More Phoenotopia on my mind (Whimsical Weekend #21)

The root cause of this sudden flash (no pun intended) of inspiration is that the Phoenotopia dev blog announced in December 2017 the existence of a dedicated Discord server. The community grew quite a fair bit in the first few months, and there was even enough interest in speedrunning the game for a #speedrunning channel to be created. This made me consider that perhaps there is more to be done for the leaderboard.

For starters, there was talk about doing any% without buying the Iron Hammer. Low% would be an inaccurate name for the category because Phoenotopia’s percentage counter is independent of the Iron Hammer, but when I thought back to the Most Dangerous Arsenal miscellaneous category idea, it dawned on me. That’s when I decided to stop saving additional categories until I do runs of them (and that’s kinda unfair in the first place) and establish two new miscellaneous categories:

  • Most Dangerous Arsenal – Based on a medal of the same name, this category involves completely filling all the TOOLS and MISC slots in the pause menu. (P.S.: The armor slot must contain Ancient Armor.)
  • Least Dangerous Arsenal – Based on the previous category, this category involves beating the game with minimal equipment. This means that only the following items may occupy the TOOLS and MISC slots:
    • Wooden Bat
    • Slingshot
    • Artifact
    • Rolling Technique Scroll
    • Bombs
    • Rocket Boots
    • Javelin
    • Floatation Donut
    • Green Bracelet / Bandit Boss / Golem Head

I also established a No Major Glitches subcategory that prohibits the use of Loot Duping, Pot Head Clipping, and Door Pushing.

On top of that, remember when I said that All Heart Rubies was a bad idea? Well, I considered the coexistence of All Moonstones and Most Dangerous Arsenal. They are more similar than you’d think (you have to get 25 moonstones (for one of the Asteroid Rocks for the Morning Star) to complete MDA), and perhaps 100% and All Heart Rubies are in the same boat (you have to have 40 moonstones to obtain the Heart Ruby at Hidden Village), so…yeah, now All Heart Rubies is also established as a miscellaneous category.

In light of the newly established categories, I wanted to do a Most Dangerous Arsenal run, but several attempts yielded nothing satisfactory. Then February came along, a telltale sign that Flash Marathon III was drawing near. Thus, I decided to put MDA on hold and focus on All Moonstones. At first I figured that the category could use a bit of rerouting, which later led me to wanting to PB before the marathon. (If I ended up not getting the PB over this weekend, I would have discarded the idea and focused on practicing the no-reset and commentary-related aspects of the marathon run.) Thankfully, I ended up doing so.

I’m honestly still kinda irritated about how Forgotten Forest pressure plate puzzle went, but everything else was totally acceptable.

So, next up is practicing for the marathon (and, obviously, running in said marathon). Afterwards, I’m considering transitioning to Any% No Major Glitches (which, if there ever happens to be a fourth Flash marathon, will probably be my next marathon category of choice).


 À la prochaine! (Until next time!)


Flash Game Speed Marathon: My marathon debut (Whimsical Weekend #11)

Two months ago, I talked about submitting for Memeathon X, specifically with an unofficial category of Phoenotopia that I call “69 HP RTA,” but my submission didn’t make it into the marathon.

Fortunately, not long before the end of March (I don’t remember the exact date), I happened upon a more esoteric submission form posted by Twitch user LaserTrap_ in the 360chrism Community Discord. That form was for a marathon initially named “Flash Games Done Quick,” but they had to change the name mid-marathon because “Games Done Quick” is trademarked. Regardless, considering that a vast majority of the games that I currently speedrun are Flash games, of course this marathon would be the perfect fit for me. So, I submitted Phoenotopia 100%, Rock Bottom All Levels, and Chompy All Levels; and I ended up being one of only seven runners in the marathon (including LaserTrap_).

It was my first time doing live commentary while speedrunning (granted I rehearsed a bit beforehand), but I’d say it went over pretty well. Highlights can be viewed below:

Phoenotopia [100%] in 1:47:16 —

Rock Bottom [All Levels] in 7:53 —

Chompy [All Levels] in 6:22 —

(Other highlights of the marathon can be found at

Because I rehearsed so little, it goes without saying that there was at least a little rust involved in all my runs. Prior to the week before the beginning of the marathon, it had been six months since my last Phoenotopia 100% WR (1:44:08), five months since my last Chompy WR (5:09), and three months since my last Rock Bottom WR (6:17).

Since the games that I ran are so fast-paced (well, not so much Phoenotopia, but still), I can’t remember off the top of my head where exactly I messed up in each run, but I do recall that I unfortunately didn’t get the 1-minute skip in level 14 of Rock Bottom. Also, apparently my keyboard doesn’t like me pressing down, right, and Numpad 0 at the same time, so I had a tiny bit of difficulty starting the timer for Chompy.

Another thing: I’m not used to talking in general, so running my mouth for practically two straight hours caused my voice to hurt over the weekend. Thankfully, though, it was nothing major.

Bottom line: I dragged myself into a change of pace by becoming part of an esoteric marathon, and the highlights linked above are the results.


 À la prochaine! (Until next time!)

New PC + speedgames (Whimsical Weekend #8)


I never really made this evident in any of my recent Vouiv-review posts, but I recently built a new PC! I spent roughly $1600 on all the relevant components, and it’s been working very well ever since I got the Windows 10 Home product key and Ethernet cable.


This is my current setup. It may not be the most complex setup in the universe, but it gives me just what I need. As you may or may not be able to tell, in front of my bed, just near the window, I have a swivel chair facing a single table with the tower off to the right (which unfortunately hides the insides from view) and keyboard and mouse + mousepad in front of a dual-monitor setup on top. Because the monitors are of different sizes and odd angles (the 25″ one being HP brand and the 22″ one being Sceptre brand), I had to use Heroscape tiles (from my childhood) to adjust the heights and angles to my liking.

Also, since it’s a new PC with a resolution differing from that of the MacBook Pro that I would otherwise normally use, I decided that I would use a different wallpaper as well.

It’s from the same series as my laptop and phone wallpapers (that being Mondaiji-tachi ga Isekai kara Kuru Sou Desu yo), and this particular image is from volume 11 of the light novel.

Another funky thing about the dual monitor setup that I have is that not only are the monitors of different sizes and odd angles, but they also produced different sorts of output in their initial state. Thus, I had to tinker with the settings of the 22″ monitor in order to make the outputs as similar as they can be. Even so, I’ve noticed that it’s easier to see darker colors on the 25″ monitor (a fact that I first came to realize while playing around with Phoenotopia Awakening Demo v0.06b), although the 22″ monitor actually has a speaker (whereas the other monitor does not).

How great it is to have a dual monitor setup with Windows regardless; it allows me to multitask better than I ever have before. For example, I can have a Twitch stream up on the 25″ monitor, some other form of web browsing (which sometimes involves writing, such as this post to this current instant of typing) on the left side of the 22″ monitor, and email and/or some other utility on the other side of the 22″ monitor. (Sometimes, I even have Notepad in the bottom-right corner with the rightmost Chrome window in the top-right.)


Also thanks to the new PC, I find it easier to toy around with OBS (Open Broadcaster Software) for better quality footage of the PC games that I usually speedrun (where I would otherwise have to use the MacBook Pro). I currently don’t have reliable sources for external audio or keyboard overlay with the PC like I do with the MacBook Pro, but the benefits far outweigh the drawbacks. I mean, just look at this sample footage in the form of a demonstration of most of the first Prince Tower segment of Phoenotopia 100%:

The splits are just there on the side to show that I can use LiveSplit (while I would normally use Llanfair on Mac) to show all 35 splits at once. (I probably could have done something similar if not the same with Llanfair before, but I couldn’t be bothered to figure out that sort of thing.)

With that in mind, I might actually go back and improve my times in Phoenotopia (and maybe branch out to that one “Most Dangerous Arsenal” category that I mentioned in an earlier post, or maybe even the glitchless subcategory) with the new PC, and I’m heavily considering post-commentating the any% run so that I have a helpful video resource for that category (in a similar vein to the tutorial series that I made for 100%). However, the critical issue in that department is that I would have to get used to the mechanical keyboard. As it stands currently, I have a harder time performing brief inputs (notably short-hopping) with the mechanical keyboard than I have ever had with the MacBook Pro keyboard, so…yeah.

Phoenotopia Awakening Demo v0.06b

But hey, the original Phoenotopia wasn’t the first thing that I considered running on the new PC. As a matter of fact, not too long after I built the PC, demo version 0.06b of the upcoming remake (Phoenotopia Awakening) was released, and the first thing that I resolved to do was write a guide on the demo. The guide is totally completed (at least as far as I’m concerned) and can be viewed here, but…well, the developer’s most recent blog post says that download links for this particular build of the demo will be closed.

Regardless, while the game was fresh in my mind, of course I was the one to submit it on, and that even led me to submit a series entry for Phoenotopia. On the demo v0.06b leaderboard, I established three separate categories: any%, All Collectibles, and All Pearlstones. (I decided to make the third one a miscellaneous category because it’s less sensible of an objective than any% or All Collectibles (similarly to how All Moonstones is a miscellaneous category of the original Phoenotopia).)

By the way, All Collectibles constitutes the following (as far as I know):

  • 7 Anuri Pearlstones (should be left with 1 after unlocking all 6 frog seals)
  • 2 Heart Gems
  • 2 Moonstones (specifically from the breakable tomb and the Giant Slime)
  • 2 Antique Bracelets
  • Energy Gem
  • Slingshot
  • Crank Lamp
  • Bombs

My PBs in the three categories are 4:29, 10:27, and 6:58, respectively. My any% PB was initially 4:59, but a competitor with a 4:56 led me to rethink the route and improve my time.

Here be videos:

Rock Bottom

I don’t think I ever mentioned Rock Bottom in any of my blog posts, but that’s another game that I submitted to To be honest, though, I didn’t continue beyond 7:45.97 for a long time, and it was back in early November 2016, before I had even bought the parts for the new PC, that I had established that time. Meanwhile, SRCom user CreepinAtMyDoor was out shredding my time to bits, even accidentally discovering a skip in level 14 that saves more than a minute, and ended up with a time of 6:22.1 before the end of that November. I felt at the time that WR was beyond recovery, but then, about a week ago, I had suddenly regained the inspiration to take it back. One day I got 6:35.43, the next day I got 6:30.67, and then finally I was able to take back the WR with 6:17.87 (after a 6:25 run that I opted against uploading).

This game is difficult as crud to optimize, so I am very happy with this time, even though I got spiked in level 12 (and at a very unconventional point too). Man, my heart was racing by the end of that run. Imagine what my reaction would have been if the recording had external audio. (The skip is at 4:15 in the video, by the way. If you miss the hard floor and hit the spongy surface, the run is pretty much over.)


Another speedgame in my repertoire, although I haven’t run it since the time I submitted it to My current record is 5:09.73, which notably includes two deaths in the 21-30 segment and a botched entry into level 11. So, even though I felt at the time that it was a solid run, there is still room for improvement, so I might actually try to get better.

Diamond Hollow II

This one is more of an up-in-the-air case. I don’t own the leaderboard of this game, and the owner of the leaderboard has been inactive for over a year (and it doesn’t help that said owner has no links to social accounts in his/her profile page). Currently there are four established categories:

  • Story Mode – Any% (all seven levels of Story Mode, anything goes)
  • Story Mode – New Save (all seven levels of Story Mode starting from fresh data)
    • The leaderboard owner also imposed a rule against “glitches,” but I honestly believe that the rule is arbitrarily imposed, especially considering the category name does not in any way imply the exclusion of glitches. Besides, I haven’t encountered anything resembling a “glitch” in the game, and I tried asking about it in the forum 9 months ago but still haven’t received a response.
  • Boss Rush – Normal
  • Boss Rush – Heroic

About two weeks ago, I submitted runs for Story Mode Any% and Boss Rush Normal, which can be viewed below:

I specified both real and in-game times because it was ambiguous which to go with, and I used real time as the submission time for the leaderboard because there was only one field.

I might do Story Mode New Save in the near future, but that seems like it would be difficult if not impossible to route out. (I’m thinking it will help to know all the Red Diamond locations, though.) Plus I would have to delete save data, and I’m reluctant to do so considering how long it took me to get a 100% completed file.

As for Boss Rush Heroic…no. No. Never. Forget speedrunning that category; I can’t even beat it casually!

Anyway, that’s all I have to say for now.

Nowi Wins À la prochaine! (Until next time!)

Phoenotopia speedrunning: resting on my laurels…for now (Whimsical Weekend #4)

I know it’s unusual to with a post one week after the previous post, but I just had to get this off my chest (so to speak) as soon as I could.

Screen Shot 2015-08-14 at 11.25.35 PM

Phoenotopia is a 2014 Flash 2D platformer developed by Quang H. Tran a.k.a. Quells. This game is hosted on a number of sites such as Newgrounds (the original) and OneMoreLevel (my preference), and playing it is absolutely free.

Since one year after the release of the game, I have written articles on how much I enjoy playing it, as well as my endeavor to do what no one else has done and make an official leaderboard of the game on First off, I wish to go through a retrospective of what I’ve written in the past.

A Retrospective of Former Articles

Spontaneous Saturday 8/15/15: Phoenotopia

This is the first Phoenotopia-related post that I’ve published to this blog, and it’s a simple review consisting of my thoughts on the game and my desire to make it a speedgame (primarily influenced by the “Speed Runner!” medal in the game and my love for the game in general). I also included compendia of purchasable items and enemies / environmental hazards inspired by my go-to walkthrough at the time, as well as this little meme:


Spontaneous Saturdays 9/12/15 and 12/5/15

I only talk briefly about Phoenotopia in these posts, and it’s just thinking out loud about the routing process of any% and 100%.

Phoenotopia speedrunning (Thought Dump Thursday 1/7/16)

This was after I had finished constructing the primary route for the any% category and was in the process of constructing a primary 100% route. The post contains the following: how I started running the game, how I discovered Loot Duping (a glitch that produces extra spoils from Rai containers and heart pots), and an analysis of my first recorded any% run with a time of 59:22 RTA (1h 2m IGT). Some of the information contained within the post is obsolete (notably claiming that Loot Duping is pixel-perfect when it’s actually a three-pixel window), although the first two videos included are still relevant. The first video demonstrates what I was doing when I first discovered Loot Duping, and the second video involves combining the pixel-based and timing-based methods of Loot Duping.

Potpourri feat. winter 2016 anime

As the title implies, this post was mostly centered around the interesting anime series that aired during the winter 2016 season (Phantom World, Dagashi Kashi, and KonoSuba), although I did briefly mention Phoenotopia. Specifically, I tried out the French version and got a new PB of 55:14 RTA (58m IGT), but I came to realize that the French version is roughly 4 seconds slower in any% due to longer blue text boxes (i.e. the sorts of text boxes that appear when you obtain an important item) and differing dismissal timings of said text boxes (more consistent in the French version, but overall slower). I also publicly wrote some food for thought regarding the routing of 100%.

Phoenotopia 100% speedrunning (Thought Dump Thursday 2/4/16)

After having demystified any questions that I may have had while routing 100% and completing the primary route, I did my first run of the category and ended with 2:06:39 RTA (2h 9m IGT). This was back when I thought that doing the Loot Dupes in Sunflower Road was a good idea. Therefore, some of the techniques mentioned in the post are obsolete, although a majority of them still apply to the current route.

Another glitch in Phoenotopia + current speedrun PBs (Thought Dump Thursday 3/3/16)

This was the point at which I discovered a glitch other than Loot Duping. A day or two after my first 100% run, I was mucking around with pots in Prince Tower and found a way to get the moonstone on the 7th floor without having to use the Rocket Boots. Later on, as I wrapped my head around what I had stumbled upon (a phenomenon that I ended up dubbing “Pot Head Clipping” (even though I don’t do drugs)), I decided to record a clarifying demonstration and include it in the blog post.

To explain, if Gale has a harmless solid object sitting on her head and independently affected by gravity, trying to jump will result in Gale being warped to the nearest available space in any of the four basic directions: up, down, left, or right. At the time of the blog post, I thought of Pot Head Clipping as abusing this concept of warping to such a degree that would allow Gale to clip past barricades and walls. I also thought that Pot Head Clipping had no practical use other than the alternative way of getting that one Prince Tower moonstone, but that ends up changing later on.

Glitch aside, I also mentioned how I got 54:52 RTA (57m IGT) in any% and 1:59:43 RTA (2h 2m IGT) in 100%.

Phoenotopia Forgotten Forest navigation + yet another glitch (Thought Dump Thursday 3/31/16)

Let’s be honest here: Forgotten Forest was atrocious in my 1:59:43 run of 100%. By the time of this post, however, I managed to figure out ways to take one of the worst parts of the run and make it faster and less risky. First off, I found another practical use for Pot Head Clipping, and that is to bypass the first locked door without having to kill any of the Arc on the screen of that door or the Plant Dog one screen to the right thereof. Second, in order to bypass the locked door on the screen with the Big Robot 2.0, I tried unsuccessfully to set up a Pot Head Clip (in the traditional manner), but then I discovered something even more astounding: While I was fiddling around with two of the boxes taken from the lower path, one on top of the other while slightly farther forward, I was overcome with disbelief when I found that pushing the duo into the locked door actually caused it to move from its normal position. Like, how is that possible!? When I look back on the concept of a locked door being pushed, I still can’t believe it. (I simply referred to the phenomenon as “Door Pushing.”)

At any rate, the new glitch and the new application of the old glitch allowed me to move through Forgotten Forest more elegantly, although I have picked up some new tech since then. Regardless, here is the video demonstration:

(Also, I would not attempt the Forgotten Forest JBJ (Javelin Bomb Jump) in a real run, let alone succeed as well as I did in the recording above.)

Phoenotopia speedrunning: status update (Thought Dump Thursday 4/28/16)

I talk about how I got 54:02 RTA (56m IGT) in any%, and I also rambled about categories that I was considering beyond any% and 100%, as well as the imminent possibility of improving my 100% time.

Back from hiatus! Anniversary potpourri (Thought Dump Thursday 6/2/16)

This blog experienced a hiatus for pretty much the entire month of May because I felt like I wasn’t taking things seriously enough (specifically overdue assignments and job hunting). When I got back, I dropped a whole load of detail on my 1:48:16 RTA (1h 50m IGT) run of 100%, including some amazing new discoveries with Pot Head Clipping—namely, you can clip through locked doors from the left without placing any objects behind you, and using a container containing a moonstone or inventory item causes Gale to obtain that item upon clipping (and also go through a jumping animation)—and also that Loot Duping is not pixel-perfect but a three-pixel window. It was when I realized the second fact that I went and wrote a Loot Duping Guide containing revised and detailed information on the glitch and all loot containers that can and cannot be subject to Loot Duping.

Nothing Specific (Thought Dump Thursday 6/30/16)

Just talking about my future plans regarding the speedgame and potential improvements to existing times, resources including a save password compendium for practice/learning, and the possibility of running Diamond Hollow II in the future.

Phoenotopia – How to softlock in less than a minute (Thought Dump Thursday 7/4/16)

Just a little filler post showing off the quickest way to force yourself into a softlock in Phoenotopia. It’s a simple Pot Head Clip in Panselo.

Phoenotopia 100% – New speedrunning guide in the works! (Whimsical Weekend #2)

This is the most recent post that I’ve made on Phoenotopia. It covers a detailed guide of the 100% category and the thought process behind constructing the guide, including some new tech in Prince Tower, specifically on the first floor (image below) and in the second tower box puzzle (video below).

Screen Shot 2016-09-05 at 7.27.45 PM

And Now We Arrive at the Present (This is now)

Transitioning from memory lane, the main purpose of this post is to say that I am on temporary hiatus in terms of speedrunning Phoenotopia. That is to say, I have established every category that I feel like establishing, and I have run times that are satisfying enough for me to stick with until they are contested. Specifically…

Any% record

For some reason, I don’t think I ever mentioned that I got 51:53 RTA (54m IGT) in any%.

The run was pretty shaky in some areas (such as failing Golem Skip), and it was before the time that I discovered the “chandelier climb” (as I call it) on Prince Tower floor 1, but it was a good run overall.

100% record

As for 100%, my most recently achieved time is 1:44:08 RTA (1h 46m IGT), and that was after I constructed not only a written guide but also a video tutorial series (which I’ll explain later).

This was also a pretty good run, even though it had a sloppy beginning and some minor execution faults throughout. I did manage to get all four skips in Ancient’s Crater regardless, and I did the new tech in Prince Tower, so that alone already makes it better than my 1:48:16 run.

All Moonstones record

All Moonstones is the first and only miscellaneous category that I decided to establish. I was initially reluctant about it due to the riskiness of the route, what with the whole run being under bare minimum conditions (no extra items other than the Iron Hammer and Lamp), but then I figured that it was the only miscellaneous category discussed in Thought Dump Thursday 4/28/16 (or thought of externally) that I saw fit to officialize. Why did I refuse to officialize the other categories? Well…

  • 99 HP RTA (Diamond Skin) / All Heart Rubies: As I mentioned in the aforementioned Thought Dump Thursday post, there is a Heart Ruby that requires 40 moonstones to obtain, which makes a category like this not much different from 100% in my opinion. So, I am completely opposed to officializing a category like this.
  • Max HP no Moonstones: I initially thought very highly of this misc. category idea because of how it was pretty much the safest middle ground between any% and 100% that I could think of, but I ended up feeling unsettled by how arbitrary the category actually is, particularly in its nomenclature.
  • All Tools: Why? Why did I even begin to think this was a good idea? Similarly to the previous category, the nomenclature just makes it seem…off.
  • 88%: At one point, this sort of idea crossed my mind: a category that involves collecting all Heart Rubies and all Moonstones but nothing else. The issue is that it’s difficult to name such a category, and 88% was honestly the best I could come up with…and that’s saying something, because the name “88%” isn’t very descriptive of the category itself.
  • Most Dangerous Arsenal: I just thought of this recently: a category in which all Tools and Misc. Items are obtained (based on the medal “Most Dangerous Arsenal,” for which the category goal is based on the condition of obtaining the medal). This might actually have potential for a misc. category, but I shudder to imagine Nebula Armlet with 20 HP.
  • Glitchless subcategory?: I was considering adding a glitchless subcategory to the leaderboard, which would forbid the following: duplication of Rai or heart drops by breaking a container multiple times on the same frame; deliberately placing anything on top of Gale’s head; and pushing objects into locked doors. (The restrictions are meant to encompass Loot Duping, Pot Head Clipping, and Door Pushing, respectively.) However, I can’t help thinking how much of a pain it would be to have to deal with 100% or All Medals glitchless (or All Moonstones, which would actually require backtracking through Prince Tower since PHC is forbidden), so I just decided to discard the idea altogether. I might change my mind if someone else were to record a glitchless any% single-segment run with an in-game time of less than an hour, though.

Anyway, I did end up doing an All Moonstones run in spite of my initially reluctant self, and I ended up achieving a time of 1:13:17 RTA (no in-game time because the category does not require beating the final boss).

Obviously not perfect, and Forgotten Forest + Hidden Village can stand to be routed before the Daea backtrack, but it’s a miscellaneous category, so I don’t plan on improving any time soon.

All Medals record

I’ve had a route in the works since the time I wrote Thought Dump Thursday 4/28/16, but I never got around to finishing the route until after I did the All Moonstones run. It took me less than a week to complete the whole run, and I ended up with 2:09:18 RTA, which is surprisingly close to my first 100% record.

I have a strong feeling that it’s possible to obtain all medals in under 2 hours, but I currently have no strong desire to improve my record.


Over the course of routing and running the categories above, I have been compiling a number of resources throughout the journey, such as Q&As (which I used to call GQIGAs (Got Questions? I Got Answers)), written routes, detailed information on glitches (through blog posts, guides, or otherwise), and perchance various specific compendia. The following remain relevant to this day:

  • leaderboard – The “official” leaderboard for the game, which I myself submitted for approval, and of which I am currently the sole moderator.
  • Any% Q&A 4.0 – Q&A for my most recent any% run.
  • 100% general Q&A – That’s what I called it at first, but it’s actually the 100% Q&A 4.0, which relates to my most recent 100% run.
  • All Medals Q&A – Q&A for the All Medals category (version 1).
  • Loot Duping Guide – An elaborate guide on the method behind the madness of Loot Duping: what it is and where/how it can be applied.
  • Save Password Compendium – A collection of save passwords for practicing/learning the any% and 100% categories.
  • Any% guide – A guide to the any% category. I have kept this guide up to date between the first time I decided to run the game and the most recent run, so note that it does not contain information regarding the chandelier climb.
  • 100% guide – An elaborate written guide to the 100% category.
  • 100% video tutorial playlist – Between Whimsical Weekend #2 and this post, I had the wild idea of coming out with a video tutorial (in spite of myself) of the 100% category.
  • All Moonstones planning/route – The planning and condensed route (the latter is on page 3) of All Moonstones. No, the route has not been updated since the recorded run.
  • All Medals route – The condensed route for All Medals. I do not plan to make a full guide of this category, because doing so would be rather redundant in my opinion.
  • Splits:


All right, I think that about covers everything: the retrospective, the point that I’ve currently reached, and a set of resources that might help provide further insight. Now that I am temporarily resting on my laurels, I will move on to Diamond Hollow II.

Nowi Wins À la prochaine! (Until next time!)

Phoenotopia 100% “speedrunning” (Thought Dump Thursday 2/4/16)

1/27/18 EDIT: I put “speedrunning” in quotes here, because this post was written just a few days after my first 100% run. For more up-to-date information on the category, use this guide:


Screen Shot 2015-08-14 at 11.25.35 PM

A sort of precursor to this post is Thought Dump Thursday 1/7/16, wherein I talked about my thought process behind speedrunning this game in the first place, presented my first recorded any% run of the game, and elaborated upon the techniques used in the run (all of which are also in 100%).

First off, in response to my question in Thought Dump Thursday 1/21/16 regarding the Floe moonstone, you do indeed have to do the following:

  • Free Fran, collect at least 5 moonstones, and talk to Fran at the table outside her lab
  • Complete the slime extermination quest at Cross Roads and claim your reward (not necessarily after the previous step)
  • Talk to Gill at Great Walls (left side of the second watchtower)
  • Head back to the treehouse at Cross Roads and talk to Floe to claim the moonstone

The slime extermination requirement makes sense, but talking to Gill…seems like a weird step to me, especially considering he says nothing about Floe when you talk to him.

Second, as a reminder, the collection requirements for 100% are as follows:

  • All 23 Heart Rubies (2% each; 46% total)
  • All 42 Moonstones (1% each; 42% total)
  • Blood Ring (2%)
  • Nebula Armlet (2%)
  • Lucky Belt (2%)
  • Ancient Armor (2%)
  • Morning Star (2%)
  • Ki Spear Technique Scroll (1%)
  • Giving Translucent Meat to the chef in Daea (1%)

Third, and most importantly, I have actually completed a recorded run of 100%.

Relevant Pastebin (please read):

Other relevant links:

Note that I will be using timestamps very frequently throughout this writing, and they pertain to the video, not the splits.

Okay, so apparently, 100% is about twice as long as any%. I mean, this time is definitely improvable, notably because this is actually my first run of 100%, not just my first recorded run. It’s still on my mind to modify the route in some parts to incorporate extra honey purchases, but I’m still not sure how exactly that sort of thing would play out.

Loot Dupes

First off, most of the Loot Dupes in any% are done in 100%, with the removal of a few and the addition of a few. In particular, the dupes in 100% are:

  1. Duri Forest interior [30 -> 60] (4:04)
  2. Duri Forest exterior [15 -> 30] (9:23)
  3. Panselo leftmost building [25 -> 50] (10:33)
  4. Panselo dojo [20 -> 40] (12:19)
  5. Sunflower Road east [20 -> 40] (13:18)
  6. Sunflower Road west [30 -> 60] (14:11)
  7. Bandits’ Lair storage room [50 -> 100] (20:39)
  8. Misty Gorge inn [30 -> 60] (28:20)
  9. Thomas’ Lab [40 -> 80] (33:22)
  10. Daea sewers [25 -> 50] (39:45)
  11. Daea prison switch room [30+30 -> 60+60] (42:24)
  12. Ancient’s Crater [50 -> 100] (1:13:02)
  13. Atai palace [45+45 -> 90+90] (1:19:49)
  14. Panselo lab [30 -> 60] (1:27:13)
  15. Prince Tower second tower [50+50 -> 100+100] (1:46:32)

These dupes, along with the rewards provided by various quests (40 Rai for the slime extermination, 30 Rai for delivering Mikan Fruit to the potion shop owner in Atai, 15 Rai for giving Translucent Meat to the chef in Daea, 300 Rai total from the first visit to Fran’s Lab, 40 Rai for delivering a Sandwich to the guard in Great Walls, 15 Rai for delivering a letter to Farmer’s Block, and 150 more Rai from the second visit to Fran’s Lab) and some chests that are broken normally (which includes the one in Panselo at 1:27:56 and was also planned to include the ones in Forgotten Forest at 1:56:30 and 1:58:21, but apparently I didn’t end up needing those chests), are just enough to cover all ~1950 Rai needed for the run. (I indicate that it’s an approximation because Gold Bar pricing seems completely random to me.)

It may seem weird that I do the Sunflower Road dupes, seeing how they are the hardest dupes in the game and the only ones that rely on enemy cooperation, but keep in mind that it is required to collect a grand total of 280 Rai (for the Chocolate Protein Shake, Rolling Technique Scroll, Iron Hammer, and Talkinator Juice) before reaching Atai for the first time. Also, if not going with those two, the best substitute would be to perform the dupe in the frog room in the Duri Forest dungeon (for 80 Rai) and the dupe in the second-rightmost house in Panselo (for 30 Rai), which wastes 2 minutes for the frog room alone and a few more seconds for the latter dupe. Compared to this, the potential time taken to dupe the chests in Sunflower Road (which each happen to be in the path of one moonstone) is trivial.

Also, I do the Atai palace dupe a little earlier than I had routed initially. I don’t know whether this is a good or bad idea, because it is in the pathway of the first Atai visit based on how I’ve routed it, but it means having to dedicate a slot to the High Quality Silk (which isn’t exactly necessary, but hey, since it’s part of this dupe and gets an easy 50 Rai (or 90 if you feel like wasting time), why not?). Now that I think about it, I’m considering routing the moonstone outside of Atai palace to after delivering the rejection letter so that performing the dupe at that point would make more sense and, at the same time, there is no need to carry the High Quality Silk all the way from Atai 4 to Atai 5.

On a final note, it may seem obvious, but I get quite frustrated over those moments while trying to get the exact right positioning when Gale ends up one pixel off every time. It happens quite a few times in this particular run, including Bandits’ Lair storage room, Daea sewers, and Panselo lab.

Man…this is a lot more talk about Loot Dupes than I expected. Now, on to the more important stuff.

Extra techniques in 100%

There are some time-saving techniques in the 100% route that are not seen in the any% run, and they are as follows:

  • The box stacking that I do (horribly) at 4:26 saves having to get the Heart Ruby while backtracking through Duri Forest. If done perfectly, this box stacking saves one or two minutes (I don’t remember) over normal backtracking, and that’s factoring in the possibility of a stamina boost during the backtracking.
  • For the Cross Roads Heart Ruby (at 34:51), pulling out a bomb right when entering the secret entrance, if done correctly, allows it to detonate the exact moment you place it on the set of boxes (which saves time over destroying those boxes by spamming the X key).
  • After the slime extermination, which starts at 35:10 and ends at 35:23, it is significantly faster to exit to the right and re-enter the area than to head all the way back left to the inn.
  • At 41:57, I show how it’s possible to complete the timed switch moonstone puzzle in Daea prison without a stamina boost: place a bomb on the top-right switch, do the same with the top-left switch, melee the bottom-left switch, and perform a full-jump midair attack on the bottom-right switch. The last maneuver is undoubtedly the hardest, but thankfully I got it first try.
  • The moments at 45:26, 47:58, and 48:35 are proof that Flying Bombs are vulnerable to the Ki Spear Technique. It’s really weird how that one little technique is the only thing that will cause those bombs to explode instantly.
  • I do the Heart Ruby puzzle in Daea prison (at 46:05) using three metal boxes. These three boxes are stacked on top of each other, and the top box is slightly farther right than the other boxes such that using the Ki Spear Technique on top of the middle box will cause the cracked ceiling to blow up. After that ceiling is blown up, the bottom two boxes and the Javelin should be sufficient to make it up to the cracked wall behind which the Heart Ruby lies.
  • Hitting the switch at the entrance of Prince Tower (at 46:54), along with exiting and re-entering the area, is pretty much necessary because it makes backtracking way easier. (I once neglected to do this in a rehearsal playthrough, and I regretted it wholeheartedly.)
  • In Dread Lands, the most efficient way to kill the Broken Golem guarding the moonstone pot at 1:00:47 is to hit it from afar using the Ki Spear Technique.
  • 1:04:54 shows off the Ki Spear Technique at its utmost usefulness. I screwed up the execution quite a bit, but the ideal way to get this bugger of a Heart Ruby in Mul Cavern is as follows:
    • From the uppermost orange mushroom, throw a Ki Spear to the left and destroy the uppermost hive.
    • From the green mushroom below the previous mushroom, drop a bomb right off the left edge and wait for it to explode. This is the main part that I screwed up.
    • Destroy the rightmost hive by throwing a Ki Spear or placing a bomb right in front of it.
    • Once everything is clear, for better or for worse, drop down to the purple mushroom and throw a Ki Spear to the left. This will end up destroying the hive on the far left, the ultimate barricade guarding the Heart Ruby.
    • Take the upper route; don’t even think about the lower one like I did.
  • At 1:26:28, I use a Bomb Jump to get into the attic of Gale’s house, which is the quickest way to do so without a stamina boost. The best part about this strategy is that any health loss will be trivial thanks to the Heart Ruby restoring all HP.
  • The easiest way to lower the barricade protecting the way to the Ancient Armor (as demonstrated at 1:31:13) is to throw the Javelin at a relatively low height and shine the Artifact upwards and slightly to the left.
  • Speaking of Ancient Armor, the section with the grounded turrets (starting at 1:31:34) may seem daunting, but a way to facilitate this section is by picking at the lower turrets using the Ki Spear Technique.
  • Because of the odd requirement to give Translucent Meat to the chef in Daea, part of the 100% route is farming for this drop from the Ghost Wasps (as demonstrated at 1:33:24). From the north entrance of Misty Gorge, walk until hitting the first arch from the right, make your way all the way back to the second-highest ledge on the far right, and spam the Slingshot as fast as possible.
  • Demonstrated correctly at 1:35:20 is a faster way to get up to the Spiky Caterpillar cave of Misty Gorge. Throw the Javelin at maximum height, use a Bomb Jump to pop up on top of the Javelin, and perform a JBJ (Javelin Bomb Jump, which involves placing a bomb on the Javelin and, right before the bomb explodes, doing a quick jump-and-throw-another-Javelin maneuver) to gain just enough height to jump up to the top. This saves time if done correctly, as it eliminates the requirement to go left one more screen and perform a series of hovers to the right.
  • At 1:40:12, I demonstrate a neat little way to exit Daea from the sewers without having to go all the way around.
  • The futuristic puzzle in Dread Lands (at 1:52:36) is much easier than it seems. You may think you have to set up a path for a Rolling Bomb to hit the lower timed switch, but, as I demonstrate, you can actually just place a bomb right above the switch, and I guess the bomb’s explosion radius is barely wide enough to hit the switch from above. Quite a design flaw there, I do declare.
  • There is a key in Forgotten Forest that is in close proximity to mark II of the Big Robot. (In fact, the rightmost trigger pixels of the key are very close to those of the robot.) However, as opposed to a near-pixel-perfect method of grabbing this key, using some sort of explosion (preferably that of a bomb), as demonstrated in 1:58:05, allows the key to be sent flying to the right and made much easier to collect without provoking the robot.
  • This is not an extra technique per se, but with the Morning Star and the Nebula Armlet, it is very possible to two-cycle Big Eye, as demonstrated at 2:02:45.

Potential areas of simple improvement

  • I mention my bladder quite a lot. Next time, I should ensure that I have not drunk any fluids within the past hour so that this sort of thing does not inhibit my focus on the run.
  • Some of the navigation and menuing in general was pretty sloppy (some of which was due to me reviewing my notes), and some Loot Dupes were rather slow, but hey, that’s natural for a first run.
  • Like I said, box stacking at 4:26 was horrible. If I had to estimate, I’d say that’s around 30 seconds of time loss.
  • At 18:44, jumping out of the lower area instead of rolling a second time does not actually save time; in fact, it actually wastes the slightest bit of time.
  • At 20:57, I believe the best way to get the moonstone in the Bandits’ Lair “pit of shame” (as I’ve dubbed it) is to start by throwing the bomb at the suspended platform to the immediate right of where you start and, if necessary, placing a bomb on the bottom rubble while waiting for the thrown bomb to explode. Then, after grabbing the moonstone, hit the switch. I’m thinking this change of order would save a second or two, but not much more (if at all) than that.
  • I feel like killing the Red Bandit at 22:06 relieves the pain of having to deal with it later, and this applies to any% as well. Not exactly time save (and probably not even time save at all), but I guess you could call it insured prevention of time loss.
  • I’m considering shifting Honeyed Tumond usage in Bandits’ Lair from 22:41 (after being purposely hit by a Bandit for invincibility frames) to 22:27 (right before exiting the first key room) so that I won’t miss the Red Bandit cycle like I did in this run. (I could save roughly 10 seconds.)
  • I’m also considering shifting the moonstone near the save statue of Daea prison (at 40:09) to after the dog-and-electromagnet room (at 41:52) to make sure that getting that moonstone will be done at pretty much the same time as the safety healing. Also, while waiting for the bomb at the right wall to explode, why not pull out another bomb to make breaking the boxes easier?
  • Also, regarding the Flying Bomb at 45:42, could I not have just thrown a Ki Spear at it?
  • If I ever consider incorporating extra honey into the route, Prince Tower is definitely a place where I would appreciate having it. At 47:00 (and 1:40:22), you can see how slow it is to navigate without honey due to having to wait for items to recharge and not being able to skip past one ladder per jump on floor 6.
  • In case it wasn’t obvious, the crud that I try to pull off at 49:34 is not practical. I should never go for it, no matter the circumstances. In this run, I wasted roughly 2 minutes trying to attempt it.
  • At 59:31, I try to pull off a strat that would allow me to climb out of the pit more easily, but since the Javelin was not at a low enough height and my safety was at risk, I just decided to give it up. If done correctly, this strat saves a few seconds.
  • When grabbing the O-key at 1:00:17, I should have waited for the Broken Golem to spew acid before jumping over it.
  • My idea at 1:02:26, obviously, was to throw the Javelin at such a height that I could just hover to it instead of going through the trouble to hover up to the high ledge. If I did this correctly, I could save about 11 seconds.
  • I could have saved about 30 seconds at 1:04:59 if I had just dropped onto the green mushroom properly instead of walking too far.
  • The proper time to use the stamina boost in Ancient’s Crater exterior is at 1:07:49, not 1:08:03, so that the Flying Bomb trap can be more easily avoided.
  • Another good place for an extra stamina boost is at 1:11:16. It’s not necessary, like I said, but it does make things a heck of a lot easier. It might even extend to 1:12:37, making navigation to the moonstone easier as well.
  • The failed Harpy Skip at 1:14:29 cost me about 11 seconds. Still faster than having to fight the thing, though.
  • The anchor box puzzle in Atai at 1:17:42 has a much simpler solution than the one that I attempted. Just break one box on the left stack and climb up using the Javelin. Plain and simple. Potential 5 seconds of time save.
  • As I said before, I am considering moving the moonstone outside of Atai palace (at 1:20:10) to after delivering the rejection letter (at 1:30:31). Might not save much (if any) time, but it does save an inventory slot.
  • To break the rubble at 1:21:44, why not just use the Ki Spear Technique? It’s roughly twice as fast (i.e., about 6 seconds of time save), after all.
  • Upon climbing the ladder at 1:27:22, throwing the Javelin (or at least firing the Slingshot) would have helped (by saving about 2 seconds).
  • A correctly executed JBJ at Misty Gorge (at 1:34:40) could save me around 43 seconds.
  • At 1:39:19, pushing the top box was not absolutely necessary and wasted about 2 seconds.
  • If I hadn’t squandered the ladder-and-arrow puzzle at 1:43:35, I could have saved myself about a full minute and a heck of a lot of trouble.
  • I didn’t really have a plan for the box puzzle afterwards. Maybe I should finally get around to formulating one and save myself some trouble and time. (I got myself in what I found to be quite a predicament at 1:45:05 and spent roughly 30 seconds trying to find my way out. Moreover, this might not even be all the potential time save.)
  • On the second visit to Great Walls, I should have talked to Gill before entering the barracks. This could save me roughly 2 seconds.
  • At 1:50:38, I could probably throw the Javelin onto the leftmost platform as a little frame saver.
  • 1:52:17. Again, Ki Spear. ~3 seconds of time save.
  • Come to realize it, speaking of the Ki Spear Technique, I recently found out that it has the special property of causing bombs to instantly explode. Maybe I could use that for the puzzle cheese at 1:52:37 and save about a second.
  • If I hadn’t beefed the collection of the final moonstone in Forgotten Forest at 2:00:17, I could have saved myself roughly 30 seconds.

Looking upon these potential time saves…sub 2 RTA might be possible, and I guess in terms of honey stops that I can only think of three places where it would be useful beyond what’s necessary (Prince Tower 1, Ancient’s Crater, Prince Tower 2), two of which are prior to the second visit to Sunflower Road. So, that would be 2 Jars of Honey on the first visit to Sunflower Road (which can easily be accounted for in terms of expenses using the second-rightmost house in Panselo) and 1 on the second visit (which might require an extra backup chest or selling a honeycomb).


While 100% is indeed safer than any%, it is significantly longer. In fact, it’s two hours long with hardly any downtime whatsoever, and some of the Loot Dupes (notably Duri Forest interior and Daea prison switch room) require difficult visual cues that strain the eyes quite a bit. Still, it is a fun run, and I would like to improve my time someday when I have 2 hours to spare.

Nowi Wins À la prochaine! (Until next time!)

Phoenotopia speedrunning (Thought Dump Thursday 1/7/16)

Screen Shot 2015-08-14 at 11.25.35 PM

A sort of precursor to this post is Spontaneous Saturday 8/15/15, wherein I talk about the game as a whole and my initial plans for speedrunning it (primarily before I discovered Loot Duping).

Now, what I wish to do is lay out the entirety of my thought process behind speedrunning this game into one blog post. It is a rather long read, which is why I included headings to aid those who wish to skip past certain parts of the reading.

How I started

The reason why I considered picking up this game as a speedgame is because I always found it to be an amazing game and there was no leaderboard for it on (not to mention there is a “Speed Runner!” achievement that is part of the game). The routing started out rather weird, involving not only breaking several chests, but also taking advantage of the NPCs that give you money for items (the guy who sells honey at Sunflower Road buys honeycombs, and the Duri Fruit merchant in Atai buys Duri Fruit, both for 15 Rai apiece). Then, I fine-tuned the route to exploit the fact that you can actually break into the mayor’s palace stash without the Javelin if you pull out a bomb and let it detonate on you at the peak of your jump, allowing you to pop up on top of the permeable part of the roof. Because of the high yield from this (which is normally 90 Rai, in addition to a potential extra 50 or 90 from selling the High Quality Silk), this was the easiest way to cover the expenses for the Iron Hammer (which is pretty much required because it makes boss fights quicker)…that is, until the moment I discovered Loot Duping.

Loot Duping

You may be wondering why I keep going on about “Loot Duping.” In short, it’s when a loot container is broken multiple times on the same frame and thus drops its spoils as many times as it is broken. I’ve mentioned it in my previous post, calling it a “glitch.” I was previously debating this nomenclature, but the SpeedRunsLive glossary defines a glitch as an “unintentional mistake in the game’s code, often exploited.” From a design perspective, I’m sure that the developer did not intend for the chest to be broken any more than once on the same frame; therefore, I now stand by calling it a glitch. I seriously don’t know whether this has been found before, because searching “phoenotopia glitch” on Google came up with nothing pertinent.

Primitive stages of the glitch

It started with the chests in Bandits’ Lair, which I had the idea of breaking by throwing them at each other. Normally, this would produce 50 Rai, because both of the chests have 25 in them, but I got curious when I realized that sometimes I was actually getting 75. Thankfully, I actually recorded a live demonstration of this happening.

Forgive the lack of professionalism; I did a simple recording in OBS and can’t be bothered to edit out the parts where I bring up OBS to start and stop the recording (because the original file is an FLV, which I do not have the software to edit). In this file, I start off with 26 Rai. At the beginning of the video, I get 50 Rai from the duo of chests (ending up with 76), as expected. However, later in the video (at around 0:50), I actually get 75 Rai from the duo of chests (ending up with 101).

This enigma led me to hypothesize that one of the chests had its spoils doubled because it collided with two objects at once. Through experimentation, I figured out how to achieve this “Loot Duping,” as I dubbed it, with the chests in Bandits’ Lair and the piggy banks in the mayor’s palace in Atai, in a consistent manner through pixel-perfect positioning.

Bandits' Lair Loot Dupe positioning

This is the primary Bandits’ Lair dupe positioning, which consists of lining up Gale so that the tip of her front foot is at the center of the second crate from the right (using the top pixels of the crate as a visual aid). If you were to throw the chest at this position, the chest would simultaneously collide with the other chest and the crate underneath it, which achieves the hypothesized effect of doubling the thrown chest.

PRE Loot Dupe positioning

This is the primary Atai palace dupe positioning, which uses the gold between Gale’s legs as a visual aid; i.e., she is positioned so that only the higher pixels of the gold are showing. This was the most important dupe of the run at first, yielding a whopping 135 Rai, more than enough to cover the expenses of the Iron Hammer.

Consistent Loot Duping

Not long after figuring out this new tech, something vital occurred to me. If this “Loot Duping” is achieved when a chest collides with two objects at once, could that also apply to a corner? Thus, I hypothesized, experimented, and figured out the most consistent form of Loot Duping. The easiest and original way to set it up is by positioning Gale so that the tip of her front foot is exactly 64 in-game pixels (192 computer pixels in the standard 900×600 window) from a corner on the same level as her. Once set up, all that’s left is to throw the chest at the corner for double the money.

Based on this new and more consistent method, the positionings of the above two Loot Dupes have been updated.

LDP - Bandits' Lair

Use the purple trinket as a visual aid. Performing this on both chests yields 100 Rai instead of 75.

LDP - Atai palace raid

Use the left vertical groove near the windows as a visual aid. Performing this on both piggy banks yields 180 Rai instead of 135.

By the way, this glitch only applies to containers that contain money or hearts. Other items—that is, Moonstones and consumable items (such as Honey)—cannot be duped. I have tried with the moonstone chest in Daea and the Honeyed Tumond jar near the save point outside of Bandits’ Lair, and neither can be duped.

As I went through the game and made note of places where the glitch can be put to use, I found that corners are not just limited to walls and floors, but also other containers and even enemies. In Sunflower Road, for instance, it is possible to dupe the chests by coaxing a Purple Spider to fall to ground level and using that spider and the ground as a makeshift corner. This is relatively difficult, however, as you have to get the exact positioning before the Purple Spider starts to return to its initial position (which doesn’t take that long) in order for the dupe to be successful.

LDP - Sunflower Road East

Shown above is how you would dupe the chest on the east side of Sunflower Road, which normally yields 20 Rai but yields 40 after being duped. While the Purple Spider to the right of the chest’s initial position is down, look at the groove in the ground positioned in the middle of the leaf to the right of the Purple Spider and make sure it lines up with the center of Gale’s front foot.

Another way to create a makeshift corner is using the Javelin. This occurred to me when I was thinking about how I could dupe the chest in the falling maze in Daea. If there is not enough horizontal space for a traditional dupe, you can throw a Javelin at ground level to reduce the required space from 64 in-game pixels to 50 (i.e., from 192 computer pixels to 150).

LDP - Daea falling maze

Shown above is how you would dupe the chest that I mentioned earlier. You throw a Javelin at ground level and use the vertical beam in the background as a visual aid.

So, to conclude, here is a full list of consistent Loot Dupe positionings that I have found in my recent experience with this game:
(If any of the ones that I have omitted are actually possible, and/or if I have missed any others, please let me know!)

4/5/16 EDIT: I just recently found out a way to make Loot Duping easier. If you are one pixel behind the required position (which happens quite frequently in my experience), you can simply place the container on the ground and roll into it. This moves Gale exactly 1 pixel forward, and that gives you the correct positioning.

Frame-perfect Loot Duping

It may not be possible to dupe all chests using the pixel-perfect methodology described above. (The exceptions as far as I know are all chests in Mul Cavern and the one chest in the second invisible maze of Ancient’s Crater.) However, regardless which chests cannot be duped using the pixel-perfect method, all chests can be duped.

When I was thinking about how Loot Duping relies on a loot container being broken twice on the same frame, I thought about other ways to break a container, and the first two things to come to my mind were: melee and bomb. If you place a bomb right near the loot container and time a melee attack so that it lands right when the bomb explodes, the loot container will produce twice the loot, just like with a pixel-perfect Loot Dupe.

Oh, but that’s not all. At the time of writing this blog post, something more appealing occurred to me. Could it be possible for a loot container to be broken three times on the same frame? Surely enough…well, just watch the video below.

The chest in the upper cave of Adar’s House, which normally contains 35 Rai, is broken in three different ways using the same approach (setting a bomb and throwing the chest at the corner as the bomb detonates):

  1. At 0:16 in the video, the bomb explodes on it before it can collide with any surface. Therefore, it produces 35 Rai, as expected.
  2. At 0:41 in the video, the chest hits the corner before the bomb explodes. Therefore, it produces 70 Rai—twice as much, as with any other pixel-perfect Loot Dupe.
  3. Now, here’s the fun part. At 0:56 in the video, as I intended with this nonsense, the chest hits the corner on the exact frame when the bomb explodes. Therefore, this one little chest containing 35 Rai has its spoils tripled and produces a whopping 105 Rai.

It is possible to do this with the Ki Spear Technique as well, but this requires a manual timing as opposed to a fixed timing (notwithstanding that the explosiveness of the Ki Spear expires after a certain time period). Whether this is less or more difficult is up for debate, but for now, I have to say the Ki Spear approach is easier.

Now, I’m not sure if it’s possible to combine hits from the bomb and the Ki Spear in conjunction with a makeshift corner to quadruple the spoils of a chest, but that would be extremely difficult to pull off if it were, because it combines the fixed timing aspect of the bomb method with the manual timing aspect of the Ki Spear method. Also, it requires a stamina boost, the Nebula Armlet, and literal first-try positioning. If anyone could pull that off and provide video proof, I would be very grateful.

Anyway, frame-perfect duping is not included in runs because it is not nearly as consistent to pull off as pixel-perfect duping. I’m not stopping anyone from trying to pull it off, though, especially because the tripling method, if done properly, can cover expenses like nothing else.

So, enough talk about Loot Duping. Let’s get down to the nitty-gritty.

Any% speedrunning

Before watching this video, please read the following Pastebin: (Also note that I will be using timestamps very frequently throughout this section, and they pertain to the video, not the splits.)

This is my first recorded any% speedrun of Phoenotopia, which ended up being 59:22 RTA (1 hour and 2 minutes IGT). I’d say it was a good run overall, outside of the blind sewer navigation (at around 28:06; wasted about 46 seconds), Prince Tower floor 3 (at around 35:10; wasted about 16 seconds), and, most of all, the end bosses (starting at 52:38; died to Big Eye once and 66 twice; lost about 1 minute and 19 seconds to Big Eye and about 1 minute to 66). My statement at 52:14 that “I probably just jinxed it” was more accurate than I expected. Also, some of the dupes could have been better, especially the Daea guard house dupe (at 24:27). On a more minor note, my menuing and mashing through text also could have been better. So, I say this time can be cut down by three minutes, maybe even more. I’m thinking sub 56 would be an ideal time for this category, and sub 1 hour in-game time is definitely possible.

Also, this wasn’t my first actual run, just my first recorded run. I believe I got 1:01:xx on my actual first run and 59:xx on some other run (which I believe was worse than the recorded one). It’s not the first time I tried to record a run either; the first run I tried to record, even though it was on such a good pace, ended up dying to Prince Tower floor 6 (which I execute first try in my completely recorded run at 36:30).

1/23 EDIT: Better (and deathless) run in 55:14 RTA [58m IGT]:

My Route

I completely routed this run from scratch, using an already completed file for movement practice through treacherous areas, and it is devoid of any Heart Rubies. My route can be found here:

The Loot Dupes used in the run are:

  1. Duri Forest interior key room [30 -> 60] (3:45)
  2. Duri Forest exterior [15 -> 30] (6:10)
  3. Panselo leftmost building [25 -> 50] (7:03)
  4. Panselo second-rightmost building [15 -> 30] (7:41)
  5. Panselo lab [30 -> 60] (7:57)
  6. Panselo dojo [20 -> 40] (9:21)
  7. Misty Gorge inn [30 -> 60] (19:52)
  8. Thomas’ Lab [40 -> 80] (23:55)
  9. Daea guard house [35 -> 70] (24:27)

The first six dupes cover the combined expenses of the Rolling Technique Scroll (required to get through Atai Bridge; costs 50 Rai), the Talkinator Juice (required to get the locked-up bandit to tell where Bandits’ Lair is; costs 10 Rai), the Iron Hammer (makes boss fights faster; costs 120 Rai), and 6 Jars of Honey (which, at the cost of 90 Rai, grant 1.2x running speed and double stamina regeneration for 2 minutes each, allowing for quicker movement and some neat tech). The Misty Gorge inn dupe, funnily enough, is just the right sum to buy the Floatation Donut (60 Rai), which is required to proceed through Misty Gorge. The final two dupes are used for extra honey, potentially leaving a bit of backup Rai for revivals if needed (which is further bolstered by the chest in Daea sewers broken normally at 29:19).

Although I feel this financial planning is pretty much optimal in terms of time saved, it’s not very safe because of the following setbacks:

  • You cannot die before Thomas’ Lab
  • You cannot fail any of the dupes listed above; otherwise, it’s a reset

The first setback can be mitigated using the chests in the Bandits’ Lair storage room, but there’s not much to do about the second one.

Techniques used

This run contains some neat little tricks that are used to save time over the intended way of doing things. In particular:

  • As with most 2D platformers that somehow involve climbing, jumping up a ladder is quicker than just climbing it normally—in this case, about 1.5 times as fast. This is first demonstrated as early as 0:28.
  • At about 1:05, when I get the Wooden Bat, I hit the first box from a distance so that I can create a little ledge with it instead of having to hit the box four times (or with two consecutive charge attacks). I gave this trick a pet name known as “Box Ledge.”
  • I occasionally abuse the mechanics of landing lag by either pulling out a bomb (demonstrated at 17:37) or using a midair attack (demonstrated at 4:48). Pulling out a bomb is the simple approach, as landing lag does not occur while Gale is holding an item over her head. Using a midair attack, on the other hand, is more complicated. Normally, when Gale falls from above a certain threshold, if her descent is uninterrupted, she will go through a unique “landing lag” animation that wastes about half a second. However, if she is in the middle of a midair attack animation before hitting the ground, the landing lag mechanics are actually reversed, meaning she will only go through landing lag if she falls from below that threshold, and the inverse is also true.
  • At 3:30, I did not know to hit the key off of its platform using the Slingshot (initially, I would use the wooden box to create an extra platform for me to climb, and that made things way more difficult) until I saw a post on Reddit from more than a year ago of someone else who was considering speedrunning the game (but, as far as I can tell, never followed through).
  • Obviously, Loot Duping, as mentioned above.
  • As first demonstrated on the Armored Toad King at around 4:52, when I try to get off multiple attacks, I start it off with a charged attack, because apparently using a charged attack can make one or two subsequent attacks charged as well. (I give full credit to the TVTropes page (specifically the “Ascended Glitch” part) for that one.)
  • At 14:59, I do a little damage boost and abuse invincibility frames and the speed boost to run past the bandits instead of having to tangle with them.
  • When I started routing any%, it took me quite a while to realize that you can actually destroy the Sand Dragon’s fire hazards with bombs (as demonstrated at 16:15). I imagine this is intended, but it never occurred to me until the primary stages of routing (which was about a year after I first played the game casually).
  • I use the fact that interiors are significantly larger than exteriors in Phoenotopia to make decisions like taking the right entrance to Gale’s house instead of the left at 19:11.
  • Starting at 20:18, I jump out of the water a lot, making it obvious that airborne movement is significantly faster than waterborne movement.
  • To solve the pressure plate puzzle in Thomas’ Lab, instead of having to free the three metal boxes that are barricaded, since I have a stamina boost, I can place the metal box outside of the barricade on one pressure plate, place bombs on the other two pressure plates, and step on the fourth pressure plate, as demonstrated at 22:19.
  • Blind navigation through the sewers, as demonstrated at 28:06, in an attempt to save having to buy the Lamp. It’s easy until you climb the ladder (after hitting the farthest point left from the second-leftmost source of light on the screen), after which it becomes a nightmare. Upon dropping down, hold right so that the very leftmost source of light is barely showing, hit the switch, blindly jump to the platform that gets activated by the switch, wait for the sixth quickened tick of the timed switch, and blindly jump to the platform that re-activates as a result of the expiration of the timed switch. You should see a dust cloud if you did it successfully, meaning that you can jump again and head straight left.
  • At 29:43, I demonstrate the safest of skirmishes, picking at the guard dog using the bombs and taking out the Spear Guard using a combination of the Slingshot and the Iron Hammer.
  • At 30:15, I skillfully evade a guard dog and two guards by jumping over the guard dog and onto the bookcase. This saves the trouble of having to fight them.
  • At 30:30, I demonstrate an obvious flaw in the fish puzzle of Daea prison by sniping a switch off-screen with a well-aimed Slingshot round.
  • At 33:48, I exploit two key factors to take care of the Flying Bomb: (1) Gale is immune to damage when entering a door, and (2) once an enemy has been defeated, it will not come back unless you exit to the world map and re-enter. In essence, I lure the Flying Bomb over, time my exit just right, and voilà—I’ve defeated the Flying Bomb without having to play tennis with it (for lack of better terminology). I also abuse this at 35:46.
  • If you’re fast enough at 33:56, you can get the Flying Bomb above to activate the barricade-lowering switch for you, which is the best way to handle it.
  • At 34:03, it is fastest to alert both Flying Bombs so that the farther one explodes while you flee from the closer one. You could probably take the damage like a man and heal it off later, but I personally prefer to just flee from them.
  • The quickest way to hit the timed switch at 34:45 is by quickly firing the Slingshot. Additionally, at 34:48, pivoting and throwing the Javelin is quicker than climbing up the provided platforms.
  • On floor 5 of Prince Tower (at 36:16), instead of taking the route starting from the far right, I skip about half of the platforms by throwing the Javelin at the peak of a small jump (not necessarily a short hop) and using it as an easier way up.
  • If you manage to two-cycle the Kobold Mercenary General, like I do starting at 37:05, you should not have to worry about its annoying Flying Bomb phase (which, if it occurs, results in some sort of damage pretty much every time a Flying Bomb spawns, at least in my experience).
  • In response to the bomb trap at 41:53, I place a bomb right in front of the first Rolling Bomb to make the duo more avoidable.
  • Similarly, in response to the bomb trap at 44:09, I place a bomb right in front of the first Rolling Bomb on the right side, and the other two blow up by running into each other, making a clear path to the exit. The slightly faster (but riskier) alternative is throwing a relatively precise Javelin.
  • Much like the sewers, I navigate blindly through Mul Cavern to completely eliminate the need for the Lamp. This time, however, it is more semi-blind because you can barely see what is not illuminated (whereas in the sewers it is pitch black). Navigation through this area abuses two key factors: (1) The hives usually take 6 hits of the Iron Hammer to destroy, but explosives (i.e., bombs and the Ki Spear Technique) destroy them instantly. I believe this is another fact that I did not know before but figured out through intensive analysis of the game. (2) If a hive is destroyed and you exit the screen, the beetles that come out of the hive will despawn and never come back. Abusing these two factors makes blind navigation easier than it could be, but it is still relatively difficult. Thankfully, there are two Spotted Lepistas in your path, which each heal 15 HP. (Consult the route for further details on blindly navigating Mul Cavern.)
  • 47:49 shows that there is a part of the third screen of Ancient’s Crater that you can just completely bypass by performing a stamina-boosted hover from the permeable platform on top of the left side.
  • On the fifth screen of Ancient’s Crater, I actually use the third and fifth laser towers to my advantage by throwing bombs at the turrets, letting the laser towers detonate them on sight. This, in tandem with the stamina boost, helps to wither the turrets down with a relatively quick flurry of bombs (3 should be enough to kill), as demonstrated at 49:00.
  • At 49:51 and 51:17, I show off how to breeze through the invisible mazes without using the Artifact.
    • For the first invisible maze, I jump just past the initial platform, throw the Javelin after a short hop, climb up via that Javelin, find the highest possible ground to the left, jump from there, navigate to the top-right, hover over to the second vertical background element to the left from where you are, fall down, and head left.
    • For the second invisible maze, I head left until hitting a wall, head right until hitting the visible platform (literally or otherwise), walk left until I’m just behind the the second gray background element, jump up-left, throw the Javelin, climb up, and navigate easily through the rest.
  • For the upside-down turret that appears in 50:19, you can actually throw the metal boxes below at it, all of which do as much damage as the Iron Hammer, and even hold one as a sort of umbrella to protect yourself from the turret’s projectiles.
  • Ancient’s Crater is more broken than you may think. I literally skip every lockout fight before the boss: the Arc fight at 50:56, the flying turret fight at 51:02, the Golem fight at 51:40, and the Harpy fight at 51:49.
    • Arc Skip and Turret Skip are simple. Just position yourself adequately close to the laser trap, throw a bomb at maximum height, and the bomb will stop the barricade from moving when you activate the laser trap, so get the heck out of dodge before it explodes. In the case of Arc Skip, make sure to hover over the Arc that will appear on the right.
    • Golem Skip is the most difficult trick of the run, and if you fail it, you might as well just defeat the golems normally (although it’s slightly slower that way). You have to position yourself as far into the laser trap as you can without touching the danger zone and throw a bomb while hovering at maximum height. Ideally, the bomb will be in between both of the currently inactive barricades. Then, activate the trap, wait a brief moment, hover left, and get the heck outta dodge. Note: if you hover left too quickly, you will hit the Red Golem corpse and fail to make it in time.
    • Harpy Skip, I have to say, is a middle ground between Arc/Turret Skip and Golem Skip. However, unlike the previous skips, it absolutely requires a stamina boost. Run as fast as you can through the laser trap, jump so that you arc over the elevated ground ahead without bumping into anything, and fall on the right side of the drop (at which point you should ideally make it past the barricade before it’s too late), making sure to slow your descent by hovering, lest you fall into the pit and have to attempt the skip all over again.
  • At 51:10, I perform a neat little exploit in the backwards C area where you can drop a bomb down onto the vertically based laser trap and activate the lower laser detector, which will ideally result in the bomb being hit by the far left laser instead of you.
  • Big Eye seems really difficult with 20 HP, seeing how it takes exactly two blows coming directly from it to kill you. However, at 54:02, I demonstrate a perfect fight against the boss. Since the best number of cycles achievable with the Iron Hammer is 3, the cycles are best initiated at the following points:
    1. After destroying the flying turret. By then, you should have only one mini eye left to blind with your Artifact (on the elevated platform, of course). The sequence of Javelin throws on the roped spheres makes this a whole lot easier to achieve, and you can blind at least one mini eye before Big Eye starts sending the spheres out. Any other mini eyes can be blinded after the wind and lasers (all but the last, obviously).
    2. After the wind and Rolling Bombs. After the first cycle, hold right to stay damage-free during this phase and, when Gale stops jittering, blind the bottom-right, top-left, and top-right mini eyes (and the bottom-left if necessary; most of the time, the Rolling Bombs take care of that one).
    3. Right after the second cycle. Gotta be quick about this one, otherwise the next set of roped spheres will kill you. (Such was the cruel fate I suffered on my first attempt of Big Eye in the recorded run.)
  • You can actually escape from 66 without activating any unnecessary barricades if you’re fast enough. I tried this twice in my recorded run, but I failed both times, so I just decided to take a safer approach and only activate the first two.

Now that I think about it, it is actually really easy to skip the first Gray Golem on the second screen of Dread Lands (the one that I destroy with the stationary Rolling Bombs at 42:02) by jumping over it quickly, but that never occurred to me until now. (The any% route document has been updated accordingly.)


This is something I’ve been thinking about along with any%, almost hand-in-hand—routing 100% has occasionally helped me find new techniques for any% and vice versa. The way 100% works is that the “collection rate” at the credits should say 100%. Even though the percentage counter was confusing to me at one point for a reason that I mention in Thought Dump 7 (long story short: 99% at first, unlock Mystery Bento and Power Potion, now at 100%), I managed to figure out what exactly causes the collection rate to rise and by how much. It works as follows:

  • Heart Rubies count for 2% each, making for a total of 46% if all are collected.
  • Moonstones count for 1% each, making for a total of 42% if all are collected.
  • The Blood Ring, Nebula Armlet, Lucky Belt, Ancient Armor, and Morning Star count for 2% each.
  • Obtaining the Ki Spear Technique Scroll counts for 1%.
  • Now, get this: Giving Translucent Meat to the chef in Daea counts for 1%. If you ask me, this is a very arbitrary requirement for 100%, because it is precisely what confused me in Thought Dump 7.

So, it makes sense that the resulting collection rate of an ideal any% run, which requires literally none of those, is 0%.

In terms of routing 100%, I’ve routed out what I find to be the fastest order of visiting the required locations and obtaining the required items, as well as covering the required finances through Loot Duping and various quest rewards throughout the game. A condensed version of the route in question can be found at:
whereas a more specific version of the route can be found at:
and my planning of the route can be found at:

I might plan to run 100% eventually, but for now, I’m kinda reluctant. Also, I feel that the route is not quite complete, because I’m wondering where stops for honey could be incorporated, if at all.

By the way, I just figured out today that the Ki Spear Technique is baneful to Flying Bombs (i.e., it causes them to detonate early). I had no idea how to deal with these things at first, but apparently that’s how. This might actually make parts of Daea prison and Prince Tower slightly easier. It’s not worth incorporating into the any% route because you need extra Rai for a Chocolate Protein Shake and some inventory space for both that and a Duri Fruit (the latter of which wastes time to find), not to mention the few Flying Bombs you encounter in the route are not that hard to deal with.

On a final note, that extra in Prince Tower where you have to climb up the ladders and avoid the arrows…is in my opinion the most frustrating thing in the game. (No amount of routing the late-game treacherous areas in any%, nor any part of Forgotten Forest, can compare.)


I submitted this game to a week ago, making sure to include the video above in my submission, but I still haven’t received any approval or rejection. Perhaps just before AGDQ 2016 wasn’t exactly the best time to submit… 1/9 EDIT: Looks like it’s approved now.

At any rate, I definitely enjoy Phoenotopia (in case that wasn’t obvious), and I find it to be my kind of speedgame, no matter how frustrating it may be at times (especially when I always end up one in-game pixel off when trying to perform a Loot Dupe due to Gale’s natural slipperiness). I also had fun routing it, as annoying as it was to figure out the treacherous areas, especially for any%.

As of now, there still remain some mysteries to me.

  • Once again, is it actually possible to break a loot container four times on the same frame? Like, if you were to have a bomb and a Ki Spear explode on the container on the exact frame when the container hits the corner?
  • At 42:22 in the recorded any% run, why was the Broken Golem not there?
  • After my death on Big Eye (at 53:50 in the video), why did the HP counter display 0 upon revival even though Gale normally revives with 10 HP?
  • What’s the deal with the Gold Bar merchant? I can never seem to figure out how the pricing works—how it’s set or how it changes.

Perhaps some of these have already been figured out (maybe even all of them), but the few times that I have searched the web for anything glitch-related, I’ve come up empty. Some sources, if there even are any, would be helpful.

Anyway, I’ve rambled on quite enough. The point is: I have a route for single-segment any% that is crazy enough to work (and something of a route for 100%), and I have shared the full extent of my thoughts on the matter, hoping that others who are interested in the game and/or speedrunning as a whole will see potential in it as a speedgame, just as I have.

Nowi Wins À la prochaine! (Until next time!)

Top Three Thursday 5/14/15: Favorite speedruns to watch

Since I spend a good portion of my time watching speedruns, I figured for today’s Top Three Thursday that I would make a list of a few in particular that I like.


3. Kirby 64: The Crystal Shards

Of all of the games in the Kirby franchise, Kirby 64 has always interested me the most due to its mechanic of combining types, not to mention it has some awesome music, particularly in Ripple Star level 3. Although I don’t remember the gameplay too well, I have a good understanding of it by watching speedruns of it. I find 100% to be the most interesting category; while it takes slightly longer than any%, at least it has more in store than just fire spam. (Fire and double fire are the fastest types in the game.) There are these collectibles called Shards, all of which are required for 100%, some of which require particular type combinations to get (which are generally indicated by color), which is why the 100% run is more diverse in what types are used. In closing, the game and the routing (of 100%) are what make speedruns of it fun to watch.

World record (WR) run at the time of publishing – 1:10:32 by Swordsmankirby (


2. Bravely Default: Flying Fairy

Even though I’ve only ever seen one run of the game (; WR at the time of publishing) via the SGDQ 2015 submissions page (shame that the game got rejected), it’s still an interesting speed game to me. First, I never figured that you could destroy the crystal to discover an alternate ending (the full extent of which can be viewed on Second, the strategy of using Reflect and casting Black Magic on teammates to increase damage output is really creative. Third, to be honest, in all of my hours of playing the game, I never considered manipulating the encounter rate like Romi did. (Rather than random encounters, he used planned encounters (e.g., in Florem Gardens) and Barbarossa to gain experience.) I also like how the route takes into consideration a number of jobs and abilities that increase damage output and/or make things faster/safer (e.g., defeating Mephilia to obtain Girtablulu, which solves the problem that the Land Turtle (one of the mandatory bosses) can cast Reflect on itself; defeating Qada for Compounding, allowing the ability to use Font of Life (Hi-Potion+X-Potion) to grant Reraise to allies; and leveling up as an Arcanist for Black Magic Amp).

Even though Bravely Default is a turn-based RPG, speedruns of it are definitely not boring (especially due to how unforgiving RNG can be, e.g., the Behemoth’s crits in Romi’s run), and my first time watching the run has taught me things that I had either forgotten or overlooked in my 100 hours of playing the game. (Then again, the alternate ending video did as well, mainly how useful Mimic can be.)


1. Donkey Kong 64

Like Kirby 64, I don’t remember Donkey Kong 64 too well (in fact, even worse). Still, it seems to be the most enjoyable speed game I can think of. It’s one of several games that is so broken that nobody runs it “glitchless.” I mean, the closest to “glitchless” you can get in DK64 runs is No Levels Early, which merely restricts accessing level lobbies before intended and skipping B. Lockers without the required golden banana count. (I suppose you could consider it like 70 star in Super Mario 64.) Really, the game is so broken that you can swim through vertical walls (sometimes abbreviated STVW), clip through ledges at a certain pixel (buffered through first-person mode, which allows you to skip Troff and Scoff doors), perform a special kick as Donkey Kong with drastically increased height (called a moon kick), and skew your character model in order to clip through walls (demonstrated on at around 14:10). On the original Nintendo 64 version, you can use oranges to lag the game and make your character fast enough to clip through certain walls. (N.B.: This does not work in the Wii U Virtual Console version because lag was removed.) It’s really fun to watch, even if it drives runners mad sometimes, especially because of Lanky Kong. Really, I’m glad that No Levels Early got accepted into SGDQ 2015, and I’m definitely looking forward to it.

WR runs at the time of publishing:


Nowi Wins À la prochaine! (Until next time!)