Tool-assisted game movies
When human skills are just not enough

Submission #4822: CogneatoSwitch's NES Nobunaga's Ambition in 12:06.37

Console: Nintendo Entertainment System
Game name: Nobunaga's Ambition
Game version: USA
ROM filename: Nobunaga's Ambition (USA).nes
Emulator: BizHawk 1.10.0
Movie length: 12:06.37
FrameCount: 43654
Re-record count: (unknown)
Author's real name: Alan Hayashi
Author's nickname: CogneatoSwitch
Submitter: CogneatoSwitch
Submitted at: 2015-09-05 06:29:39
Text last edited at: 2015-09-10 05:54:46
Text last edited by: Spikestuff
Download: Download (7822 bytes)
Status: published
Click to view the actual publication
Submission instructions
Discuss this submission (also rating / voting)
List all submissions by this submitter
List pages on this site that refer to this submission
View submission text history
Back to the submission list
Author's comments and explanations:
A brief run-through of the classic Japanese Strategy game, Nobunaga's Ambition

Game objectives

  • 17 Fiefs
  • 25 Seasons
  • No Wars!
  • Emulator used: BizHawk 1.10.1


This is my third submission of a Nobunaga's Ambition TAS. (Here is my first) (Here is my second)

This TAS improved upon my previous techniques and used granting to move through the game at a lightening pace. As requested, I tightened the inputs as much as possible, with a special note below, and rounded up or down numbers to reduce the number of inputs. Finally, I added subtitles to the top of the video to aid in understanding why each move was made.

Most of this TAS is about getting Assassination Power, or AP, as I call it. AP is the sum of a lord's Char and IQ, and must be greater than then enemy's for a successful assassination.

In the stats screen, I spend about 200 frames maximizing Char and IQ. This is far more efficient than using seasons (~800 frames each) later on to increase AP.

Luck is the next most important stat, as it must be greater than the enemy's luck-30. Health and Drive affect random events and AI, but only if they are low.

Fortunately, the random numbers are essentially sequential, so getting higher values for the first 3 stats did not affect the time to first turn at all.

However, I found doing this stat selection resulted in animated opening events, namely Fief 1 having a revolt, which takes a lot of time and occurs before I can turn off the animations. I found that the RNG is directly manipulated by waiting extra frames at the first button press (around frame 14). Waiting 2 frames at the beginning entirely prevented any opening events, saving over 800 frames, and sped the stat selection tremendously.

Most of my actions are straight-forward: assassinate when able, move men to the front, and keep the lord's fief filled with gold for bids. Sometimes, however, I had to make unusual moves to manipulate the RNG, such as moving 1 man so an assassination would happen faster, or waiting to assassinate until later to prevent counter attacks and manipulate AI behavior.

The most noticeable and slowest oddity was not granting Fiefs 13 and 12 when building to assassinate Fief 5 and 9. This was because granting them did not raise AP as quickly as doing it manually, and caused Luck to plummet as to prevent assassinations. Other permutations of granting and un-granting fiefs had similar affects to Luck and were slower.

Another was the no-bid on Fief 4. This was done because bids only occur when two or more lords can bid on the fief, and would have caused two battles to have been fought instead of one.

I was happy to discover the final move of granting a fief to military to prevent the final war, which saves 3 minutes or so. I tried to use the tactic earlier, but found it took longer to move men into position and then grant the fief to attack than to just build up AP and assassinate them.

Possible Improvements

Honestly, there isn't much left. This particular run might be faster with other permutations of turns to avoid some of the RNG manipulation moves, but I was unable to find them despite my best efforts.

Also I searched through permutations of the stats selection and opening frames, and found over 300 other combinations with similar “frames to first turn” values. I imagine some of them would yield faster solutions.

There is still the 50-fief version to go!

ars4326: Claimed for judging!

ars4326: Updated movie file. Previous time: 12:06.57 (43,666 frames). Current time: 12:06.37 (43,654 frames).

ars4326: Hello again, CogneatoSwitch. As already expressed by fellow forum members in the thread, outstanding job on coming back to this run and implementing some major improvements (especially in input optimizations)! The Fief granting strategy you mentioned in the prior submission really made a substantial difference here, also. And despite being a bit limited on entertainment value, it is certainly now publish-worthy. Congrats on your first published run!

Accepting for publication to the Vault!

Spikestuff: Publishing.

Similar submissions (by title and categories where applicable):