodging tackles for 5 minutes works as a speedrunning strategy and as a playaround. It's possible to end input a bit sooner with a different ending strategy. Here's an example that is 438 frames faster. I modified your input to kick a field goal with 50 seconds left, and then made a long kickoff that runs out the clock to 0. The longest kick return I was able to luck manipulate was 50 seconds of in-game time, but longer ones should be theoretically possible. It may be possible to end input by punting the ball with the score at 0-0, and manipulate the other team into catching it on the 1-yard line, fumbling backwards into their end zone, recovering the ball, and then getting tackled for a safety. This would be improbable to manipulate, but with the kicker at the right distance, it might be doable.
Happy Super Bowl Sunday.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
arkiandruski wrote:
So, the more I try to look into this run, the harder time I have giving it a succinct objective goal.
This run and Chamale's run before it seek to unlock a final boss fight where Slick calls on three others and you fight four people. Other than that, it basically follows an any% route, except they both go out of their way to recruit Ivan, which requires talking to multiple NPCs.
Currently the run is listed as "Secret Bosses". The justification for that is unlocking the final boss fight with four enemies, but if that counts as a secret boss, shouldn't those other fights mentioned be as well? Also, there are bosses that are optional which can be seen as secret or extra, like Benny and Clyde who require going on a branch from the main path, Alex/Ryan (depending on the PC you choose, you can fight the one not chosen) which requires returning to a previous area, and Jesse who requires beating a certain boss five times.
The decision to recruit Ivan is also difficult. Like the Slick boss fight, I can't really say it's the only character recruit that has hidden conditions to happen. There are some characters where you have to refuse invites in order to have a chance of recruiting them for example. Also, having Ivan is not a requirement for the Slick fight, nor do I think recruiting him is the fastest way to fulfill those conditions.
As far as I can tell, the best this submission can get is to be a Playground run, unless there is something I'm not seeing here.
I don't agree with tom_mai78101 describing it as a "secret events" run, because there are a lot of secret events and this run doesn't do most of them. I think the alternate final boss fight against 4 enemies is worth including as a separate category, but I think that's up to the voters.
Recruiting Ivan is the fastest way to get four members of the party. Gary or Rick require starting with only one character, and Conan requires defeating thirty Generic Dudes.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
jeff_town wrote:
so the computer (if you can call a 2600 a "computer") blanks the screen and thinks for up to 2 minutes between moves?? wow that's terrible.
There was even a chess game on the 2600, which took 12 minutes to move on the second-highest difficulty and 10 hours on the highest.
I voted yes on this movie. I think it makes an interesting challenge to see how quickly you can play a game against such a primitive processor, when limiting its thinking time is one of the biggest obstacles.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
I'm working on an improvement to this run. I think it's possible to save time by manipulating enemy positions. I'm 966 frames ahead as of the beginning of the Rocko fight, but I don't understand how to do the glitch where you throw him off the map - could you please explain it in more detail?
Also, I can't find the exact ROM you used for this run. The good dump I found has a CRC-32 of 8686436E.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
I take it the objective is to unlock and win the secret, four-person final boss fight. I like it, this makes it very different from the NES version of the game. Voting yes.
I've published a couple TASes for River City Ransom on the NES. Did the Dragon Kick turn out to be faster than using weapons for fights? I remember how difficult it is to optimize fights with the erratic enemy movement involved. On the NES, the trash can and the metal pipe deal much more damage than fighting with bare hands. I suspect that using Seraph Cola with luck manipulation could give a big boost to your weapon stat of choice - I think it merits testing.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
DJ Incendration wrote:
What is this, May Fools? I'm confused.
It is a semi-serious run. Battle Chess has had a number of submissions over the years, none of which were published. This seemed like an interesting way to create a non-trivial category for the game.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
RangerSquid wrote:
Considering it could take upwards of 4-6 months just for a new one, that is mainly up to you. I’m fine with the judging for now because I could lose motivation for this project halfway through, but I don’t know until I have a new caterpie to use because I’m also implementing the tackle animation mentioned earlier in this post.
Good luck. You might want to test whether you'll need a new Caterpie manip or not. It may be faster to not touch the options menu until after the Sandshrew fight, or it might be faster to turn off animations before fighting Pidgeotto and then turn them back on. Thundershock is about 60 frames slower with animations on, and Gust is about 30 frames slower, so the animations for 2 Thundershocks and 2 Gusts should be marginally faster than entering the options menu twice, but that's based on rough math.
One other early game wrinkle: Catching Pidgey in the Bug Forest slows you down when you use the Pokemon Center in Cerulean. It would be better to catch a Pidgey on Route 6.
I know that your current Caterpie has DVs of FEEF. The max Attack and Special are necessary; if you end up manipulating a Caterpie with more HP that will probably be a good thing for Take Down but I'm not sure how it affects redbar throughout the whole run.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
RangerSquid wrote:
Since you actually made a test run, I will look into actually making a full fledged version of this. I still enjoy the idea of the meme game corner Hyper Beam pickup, but it definitely is incredibly slow.
Mega Drain will be a hilarious move to use. Confusion had some utility in the final battle because its animation is faster than Psybeam, but I'm pretty sure replacing it with Mega Drain (instead of tossing TM21) will be worthwhile.
Edit: If you use Mega Drain on Giovanni's Rhyhorn, against Sabrina you can:
KO Abra with Take Down
KO Kadabra with high-damage crit Mega Drain (at least 21 damage, out of a max of 22) followed by Take Down
KO Alakazam with two Take Downs.
This leaves Butterfree with exactly 1 HP.
I'm not sure when is the best time to use Mega Drain before the Viridian Rival Fight - potentially on Giovanni's Dugtrio or Rhyhorn, or on the Rival's Sandslash (which would require two hits).
Also, I looked into redoing the Metapod section with animations on. It's barely worth the menu time: 234 frames saved from the faster Tackle animation, but about 190 frames lost from going into the options menu twice. There's also no benefit to delaying Caterpie's evolution, even though its attack is slightly higher than Metapod's.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
I made a test run that avoids the 9-minute detour for Hyper Beam. It's 5 minutes and 7 seconds faster than this video. I used your input for the first half, and my part is very sloppy - this is just a route test. A good version would save at least a minute with tighter input and some route changes.
Because of the need to use Take Down in more battles, this run has two additional Pokemon Center visits. They could be avoided by teaching Butterfree Mega Drain after it learns Psybeam - this would let it heal partially before it needs to use Take Down, such as by using it on Giovanni's Rhyhorn before Sabrina. That way, the run could stay in Redbar longer. It also saves a single turn against Lorelei's Slowbro.
Experienced Forum User, Player, Published Author
(162)
Joined: 10/20/2006
Posts: 1348
Location: Canada
RangerSquid wrote:
I appreciate the idea, but I have absolutely no interest in attempting the fight unless I find a faster overall route for the TAS than this one
That's completely fair. I don't have the time to create TASes these days so I can only suggest route changes.
I wonder if you could avoid needing Hyper Beam by using Psybeam to confuse tough enemies into hurting themselves. Starting with Psybeam and then manipulating confusion damage would be easier than trying to manipulate Gen 1 Misses for some fights. My River City Ransom TAS defeated the later bosses by making them attack themselves.
