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

Submission #6979: Challenger's NES Castlevania "pacifist, minimalist" in 11:07.52

Console: Nintendo Entertainment System
Game name: Castlevania
Game version: USA PRG0
ROM filename: Castlevania (U) (PRG0) [!].nes
Branch: pacifist, minimalist
Emulator: BizHawk 2.4
Movie length: 11:07.52
FrameCount: 40117
Re-record count: 62238
Author's real name:
Author's nickname: Challenger
Submitter: Challenger
Submitted at: 2020-12-27 10:23:26
Text last edited at: 2020-12-30 18:46:39
Text last edited by: fsvgm777
Download: Download (10624 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:
As I said on any% TAS, I was working on pacifist category too. Like the previous two runs, this category does not contain any destruction in the levels.

(Link to video)

Game objectives

  • Emulator used: BizHawk 2.4 (works even on most recent versions)

  • Uses death to save time

  • Takes damage to save time

  • Heavy glitch abuse

  • Heavy luck manipulation

Once again, this run syncs on PRG0 version.

The improvement is 4674 frames (1:17.77 seconds) faster than the previous run. Like the any% run, most of these improvements came from that scroll glitch discovery.

New tricks and glitches

See the any% run for more details.

Development of the run

I started my work adding those new skips, but stage 15 caused a huge problem: while Simon dies at the start of room 1 on previous runs, this time you can't, and having less health means dying during the boss fight. Redoing that battle would be very annoying, so I decided to start optimizing this game until reaching that point again. I haven't tested scroll glitch on bosses at the time.

Since I worked on any% run at the same time, it was easy to improve pacifist run.

Well, I originally intended to submit this and any% run on the same day, but I ended finishing any% run first because I decided to show it on an marathon (which occurred this month). In other words, improving both runs took more time than I thought.

After finishing those optimizations, I resynched this run to PRG0 version, because nearly every transition is a frame shorter. Here's the original work done on PRG1 version: http://tasvideos.org/userfiles/info/68428051265792762

Resynching wasn't simply removing frames. There are some rooms which I got a lag frame, and some others with a single lag frame removed. You can see these points during stage-by-stage commentary:

Stage 01

  • There's no improvement for this stage, but scroll glitch is performed for a late skip.

  • The point for this skip is after the starting point from the last stairs, but there's a problem: the glitch can be lost if you return a bit.

  • So, in order to work properly, I spawn some tiles at the top point (glitch must be performed twice). With enough distance and performing at that new point...

160 frames lost

Stage 02

  • Simon falls from the starting point and lands next to the stairs. This skip saves 271 frames but since I wasted 160 frames on previous stage, the improvement is 111 frames.

  • This change also affected both room transitions (no more extra lag), so 2 frames were gained.

While PRG1 version (and the previous runs) got a lag on the room below, PRG0 version didn't.

274 frames gained, now 114 frames ahead

Stage 03 - Boss defeated a frame faster, and with a frame-perfect trick, orb grabbed a frame earlier = 2 frames gained.

The new run finishes with 236 time left while the previous run finishes with 234. 2 frames lost by having those extra seconds.

0 frames gained, still 114 frames ahead

Level 1: 114 frames faster.

Stage 04 - Improvement on the first room: if whip was useful at the orb grab, that trick can be more useful with enemies. 2 frames gained, and one more for waiting a frame before moving (after the first jump).

On PRG0 version, I added an extra jump to avoid lag (second room).

3 frames gained, 117 frames ahead

Stage 05 - Same strategy from any%.

This run got an extra lag transition on PRG1 version while on PRG0 isn't.

48 frames lost, 69 frames ahead

Stage 06

  • First room completely skipped and I got an extra lag transition this time.

979 frames faster

Second room

  • Same scroll glitch from any%.

  • Since I have no whip upgrade, it's not possible to defeat a boss with a single hit.

  • So, the strategy is doing a normal critical hit + 2 whip attacks before doing that critical hit = 128 frames faster.

The new run finishes with 346 time left while the previous run finishes with 330, losing 16 frames.

Room 2 is 100 frames faster at the end.

Total: 1079 frames gained; 1148 frames ahead

Level 2: 1034 frames faster.

Stage 07 - First room has the same strategy from any% run.

This run got an extra lag transition on PRG1 version while on PRG0 isn't.

Second room is lagless on PRG1 version, while PRG0 version got a lag.

86 frames gained, 1234 frames ahead

Stage 08 - Pause manipulatiom fron 2018 any% run added and a lag transition removed.

9 frames gained, 1243 frames ahead

Stage 09

  • Scroll glitch used twice: the first one is to force one of these skeletons for keeping away, while the later one is for boss battle.

  • The published run got a lag before, but this time it's avoided.

  • With good skeleton manipulation, that bone dragon fails to spit fireballs (sprite limit).

I waited 2 frames before loading boss battle for better pattern manipulation (right mummy). PRG0 version wait only 1 frame because I used that other frame to avoid lag.

10 frames faster

  • A completely new strategy: defeating both mummies with major critital hits using a new point created through scroll glitch.

  • While waiting for both mummies, I still must attack them.

  • Critical hits normally works by facing opposite direction, but doing on that point would result breaking one of those candles. Fortunately I managed to doing those critical hits without turning around.

It works by jumping to that point then taking damage between the ceiling and the floor (narrow height once again).

Battle is 123 frames faster.

The new run finishes with 423 time left, while the previous run finishes with 421, losing 2 frames.

Stage 09 is 131 frames faster at the end. 1374 frames ahead

Level 3: 226 frames faster

Stage 10 - Performing scroll glitch while standing on a moving platform is bad (when it stops moving, you have a small amount of time), but fortunately it worked since that point starts on a solid platform.

  • I got a lag transition at the end this time.

PRG0 version got a lag after the moving platforms path while PRG1 isn't. And the playaround isn't exactly the same, since lag sucks.

39 frames gained, 1413 frames faster

Stage 11 - frame-perfect strategy.

A framerule (16 frames) improved, 1429 frames faster

Stage 12

  • Same scroll glitch from any%.

Both previous run and this run (PRG1 version) got a lag while PRG0 isn't.

  • Boss defeated with 2 critical hits and 2 whip attacks = 200 frames faster.

The new run finishes with 327 time left, while the previous run finishes with 323, losing 4 frames.

Stage 12 is 173 frames faster at the end. 1602 frames ahead

Level 4: 228 frames faster.

Stage 13

  • First room: I only glitch the stairs, since not killing enemies ruins that strategy used on any% run (spawning a glitched tile too).

PRG1 version got 6 lag frames while PRG0 got 5.

203 frames faster

  • Second room: Same scroll glitches from any%.

416 frames faster

Total: 619 frames gained; 2221 frames ahead

Stage 14

  • First room: once again no damage shortcut. Despite losing time since I can't kill that axe knight, this skip is indeed faster.

60 frames faster

  • Second room: same skip from any% run, although I lost some time like the previous room.

656 frames faster

Total: 716 frames gained; 2937 frames ahead

Stage 15

  • First room: same skip from any%.

764 frames faster

  • Second room:

  • With that whip trick and perfect timming, it's possible to bypass both axe knights much faster. While the first one still needs some waiting, the later one is completely avoided.

  • Since I have less health this time, the battle must be redoing (only one critical hit instead of two).

  • If fighting Death boss was so annoying before, fortunately that critical hit solved this situation. He also was defeated much faster - 142 frames.

This run finishes with a lot of time left (421), while the previous run finishes with 260 time left, losing 158 frames. The reason is because I didn't die this time, because of that skip on the previous room.

At least room 2 is 62 frames faster at the end

Total: 826 frames gained; 3763 frames ahead

Level 5: 2161 frames faster

Stage 16 - No gameplay improvement. And once again PRG0 version removes a lag during transition to stage 17.

Well, so 1 frame gained, 3764 frames ahead

Stage 17

  • First room: same strategy from any% run.

7 frames faster

  • Second room: same scroll glitch from any% run, and I got an extra lag during transition to stage 18.

437 frames faster

Total: 444 frames gained; 4208 frames ahead

Stage 18

  • Dying before fighting Dracula is still faster.

Unfortunately PRG0 got a lag frame while PRG1 isn't.

  • Before the battle starts, spawning a platform is a very good idea.

  • Normally I can manipulate the first cycle by entering Dracula battle a frame later, but this time it's not necessary, because I already got a good RNG for the next cycle too.

Phase 1

Cycle 1 is the same, cycle 2 manipulates Dracula position (through pausing) and closed cape timming, for an extra attack while standing on that raised platform and so finishing cycle 3 (pausing, again) with a single critical hit.

21 frames faster

Phase 2

  • Attacking while standing on this raised platform is much faster than attacking on ground.

  • Before finishing phase 1, with another pausing the cookie monster can stay on the ground for more time.

  • It's possible to force the cookie monster to jump earlier by pressing (on certain frames) or holding B button - pausing also determines if he will jump lower or higher.

  • By forcing fireball attack pattern, cycle 3 lasts longer, saving some seconds.

Attacking when jumping while avoiding those candles is frame perfect.

443 frames faster

Stage 18 is 453 frames faster at the end input, but 445 frames after time left countdown.

Level 6: 898 frames faster.

4661 frames ahead of published run (the remaining frames were from most room transitions)

Other comments

If the any% run was improved by nearly a minute, pacifist improvement was more impressive, since 2 years ago I got a very solid improvement, through bruteforce optimization.

After working on both categories, I don't know which game will be next, since I put on hold some of my works due to difficult manipulations. But I still like TASing.

And once again, thanks to InfoManiac for the scroll glitch discovery, scrimpeh, for discovering most of those shortcuts, as well as discovering stair glitch, and EZGames69, for encoding this run.

ThunderAxe31: Judging.

ThunderAxe31: This is yet another big breakthrough! Accepting for Moons as a notable improvement over previous run.

fsvgm777: Processing.

Similar submissions (by title and categories where applicable):