SMS Strat Creations
So I had the idea of doing a countdown of the most useful contributions I made to SMS strat development. There’s a disappointing amount of inertia in spreading strat ideas among players, especially on influencing beginners who aren’t familiar with your work, so the perspective I write this from isn’t the impact that these ideas have had, rather my take on the impact they might have with a fair shot vs established strats.
But it’s really just a little soapbox for me to shine a light on my most useful ideas, imo. There’ll be a little description for each of what the strat brings to the table, who may benefit, and some history of how it was made.
#8 | Noki 100: IL Shine Grab
Out on the IL tour, I can spam segments out of save-states and rapidly learn what’s fast, especially if the timer stops or freezes at the end of the segment, like for a shine grab (if it doesn’t, sometimes I can displace the shine to make it so…). I also have a tendency to try to simplify strats for endurance ILs, which led me to simplify a horrific instant ledgegrab in this level, with this easy-to-hit, hard-to-perfect momentum-spin rope-bounce. A similar thing happened where I noticed a jump-dive into the shine was faster in Gelato Hidden than the tryhard hover-dive-rollout in the WR, but many people noticed that one tbf :).
My best segment saved 0.73s over the WR at the time, but I think a realistic estimate factoring in the time-consistency of both strats is 0.5s. Not bad! It’s niche as strats go, being not applicable to 120 shines since that uses the turbo nozzle. But a very original strat, so in at 8th place.
#7 | Bianco 6s: Lumacor Movement
2021 had just rolled around and it was high time I learnt the Bianco 6s buttslide for my ILs. The strat had evolved in 2020 from being spoken about in hushed tones as the voodoo of famed IL occultist Snide, into a strat that was being pulled off by ILers who averaged bottom 1/4 of the leaderboards and had (semi-successfully) been incorporated by Nindiddeh in his any% WR grind. I spoke to lumardy as he finished a bad run of it that he was proud of, telling him it was 2021 and this shit easy now. He promptly got it together and got WR. Along the way, he invented a dive-rollout-only-based approach onto the first set of moving platforms, and thought it saved a frame or 2.
I had been using Metacor’s jump-dive-based moving-platforms-to-first-cube movement, largely because I literally couldn’t do the standard sjd movement, and so put the two together to IL the level. My eyes opened 👁️ I started getting consistently good first cube approaches, cut down the WR cube grab, and felt in fine control with the freedom to space Metacor gave me. Dogecyanide’s Sandbox Any% relay tourney was in full swing, so I started hyping up the mic with Metacor strat while commentating. “New strat?? Timesave??? Will SB go for it?????”. (IIRC he did!). Tbh I was just excited cos lumacor was the first novel idea I’d had that really saved time :).
A few days later, lumardy tied my best approaches using his own strat. A few people had grown to like lumacor in that time, and the Metacor part became a cute preference that alas didn’t save time but, at least for me, made it easier to be consistently fast. But over the years, more and more top ILers started to agree with me. Implementing the lumardy part from IL into RTA (just with some shortened rollouts for safety, exactly as I had done in my ILs), lumacor became the top choice for top-level RTA. lumardy saving 0.1 and Metacor making it easier to go consistently fast.
Doing nowt but introducing one person’s breakthru to another’s and taking the credit here (Keepo), I’m very much the Maurice Wilkins of this discovery. But I’m happy to have played a part in what’s now the top-level strat of choice. All this over a bit of movement to the first cube… but a fun story :p.
#6 | Beginner Bianco 6s and Pinna 6s
Beginner secret strats is a research area that’s always moved me, since the culture around it of “just do casual strats” seemed to really be taking a toll on beginners who couldn’t expect any kind of consistency from these levels and just kept dying. For after all, aren’t those casual strats the reason so many casual players tremble in fear of SMS secrets?
I worked on all the secrets but put together complete strats for these 2 in particular, strats I felt were completely suitable and would give total beginners a nice experience with these levels. And both save a good few seconds compared to more usual beginner strats, with just a little kick of execution challenge to make a couple of cycles. Also, I managed to avoid deadly spinputs entirely in both. I hope these age well and become a foundational block for beginner SMS.
I made a lot of progress with fast-yet-safe beginner strats in Ricco 4s, Sirena 2s and Noki 6s, but was always left with a few niggles. My best drafts are linked in my strat demo playlist. The Ricco 4s one I’m almost confident enuff to promote to join Bianco 6s and Pinna 6s here.
#5 | Stickless GBS
Stickless… the #help channel’s favourite topic and initiator of drunken brawls. After some breh made a video that blew up that disingenuously claimed it made GBS free – “just hold neutral and you’ll get the clip!” – many beginners were due big comedowns when they lost the ability to do it for reasons they couldn’t understand, including the fact that there are positions on the roof where setting down the coconut results in no frames working for timing a stickless hover. The Discord got Angy and told absolute beginners that they had to react to do GBS, despite GBS being a largely fast and unreactable trick that only allows a bit of flexibility to slow down if your reaction time is good enuff. Many beginners couldn’t get it at all in hours of trying.
I always had faith in stickless, but for years, that faith was tested. We introduced a downward tap during the hover, which had the advantage of working in many more situations than stickless, and so was deemed more consistent. I advocated it for beginners since it avoided reactions, but in my own runs found I was getting janky animations (“why did the roof pull away from Pinna Park???”), no-clips and the like, and it was all less consistent than it ought to have been. Was GBS just a “feel-based” trick after all? Was my life of anti-feel propaganda a lie?
I cracked it eventually! After trying to diagnose why I often wasn’t clipping at all, I realised there was a 0.5-unit window in the forwards/backwards axis that always gave no-clips and was dangerously close to my position cue. But it stood to reason that if this 0.5-unit window consistently gave no clip, then other windows would consistently give non-janky animations – maybe even stickless?
And sure enuff, I found a 3-unit window where stickless worked on not one but two frames, and downward tap covered the 2nd frame and a further 3rd frame. The regions either side were still nice, and the no-clip zone could be easily spotted and avoided.
Horizons broadened with the strategy this enabled. Beginners often are happy spending 3 mins in GBS to avoid doing Gelato 1–6, so would be totally fine learning some visuals for a 3-unit-precise position, and giving stickless a few attempts before the coconut dies. It’s hard to quickly position so precisely as one gets faster, so maybe the 2f downward-tap window is alluring, and works nicely whether the position is good or slightly off (adding up to a 9-unit window). And an experienced player can accurately size up Mario’s position and decide on the fly whether to go for stickless or downward tap, and which visual cue to use. And not a reaction in sight! For those of us who are almost 30 🧓.
Ι’ve not invented a new strat here, rather taken a traditionally simple-but-inconsistent strat and made it consistent with more knowledge. Big implications for beginners, plus the new alternative of an exact flowchart to a traditionally-feel-based trick, take this to #5.
#4 | Noki 1: Better Bombs
This level was one of the hardest for me to learn, 2 years spent thinking “will I ever get good?”. But when I started analysing my PBs, I noticed I was consistently beating most top ILs in the bombs segment, and found 1f of timesave over the old EquivocalGenius 53.41 WR.
I don’t know why I started throwing the first bomb from the shine-spawn position as a beginner. I imagine I just found it hard to get into the position during the throws themselves. But it seemed to be letting me throw the bomb much earlier, and I was often early to the 3rd bomb, waiting for it to fall. I told EG about this, but he disapproved, saying he preferred fluid throws and not waiting.
I only realised 2 years later (in 2023!) that my throwing position meant the bomb’s arc missed the arc of the mole’s own bomb throw, which vastly expanded how early I could throw. This gives more time to handle 2nd bomb and prepare for 3rd. It also means the 2nd bomb walks nearer to you, so you can turn faster to grab it via tilting, rather than a turnaround. These together explain why I’m always so early to the 3rd, literally timing a spray as it falls out of the air.
Revisiting in 2022, I saw that, not only was I usually beating the 53.41’s bombs (meaning I would’ve likely had first-try WR if I could catch the fastest cycle), but I was at best 7f ahead. Later, I recorded two off-the-cuff strat demos for Noki 1, and both of their bombs segments beat that old WR. The penny dropped: even tho most of the timesave has now been replicated with the classic bombs setup in the new IL WR 53.27, I am convinced that my setup makes it easier to be consistently fast – just like Metacor’s strat in Bianco 6. And such a universal buff at all skill levels is worthy of 4th place.
Special mention to the fast goop clean I invented here as well: slide in goop with hover nozzle, then hold digital R and quickly slide thumb A → X twice. It works!
#3 | Honey Skip: Exact Setup
Honey Skip is to me a big symbol of the SMS community’s historic reliance on “feel”… which left it with the lack of a single great player pre-Nindiddeh and a dearth of top-level competition that continued for years until EquivocalGenius, Despin and both of their collaborators started to get good with a more serious approach. Many Honey Skip cues had been invented but they largely left a lot to be desired, and there was a great lethargy towards improving them despite many players reporting inconsistency with the trick. But given the rich visuals it provides, I went into labbing it confident I could do better.
The science of good cues is surprisingly deep. For z-position, the traditional cue is to creep until Mario becomes visible. But that’s a one-way cue – once he is visible, you have no idea how far forward you are and can’t adjust. I see runners standing at all sorts of z-positions as a result. I swapped it with a ground-texture cue.
For timing, reaction cues are imprecise because of the inconsistency of reaction times, so anticipation cues work better. The traditional cue for dropping hover is audio, which is usually better than video – humans are good at detecting and anticipating rhythms. However, that requires sharp attack on the pulse – the whooshing sound has a very long attack so this cue is flimsy. I saw that the FLUDD HUD moved in sync with the audio, and made a timing cue based on anticipating its rhythm and splitting one beat into 3 (for the 3 sequential inputs), which humans can instinctively count out very accurately (if they know what it sounds like haha).
What this brings to the table is a calibrated exact setup – forwards and sideways positions can both be read to 1-frame-equivalent accuracy, and the timing cue is reliably hit within 1 frame as well. The cues target the exact middle of the hover-length window, and their summed error bars still fit in the window. And there’s a visual to guarantee no death without hitting the cutscene.
Akane published a setup 2 weeks after mine that included a dynamic timing cue that allows a much greater range of positions in theory. That could be a good bit faster, but its consistency remains to be seen. My setup is universally applicable, very difficult to fail, and easy to see what you did wrong: a great new foundation.
#2 | Noki 6s: Single-Jump Shine Grab
A bit of a silly one to include, let alone at #2, since jumping is how that gap was designed to be crossed by casuals… but over the years, SMS players internalised a need to perform ever-more-elaborate spins, sideflips, triple-jump-dives, to where they see the gap as A Thing and think they’re playing safely doing unnecessary, risky shit.
Introducing formal strats that include a single-jump will hopefully normalise doing genuinely safe movement, but moreover, this safe movement turns out to also be fast! The sprint-into-single-jump ending setup is beginner-level difficulty, cycle-independent, and loses at most 1s to jump-based endings. The example I show below also cuts out another sjd from green cycle, and so loses another ~1s, but the variant that only cuts the final sjd seems like a big idea for most 1:20–1:17 players, and rebukes another dogma – that Noki 6 is Just Hard so do all the proper strats and Get Good. Meaning in this case, time a risky spinput and jump on a surface whose hight is rapidly changing.
#1 | EYG: Sidestep 5YG
The #1 spot is a no-brainer. EYG has always been considered among the hardest tricks in the game to learn, since it is off-the-rails and (relatively) inconsistent when done at competent level, and loses 18+s for any beginner variant, all of which still have a precise component. And every variant costs a lot to fail. It has always been advised to not do EYG when starting out.
Noki Doki invented a sidestep for clipping into the inner wall when doing 1-Stu, and I later realised that extending this sidestep lets Mario get thru the entire wall. Albeit slowly – this requires the slow stus of 5YG. Pretty much every innovation in this countdown is a really small idea like this one, but the devil’s in the impact…
This removes every risky precise input and fast movement from EYG, leaving only an easy precise input to get the clip itself, which can even then be attempted repeatedly by baiting the stu to pound. EYG thus becomes one of the easiest segments in an entire beginner Any% run in terms of input difficulty, mostly being doable at one’s own pace and not requiring any movement skill whatsoever to perform.
It still saves 40+s vs not doing EYG at all, so the set of beginners who are recommended to first learn non-EYG (short of 120-Shines runners) is cut down to only those who don’t want to learn anything at all, rather just improvise. And I think that’s revolutionary…
Special Mentions
For more minor but still useful stuff.
Bianco 4: Coin 8 Spin-Jump
An alternative to the GWKH that’s mechanically much easier and optimally loses only ~0.1, tho that’s with very precise timing.
Pinna 3: True Boatless
A hover to the grate on the left, avoiding the boat.
Bianco 6s: 2nd Cube Dismount Setup
A cue for getting off the cube fast, making the standard left-side strat much more consistent and approachable.
Ricco 4s: Basic
The best candidate yet for a reliable beginner ricco 4s strat with forgiving cycles.
Ricco 4s: Spinless
A yellow-cycle variant that adds a lot more leniency to both the cycle and the peg spacing when approaching the big rod.
R-Only Rocket Storage
An intuitive variant with fewer inputs, so easy to understand, internalise, and replicate by beginners.