Ikari Warriors II: Victory Road (ntscus) (nes) [Any %] [Single Segment] [0:07:52]
Run Information
Completion Date Oct. 20, 2020
State Published
Runner 'ktwo'
Verification Thread
Run Comments

I did this run on a Powerpak. In case it matters for future reference, I used a ROM with checksum 94edaed6f4a80e27dea80f34d16baf7f.

SDA history
22-Aug-2013 - 0:08:47 by Master-88 (http://ia902503.us.archive.org/23/items/2014-06-18-SDA_Speed_Runs/)
14-Nov-2014 - 0:07:58 by ktwo (http://www.archive.org/download/2015-05-12-SDA_Speed_Runs/)

Speedrun information
For information about speedrunning Ikari Warriors II, go to the sda knowledge base (referred to as sdakb in my comments below):

My previous run for this game was without any big visible mistakes. Since then, a few new routes/techniques have been found, allowing the lower time of the new run:

  • When working on the first Ikari Warriors game, I got a better understanding of the movement speeds. It turns out some of that can be applied to Ikari Warriors II as well. In particular, the first and last movement of a diagonal movement are a sort of "half-diagonal" movement. These "half-diagonal" movements are more efficient than the full diagonal movements. So when you need to move in the horizontal direction, it's faster to divide the diagonal movement into as many diagonal movements as possible (and thereby increasing the amount of "half-diagonal" movements). The effect is very small each time, but they do add up. Maybe one to two seconds saved over the course of a run.
  • A more direct route through the flying section of the second level. Found during the TAS creation of this game (made shortly before this run was done). Around half a second faster.
  • During the work with the TAS, I finally looked into what triggers the content of the end-of-level tunnels. Previously, several of them were more or less 50/50 to contain items (and not enemies). This created severe bottlenecks in the attempts and forced many resets. With the new findings, the content of every tunnel can be manipulated to always contain items. With this resulting in more attempts reaching further, a bit more risk could be added elsewhere. E.g. slightly more aggressive menuing or the more direct routes in levels 4 and 5. Difficult to estimate how much time this saved, but maybe a second?
  • When doing console attempts, I found a faster way to pass the statues at the beginning of levels 3 to 5. Instead of passing more in the middle between the status, I now start by going left, while taking out the left statue with a grenade. Around half a second saved per stage, so ~1.5s in total.

Run comments
There isn't a lot to comment on in this run that's not mentioned in the sdakb. This is an "easy" game to speedrun. I was therefore basically resetting every attempt that had any kind of visible mistake. The end result is therefore essentially a "flawless" run. That being said, there are still minor things that could be improved in many places. The menus are quite unwieldy in this game and easy to make mistakes in. More aggressive menuing could therefore chip away small amounts of time during each menu trip. Same goes with trying to stutter-walk diagonally even more (see the sdakb for the reason) or finding the most optimal line to walk in in every situation.

Final thoughts
I didn't want to settle for a run with any kind of visible execution mistakes. This run achieves this goal. It's the only proper attempt reaching the end. I had two other attempts arriving at the final boss, but missing the quick-kill. The pace of this attempt was the best of the three though. Once my strats were defined and I made serious attempts, my goal time was around a low 7:54, but with the feeling that it would probably feel a bit slow. I was therefore hoping for a 7:53. Getting a 7:52 was therefore a positive surprise. More time could come off by really learning how to optimize the menuing in this game and, to a lesser extent, the diagonal movement and taking straighter paths. Unless something new is found, I don't plan on trying to lower the time further though.