Intro to Okami Speedrunning

Ōkami

Ōkami is an action-adventure game originally released for the PlayStation 2 in 2006, and since re-released and remastered on several other platforms. In terms of gameplay, it's similar to most 3D Zelda games; The player goes through various dungeon-like areas, solving puzzles, obtaining new abilities, and defeating bosses to progress across the world and defeat a final antagonist. This similarity extends to the realm of speedrunning, where various tricks, techniques, and glitches can be used to speed up or skip segments and sequences of the game to reach the end faster.

All are welcome to join the Okami Speedrunning Discord to ask questions and get help getting started.

Versions

Main Article: Game Version

Okami has been released on 7 different platforms: PS2 (2006), Wii (2008), PS3 (2012, first HD remaster release), PC (2017, Windows only, Steam), PS4 (2017), Xbox One (2017), and Nintendo Switch (2018). While each version has its own quirks and differences, these 7 versions can be divided into 2 main groups for the purposes of speedrunning:

• Classic Versions: PS2, Wii, PS3. These 'OG' versions allow you to perform the full KT Glitch, 1 Frame Cutscene Skip, Veil of Mist Warp, and Acid Skip, which are patched in the later versions. The PS2 and PS3 versions only allow you to skip cutscenes in a New Game+ file, and only if those cutscenes were watched in the preceding playthrough. These factors make the OG versions better suited for the NG+ Any% and NG+ All Brushes categories.
  • While the Wii version does allow cutscene skips in NG, it is generally disliked for speedrunning due to its motion-based control scheme. It is also prone to randomly crashing.
  • Of these 3 versions, the PS3 version is generally best, due to its greatly reduced load times and lag compared with the PS2 and Wii versions.
• 2017 Versions: PC, PS4, Xbox One, Switch. These versions have the mid-air KT Glitch patched, though this can be worked around by performing Bork KT's. 1 Frame Cutscene Skip, Veil of Mist Warp, and Acid Skip are also patched in these version. These versions allow you to skip any skippable cutscene even in New Game. These factors make the 2017 versions preferable for the NG Any% and NG All Brushes categories.
  • As mentioned above, the Wii version does allow cutscene skips in NG, and even has access to the glitches patched from the 2017 versions. Despite this, it is generally slower for NG runs due to its higher load times. Its motion-based control scheme can make optimal execution far more difficult, and its instability can cause runs to be lost to crashes relatively often, especially for longer categories.
  • All of the 2017 versions are basically equivalent in terms of load times. This is because each loading zone has an effective minimum load time, which each of these versions can easily meet under most conditions.
  • These versions are virtually identical in terms of gameplay, including available tricks, skips, glitches, etc. See the main article for more details on version differences.
  • Notably, the PC version suffers from a glitch with the Steam API, where most inputs are not recognized for a single frame when switching contexts/Action Layers. This does not apply to keyboard+mouse controls. See the PC Section of the main article for a more detailed explanation.

Categories

There are currently 6 major categories for Okami speedrunning:

Category Name	Best Version(s)	Typical Length	Objective	Core Appeal
NG Any%	PC, PS4, Xbox One, Switch	2 - 5 hours	Beat the game as fast as possible from a new file.	This category offers a huge range of available routes scaling in difficulty from the Beginner Route up to the hardest category at top level. There are few true glitches, and nearly every skip requires a creative and context-sensitive application of game mechanics. Without the free resources of NG+, combat and item/Yen routing are key to success.
NG+ Any%	PS3	1h 15m - 2h	Beat the game as fast as possible from a New Game+ ("Start from the Beginning") file.	This is the shortest, most broken category. Through heavy use of the KT Glitch, anything that can be skipped is skipped. The String of Beads and Exorcism Slips are used to greatly speed up combat. Good movement and consistency at the KT glitch are some of the most important factors to improvement.
NG+ 2017	PC, PS4, Xbox One, Switch	1h 20m - 3h	Beat the game as fast as possible from a New Game+ file, but only on the 2017 versions.	This category can be run in a variety of ways. With Bork KT's, virtually the same route as on PS3 can be used, though this requires even greater consistency due to BKT's being double-frame-perfect. If BKT's are avoided, a wide variety of skips are still possible using only Slope and Liquid KT's. This turns the category into a sort of compromise between NG and NG+. KT's can be omitted entirely, effectively turning it into a version of NG Any% without any need to worry about combat or item/Yen routing. This is sometimes done while learning other aspects of the NG Any% route.
NG All Brushes	PC, PS4, Xbox One, Switch	3h 30m - 5h	Beat the game while obtaining all basic and hidden/upgraded Celestial Brush techniques.	This category requires the player to cover far more of the game's normal content than NG Any%, offering many opportunities for interesting routing and unique tricks which have little to no application in a normal run. Even though much of the content that's skipped in NG Any% is required, many of those skips still save time (albeit far less) in NG AB. Sequence breaking is used less to skip content, and more to access late-game resources and abilities to optimize early game areas. Due to the Power Slash and Cherry Bomb upgrades in particular, yen routing is extremely tight throughout this run. This leads to unique movement through areas to collect valuable treasures, as well as creative fight strategies to minimize the use of expensive battle items.
NG+ All Brushes	PS3	2h 20m - 2h 40m	Beat the game from a New Game+ file while obtaining all basic and hidden/upgraded Celestial Brush techniques.	Thanks to the extra objectives, this category is a more complete demonstration of the game-breaking potential of the KT Glitch than NG+ Any%. Routing which wouldn't be possible or logical in any other category gets seen here. This, along with its well-balanced length, makes it entertaining for both viewers and runners.
Top Dog	PC, PS4, Xbox One, Switch	8h 30m - 10h	Beat the game from a new file while fulfilling all of the requirements to obtain the Top Dog achievement (Get all other achievements).	The closest thing to an official 100% run, this category captures the absolute broadest range of gameplay of any major category. Due to its length, an up-to 9 hour break (to sleep and eat) is permitted once during the run. While sequence breaking is limited by the wide array of requirements and objectives, a number of unique tricks and routing concepts can still be applied to optimize this immensely complex run. This run features objectives such as collecting every Stray Bead, fishing 1 of each fish, feeding 1 of each animal, collecting enough praise to max out all abilities, etc.. Notable omissions include not needing to feed 100% of all animal types, obtain every weapon, or obtain every Sun Fragment.

Rules

The most current and up-to-date rules for Okami Speedrunning can be seen by clicking the "Rules" button at the top right of the leaderboards located at https://www.speedrun.com/Okami. This section is a basic overview of the most important rules for newcomers. If you have any questions, see the full rules, and feel free to ask the leaderboard moderators in the Okami Speedrunning Discord.

Keep in mind that these rules apply exclusively to runs submitted to our leaderboards. You are entirely free to run the game however you find fun, under whatever rules you prefer. You can also do runs which break some rules just for practice, and submit valid runs later.

Timing

All major categories are timed via the in-game timer (IGT), which is shown upon beating the game on the "Total Results" screen. This is done because the in-game timer is generally very reliable in its correspondence to real time. It has a few notable quirks:

• The timer technically starts running a few seconds after the options is selected to start a file from the title/file select screen, during the black loading screen. Runners who use a real-time timer when running (for comparing splits) generally start their timer upon pressing the button to star the file, so their in-game time is a few seconds shorter than their real time. This can be observed in runs which use a memory-based autosplitter. For example: https://www.speedrun.com/okami/run/y81j1o5z.
• The timer stops as the text for the Total Results screen begins to appear.
• The timer counts time spent in the fan menu, options menu, and on loading screens, but it does not count lag frames. This can cause greater deviations between in-game time and real time.
• The timer generally pauses if the game is suspended/exited (e.g. by pressing the Home button on the Switch version). Abusing this is not allowed, but accidental occurrences (e.g. you unintentionally press the Home button, then quickly resume the game) will generally not result in a run being rejected, so long as its clear no advantage was gained.

Saving

You're not allowed to make any saves at Origin Mirrors during a run which you submit to the leaderboard. This is because loading a save reverts your in-game time to the point where the save was made, essentially turning the run into a segmented run. You are permitted to Continue the run if you die after running through a Golden Gate (e.g. before a boss), since this preserves in-game time.

On the Total Results screen at the end of the run, your number of Saves is displayed. If you don't make any saves during a run, this number will be 0 for New Game runs, and 1 for New Game+ runs (regardless of how many saves were made in the preceding play-through). If your submission shows a higher number of saves, it will unfortunately be rejected on the assumption that it is segmented.

Recording

For runs which aren't within a certain range of the World Record time in their respective categories, you can submit the run to the leaderboard with only an image (directly captured or taken off the screen with a phone camera) of the total results screen, showing the Play Time. This allows new runners to easily complete runs and record their times without having to make a recording set-up which meets all of our standards for proof. While not required, new runners are still encouraged to provide video recordings if they are able, if only so they can compare to their own previous times.

The full proof standards come into effect when your time is faster than 110% the length of the current WR in that category. For example, if the WR in NG Any% is 2h 03m 34s, then any NG Any% submission with a time of 2h 15m 55s or faster would require full proof, while times of 2h 15m 56s or slower could be made with only a picture of the Total Results time. You can see the full proof standards in the leaderboard Game Rules, under Recording, but here is a brief summary:

• You have to provide a complete video recording of the run from file start to Total Results.
  • A camera recording a TV/monitor is okay if all the other requirements are met, but
• The recording can't have any major disruptions (missing sections) of video or audio.
  • Minor video stuttering (intermittent frame losses) may be accepted at the moderators' discretion.
• In-game music must be enabled and audible throughout the recording.

Turbo

Turbo controllers or software (tools which send a simple, rapidly repeating on-off-on-off signal when holding a button) are permitted, so long as they are only used for the following actions:

• Opening/closing the Fan menu
• Skipping Cutscenes
• Jumping
• Swimming rapidly

Your submission must disclose whether turbo was used, and the description must include an explanation of how it was used during the run.

SSD in PS3

You are allowed to put an SSD into your PS3 for improved loading and more consistent KTs. However, your submission must disclose when SSD is used.