Red Ball 4 Vol.1 Sub 2

Well, here we are.
After over three months, 550+ sub 2 pace runs out of 6, 20 sub 2 pace runs out of 11, 200+ hours of attempts, and well over 100 hours of practice, I have finally achieved the holy grail of Red Ball 4 speedrunning: Red Ball 4 Vol.1 Sub 2. This grind has been the most grueling single task I have ever undertaken, in gaming and otherwise (as of writing).

This run isn’t much different than my last personal best, with the only sizable timesave being hitting the final DJ in 13 which allows for a far faster 14. If you are new to Red Ball speedrunning (which I suspect many more of you reading this will be than usual), this run features one massive frame-based strategy found by Starline that retains very large amounts of speed through all of the platforming levels in the game, from 1-14. Pulling it off, however, is far easier said than done, requiring 40 consecutive frame perfect inputs (which each have a window of 33 milliseconds to be performed) at minimum; I was additionally using my own setup to retain speed into 14, which makes for a total of 43 frame perfect inputs in this run.

While I don’t normally do this, I feel it is necessary to share more of the finer details of this grind in the hopes of answering some questions that spawned due to me doing much of it in private.

When I announced my plans to go for sub 2 in late May, the deadline I set was a source of confusion for some. Forty days to get sub 2 was, while certainly possible, a very ambitious assumption. However, the real reason for this deadline was not one of arrogance. My real goal was to break the barrier before mid-June because I had a summer university course scheduled to start then; I said the deadline was at the end of June because a specific date would have drawn attention to the class which I didn’t want (the class was later cancelled so it wasn’t a dealbreaker like I originally thought).

Of course, 21 days to get such a monumental barrier would have been an immeasurably large task if I had to start from scratch. Herein lies the reason for my confidence in beating my deadline: before summer break began, I practiced the sub 2 strategy in private for multiple weeks so I could learn it without potentially starting a sub 2 race while I was unprepared and/or burdened with studying. While it wasn’t particularly dense as I was balancing practice with finals during that time, it was enough for me to be decently experienced with it by the time summer rolled around, having ironed out some physical issues such as my grip for the left-tap + DJ combo in 7, which I fixed by rebinding the C key to left(this made typing very awkward for the entire grind), and having hit 6-15 twice and 1-7 over ten times. Even though it wasn’t a crazy level of consistency, I was still extremely confident that I would have sub 2 in a couple weeks tops, or maybe even fluke it on day 1.

I couldn’t have been more wrong.

My hubris was met with the most potent reality check I could have received. When I started doing real attempts, I let the mental barrier and the pressure of expectation get into my head and my DJ consistency and mentality fell off a sharp cliff. Playing the game with a bad mentality is very difficult as you need to time 1-frame jump taps for the first jump in every DJ, and often leads to a feedback loop that ends with a ragequit. This cycle persisted for a week until I encountered an entirely new problem: my muscle memory for the screen triple jump at the end of level 4 had evaporated. Whenever I went to hit my spacebar to do the inputs, my arm would spasm and often input nothing. This dropped my consistency with the triple jump from over 70% to around 5%, which was the source of an entirely new mental block. Another week was spent trying various key rebinds and postures to salvage my consistency until I figured out an approach that worked: leaning over and using the spacebar on my laptop keyboard. With this, my consistency rebounded, and I began making progress in my personal bests. This all culminated in a 2:00.967 time I got on June 28, which died to the final difficult DJ in the run (you can die to the DJ in 14 but you suck if you do that). I tried not to let this deter me and continued pushing hard to get sub 2 before the end of June, an endeavor in which I was unfortunately not successful. Still, I figured that with a run right at history’s doorstep, it was surely only a matter of time at that point.

*EXTREMELY LOUD INCORRECT BUZZER*

For the entirety of July, I bashed my head against the wall of inconsistency, desperately trying to get another run to the end of 13. Doing attempts had crossed from dull to mentally exhausting, with each day bringing a new plague of inconsistency to some part of the strategy. Additional, real-world responsibilities were also rising up during this time, which split my attention and made doing attempts feel even more like a waste of time. After three weeks of being stuck in the first ten levels, I finally broke through and got another run to the end of 13… that died to it again. Thankfully, on the same day as this choke, Maximum discovered a QMK-compatible keyboard with a customizable debounce time. By setting the debounce time on the keyboard to 33 milliseconds, all key presses would essentially be forced to take at least that long, which would virtually eliminate all 0-frame presses; additionally, with the window for a 1-frame press now being doubled, almost all 2-frame presses could be eliminated as well! Maximum generously offered to pay for one of these keyboards for me to use in my attempts. Though I was already pretty consistent at 1-frame taps due to honing my DJ technique down to a science, I still expected the keyboard to be a nice quality-of-life improvement. After about a week of waiting (we used free shipping), I finally got to test it out.

I was wrong, once again. The keyboard was a game-changer. With it, my consistency skyrocketed across the board, no longer muddled by frame inconsistencies. I almost immediately broke my old record of 23 runs out of 6 in a day with my previous keyboard, getting 32 runs in about half the time. Late-game runs, which were once very scarce, were becoming more and more common, with me easily ironing out the sections that gave me trouble. Paradoxically, however, despite my massive jump in consistency and the perfect keyboard at my disposal, my motivation to grind was at an all-time low. Since the beginning of July, I had been getting progressively more burned out as the days went on. Despite what some may have expected, I rarely played for more than eight hours a day beyond early June; I mainly averaged only a couple hours each day, if I played at all. I just could not see a future where I had sub 2, even though I was right at its doorstep. Still, I knew that calling it quits after everything I had done and getting so close would be unimaginably foolish, so I pushed forward, albeit at a slower pace. After another couple weeks of on-and-off grinding with my consistency increasing to almost comical levels for not having sub 2, the day of reckoning finally came.

On that day, I opened up the practice hack and went through the motions with setting up savestates and drilling the various sections. Something I noticed is that I was getting through lategame consistently, even from the earlier levels, without sacrificing earlygame consistency. I could sense that something was different. When I threw myself into the vortex of attempts a few hours later, I would find out why. When I got into 11 on my final run, I realized that I wasn’t sure if I was recording or not. I knew I had stopped the recording for some reason about twenty minutes prior, but I couldn’t remember if I turned it back on. This basically annihilated all my nerves for the rest of lategame, as I just decided to play it out and see what happened. When I got into boss, I waited for a moment of downtime and checked OBS to see that I was, in fact, recording. This wasn’t enough to snap me out of my nerveless trance, though, as I clutched out an odd start to the boss roll and finished the run.

Massive thanks to Starline for creating the 1-14 strategy that made this run possible. I cannot explain how blessed I and the community as a whole are to have you around. Your strategies and discoveries have singlehandedly redefined speedruns of the Red Ball 4 volumes, and I’m excited to see where things go in the future.
Big thanks to Maximum for paying for the Keychron keyboard which made the tail end of this grind so much more bearable, and for getting me into speedrunning as a whole. Your work across Flash game speedrunning as a whole is nothing short of incredible and has been a very large inspiration to me and likely many, many others.
Shoutouts to Lombre for being the first person to look into sub 2 for RTA speedruns. I want you to know that, despite all the rocky relations between you and the community that I won’t bother unraveling here, I will always respect you as a player and a speedrunner.
Of course, thanks to everyone who has supported me over the course of this grind by being in VC with me or wishing me luck along the way, including but definitely not limited to NorXor, Starline, Motorjam, Queenbea, Mukmuk, Gradient, Xdxboxjaja, BirdBanana, Poguin, Apple, and Eugene Fedoseev. You all are what kept me somewhat sane over the course of this grind. As always, naming all of you would make this list way too long but know that if you ever supported me, you have my gratitude.

The run, as one may expect, was amazing. 1-13 was flawless for the strategies I was going for. There is about 0.3s of timeloss in 14 from doing the 13 transition DJ slightly early, which prevents getting a boost off the boulder in 14. There is also 1.5s of timeloss in boss over a perfect boss roll. The limit with the strategies I used is a very high 1:54/very low 1:55. Though the future of the category may look grim, there is actually still room to go lower. Starline has developed another sub 2 strategy that is slightly faster in 7-14, yielding a time of 1:53 with a good boss. Additionally, for the truly deranged, there is a 13-frame perfect boss carry setup that saves ~6 seconds over a great normal boss. With these timesaves, a 1:4x is technically possible RTA, though it would require a far faster (and likely far harder) 1-14 strategy or boss carry.

As for the future of this category, it’s hard to definitively say. During my 30+ months as a speedrunner, I have seen so many categories that I thought were solved completely revolutionized and turned on their heads, often bringing a completely new challenge to the table. Yet, after so many cycles of this happening, it appears that I have finally found a category where the end is truly within reach. Sure, the strategy for 1-14 may become a few seconds faster, but at the end of the day, there will never be another Vol.1 any% record without someone flying through every level except the boss. Boss carry may be implemented someday, but there will never be another boss breakthrough quite like it. Red Ball 4 Vol.3 Any% Glitchless has notably experienced another strat revolution in the months I was going for sub 2, making 3:1x RTA-viable. However, the TAS for it is currently a 3:13, with just slight optimizations over the real-time strategies. It, just like Vol.1 any%, will soon have almost nothing left to give.

If I ever return to this category, it will be for a 1:4x using boss carry and the fastest 1-14 available, and it will be in tandem with an MRBG any% grind after I have fully mastered the major strategies for Vol.2 and Vol.3 in an effort to push both categories to their limit. When or even if I’ll reach this point is something we’ll have to wait and see. As for others in the community, while no one else is currently going for sub 2 for a variety of reasons, I have no doubt that we will eventually see players with more time, talent, and dedication than me give this category a shot.

In speedrunning, most of our limits are fickle; there are often many points where a community thinks their human hands can go no further. Daring to defy these limits is not about breaking the barrier of impossibility, but rather about pushing through our own illusions of where that barrier lies. Theoretically, there is no limit to what we could achieve with enough time and dedication; however, there will most certainly be a point where no one finds any benefit in pushing a category any further. I firmly believe that, for most games, humans will never achieve perfection, even in the far future, for this reason. I am in no way saying that we shouldn’t attempt to reach for the stars, though. What I am telling you is to be optimistic, but also realistic. Do not be afraid to stare into the chasm of impossibility; while you may very well see the endless abyss, you may also see the way across staring back at you. With our dedication, effort, and combined force of will, we define the human limit.