Trying to TAS this game
2 years ago
Québec

Greetings!

I intend on TASing it game as it is a fairly simple one to manipulate RNG for (at the very least, for the battles). I have made a test for the first scenario which you can look at here:

Any comment or suggestion would be very much welcome. Thank you!

DracoREXXX and Amune138 like this
Iowa, USA

Hey juef,

Keep us posted! Even though I'm the moderator of the game, there is so much to it that I have yet to learn.

I'm interested in knowing definitively if the jackpot actually has a rhythm/ pattern/ cycle to it? Something I'm sure you've figured out the answer to with this...

Québec

I do not believe so. I've written a Lua script that checks how many frames it takes to get a jackpot for each of the possible "items", and selects the shortest one. In the two fights in this scenario, it was not the same item which was the fastest, and it was not the same number of presses between items for both battles.

The most "interesting" I have found so far is simply that the game seems to ignore any kind of single-frame input. To actually do anything, one has to press a button for at least two frames. I'm not sure how common this is for NES games, but the only other I've TASed wasn't like this.

Oh, and I may be wrong about the following because I didn't keep the input, but I think the battle RNG is independent of the map movements. I tried fighting Godzilla with both types of planes, and if I remember correctly, both battle jackpots were the same, and I believe with the same input, despite the map movement being different, the fighting unit being different, and the framecount at the start of the battle being different. So if the RNG does not depend on uninitialized RAM and someone trains himself or herself at pressing the buttons at the exact same frames for the first battle, and the right set of frames for the second battle, and so on, then that could work. But obviously, that is very hard and quite unlikely possible, except for the simplest of the scenarios.

Québec

Hello again! I have done the 2nd scenario (sorry for the lack of sound):

I might have to do it again though, notably because I didn't realize one can speed up exiting the battle by pressing buttons (or am I crazy?). I also want to test a few more things.

I have a question you might be able to answer which would help me a lot: how do you choose which weapons to use? So far, it seems like the bottom one always does the most damage.

Also, I can pretty much confirm that map movement has absolutely no impact on the battle RNG... but of course, it does influence enemy map movement. For instance, in my test TAS, I move the Attacker just one tile because if I don't, Baragon will disappear instead of attacking the Fighter. One of the things I have to check is if it still works by just selecting the Attacker and having it stand by (or selecting another unit altogether).

Edited by the author 2 years ago
Québec

Never mind my question about the weapons - I read the game manual and unless it's wrong, it has all the info I need about the weapons :)

Ohio, USA

yeah the reels and the battle screen are independant.

it could be possible but extremely unlikely to teach yourself with 3 frame perfect inputs per battle to always hit trips, but the rhythm would be different from the first set to the second etc... it just uhhhhhhh

would be extremely difficult to carry through such a long run. or even 1 scenario in that matter.

i am like 85% sure the manual is wrong about what the reels that you hit are supposed to do. a lot of the time red seems to benefit you.

because of this i dont know about the weapons being totally accurate either. some stuff checks out but some doesnt. the information in the manual and in the game feels like it's very much in the air.

i think it might be possible to pull the same enemy map inputs if you do the same map movements every time too. ive only tested it through like the first 2 or 3 moves in a map.

Québec

i am like 85% sure the manual is wrong about what the reels that you hit are supposed to do. a lot of the time red seems to benefit you. because of this i dont know about the weapons being totally accurate either. some stuff checks out but some doesnt. the information in the manual and in the game feels like it's very much in the air.

I agree! That, along with the fact that weapons' animations don't all have exactly the same duration, implies that I have to test most weapons everytime anyway. But the difference is somewhere between 10-20 frames for each battle, which is pretty much negligible given the importance of RNG in this run.

I have completed the 3rd scenario, and in this one too, I managed to finish before the last monster appears (in this case, Mothra). Hopefully the trends keeps up with the next scenarios :)

buffalax likes this
Québec

Greetings! Time for a new TAS WIP! I think I'm getting the hang of this. Please let me know if you see any possible improvement or if you have any question.

Ohio, USA

i havent sat and watched the whole thing but it looks great i am super happy you put a lot of work into this game!

juef likes this
Game stats
Followers
22
Runs
79
Players
6
Recent runs
Latest threads
Posted 2 years ago
9 replies
Posted 6 years ago
0 replies