Submission #4785: CogneatoSwitch's NES Nobunaga's Ambition "17 Fiefs" in 23:24.39

A brief run-through of the classic Japanese Strategy game, Nobunaga's Ambition

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

The main purpose of this TAS is to beat NA on the hardest difficulty, which results in a vastly different strategy than the first video. I focused on winning fiefs by assassination, which turns out to be much faster than any battle-based victory, and as a result, this video is even faster than the original.

This video is optimized for time, rather than seasons. There were a few places I could have saved seasons by starting a war, but chose instead to focus on speed and assassinations.

Throughout this run, my action of choice was “Give – Peasants – Rice” (GPR). This boosts the Char of my leader, which is key to assassinations.

I plan to make another TAS for the 50 fief version.

Fief 15 was chosen because of its strategic location next to two easily-assassinated fiefs that also protect it. This “wall of fiefs” is important, as the AI is able to create better fiefs faster than the player, especially on difficulty 5. Fortunately, they are still weak against assassination.

On the first turn, I was able to kill and take over fief 10.

It took a bit of luck manipulation to find a solution where Rokkaku wasn't killed by Asai or Kitabata. In this case, by assassinating both fiefs at the same time, Rokkaku's land was preserved. Also, order was important, as killing Rokkaku first caused Ashikaga to attack instantly.

Asai and Kitabata were the next weakest of the Daimyos. After moving my men forward, I GPR'd until I could kill them.

By this time, Oda had taken over Fief 9. I needed 60 more Char to assassinate him, but only 12 more for Asakura. The choice was obvious...

I also lucked out that Fief 5 went first the following Season, allowing me to buy troops to prevent Oda's attack.

I had trouble mapping this area. Oda would attack me if I had too few men in any of my front fiefs. It happened that, when I took over Fief 6, Oda attacked immediately. I managed to find a path that allowed Fief 11 to move men to Fief 5, which moved them to Fief 6 before Oda could react.

I also started having money problems. If the Fall comes and the fief doesn't have enough money, the men leave, opening me up to attack. The solution is either to raise taxes or send money, but both take up precious turns, and increasing taxes causes Char to go down. I solved the problem by sending money in advance, which also manipulated luck enough for me to avoid getting attacked.

Believe it or not, this was the climax of the video. Oda is the most powerful Daimyo in the game, and all throughout he kept me on my toes, forcing me to move money and set taxes in such a way that he didn't attack.

When I finally had enough Char, I took him down. After that, it was all essentially end-game. Because Char was no longer needed, I started Granting the more useless fiefs to speed things up. (Grant operates the fief using an AI, rather than my explicit orders)

With my Char built up, I assassinated these two easily. At this point, I was concerned about the way bidding is done: you may only bid on a fief if an enemy leader can also bid on it. Therefore, I had to kill them in different seasons, otherwise the isolated fief would gain a new leader and I could never win it by assassination.

This was a straight-forward assassination, but it was serendipitous that I could do it in the same turn as the next assassination.

This was an interesting puzzle that threw me for a loop. To win with only 1 war, I needed to ensure that there was always a leader who was able to bid on the conflicted land. Killing Uesugi would cause a new leader to spawn in Fief 2, and if I won Fief 4, there would be no way to get Fief 2 without an extra war. The solution was to NOT bid on Fief 4, which luckily was won by Hatakeya of Fief 1. This allowed me to kill the new leader of Fief 2 (Sira) and still bid on the land.

Now there was only Fief 1 and Fief 4 left. Fortunately, killing Hatakeya caused a new leader to spawn at Fief 1 (Kite), and that new leader was allowed to bid on Fief 4, giving me the land.

Given how I had collected all my troops from every other fief, this battle was just silly. 744 vs 53, and he still killed 145, showing just how advantaged the AI is in the war. I opted to move unit 5 along with the group so I wouldn't lose time switching screens.

For most of my GPR's, I automated using ¼ of the current rice value. However, numbers like 5 and 6 take longer to get to than 0. Given how much time was spent on that action, significant time savings might be had by rounding to the nearest 10 when possible. We will likely see that optimization in the 50 fief version.

There were also a few times I GPR'd when I didn't need to, and I probably could have granted fief 5 after Oda was dead. Also, it's possible granted fiefs also give me Char, which would further speed things up. Strategies involving granting in general haven't been explored.

This run is NOT optimized for seasons. In fact, my first mapping in difficulty 5 won in 19 seasons, but took significantly longer with 3 extra wars. I believe a season-based optimization would be enjoyable in an entirely different way.

ars4326: edit: Shortened branch name to just '17 Fiefs' (since highest difficulty is already required) Claimed for judging!

ars4326: Hello, CogneatoSwitch. I've gotta be honest with you on this run: Your sub-optimal inputs (whether by automation scripts, or otherwise) cost you significantly on time. By using frame advance and choosing/confirming commands at the earliest possible frame, I expanded on that test WIP I made and beat your time by 898 frames (almost 15 seconds) at the point of the first command phase of fief 10. Other savings were made by rounding amounts to the nearest 10, as well as from a better optimized stat creation screen (mentioned in the thread).

I'd really like to see a better optimized run from you. Your big apparent sticking point here is just learning to patiently construct a run with frame-advance, and not relying on auto-scripts to generate input. If you can get this down, 3-5 minutes could easily be shaved off of this run (not counting the granted fiefs strategy you mentioned). Take your time and also explore any alternate strategies which you think could be faster, as well.

Rejecting due to sub-optimality.