I love this game, it's the best. So onto info

Basic info: I'm not going to break down the basic inputs and whatnot, but I will comment on the fact that I cancel any animations I can into running, and immediately end the run. This puts an early end to the animation, which lets me attack faster constantly and pull off some stuff the developers didn't anticipate (especially tapeworm mashing). Besides that, I gleam at every opportunity, since gleaming does something like tripling your damage output and recovers mp faster (which doesn't often matter with the strats I am using). Knowing who to gleam off of and how to maximize the damage you get off a gleam is critical to getting a good run. Hence, I need to know EXACTLY where every enemy is going to spawn and EXACTLY how to best coax them into attacking as fast as possible

As far as tape use, for stances, power gambit is pretty much the only really useful stance for a speed run. Some of the others have uses (most notably I'll equip balance or training wheels if I pick them up before the first chest, and stunner if it pops out of the chest), but power gambit really speeds the flow of the game and puts your damage into the range to kill things as fast as I do here. As far as skill tapes, knee drop and lightning are the only two skills that see use. Knee drop does a ton of damage quickly, and allows you to play a lot riskier since it gives you a way out when things go wrong. Lightning is used to clean up some fights in the tail end of the game. There are some skills I don't use that I know people might ask about, Spin kick being the primary one. Spin kick is a decent skill, HOWEVER it takes a long time to get to the real meat of its damage, and it costs quite a bit. Basically, knee drop does everything spin kick does, but better. Bro-dozer used to have some use during the tube ride, but they didn't pan out, and bomb toss allows for big combos on bosses, but just letting them continue their attack patterns and re-gleaming into knee drops ends up being a lot faster.

Anyways, more details on game play stuff will come in on individual levels.

Mission 1: Main Street. AW MAN, NOT THIS AGAIN is a phrase that will haunt me forever. This quote is such a perfect way to start a run that resets 75% of the time in this stage. Overall, stage 1 is extremely odd. On one hand, almost everything has to go right in order to not lose a large amount of time due to being underpowered from not having a stance. On the other hand, going right can actually mean a lot of things because of how the fights work out. The second fight in particular has about 4-5 variations on how it goes that all end up within about a second of one another. Stage 1 also makes heavy use of technical tricks with the inputs and precise positioning in order to make up for that fact that I don't have Power Gambit yet. Regardless, the chest has a 1/20 chance of dropping a power gambit (which is doesn't here), and Abobo has a 1/4 chance of dropping them. If I don't have power gambit after the first Abobo, it's an automatic reset since I will lose something like 10 seconds in the very next fight, and it only gets worse from there. This run gets the drop, and so it continues. Since damage values are on, it's pretty easy to tell how big of a difference a stance makes (gleamed whip hits go from doing about 130 damage to doing 200). I'd like to have sub 2 minutes going out of this stage, but 2 flat is more than an acceptable time considering how much can go wrong.

Mission 2: Back Alleys. For some reason it took me forever to figure out that if I stand too close to the first Linda, she won't kick. I had some pretty stupid resets because of her. This starts to show off how important knowing where to gleam is, and also shows off some of what run canceling attacks does. Normally you wouldn't be able to kill a Linda in the middle of a whip animation before she hit you, but thanks to run canceling I can rush her. If I am lucky on the rooftop fight, the barrel can kill both Lindas by itself and save some time, but it's pretty rare and entirely RNG by what damage roll I get on the barrel. The Abobo in this stage gives me a second power gambit, which is nice.

The Skullmageddon fight goes exactly as planned, his pattern is extremely predictable. The only way I could save time here is if I had 3 power gambits or score a crit, but either of those would only save a second. I don't bother picking up mithril since it doesn't cost any time and wouldn't be used anyways.

Mission 3: Space Lair. A lot of little things went wrong here, none of which individually are much but all add up to a noticeable loss. I stay on the left for the first fight because throwing the Hoverbizzle at the two approaching from the left is very inconsistent. I thought the explosive Hoverbizzle would finish off the last Williams in the second fight, and I miss a gleam that I can easily make up for in the 3rd. The last wave in the 4th fight is very tricky, there's not really a good place to stand to get the Geisha and be in position to attack all 3 Williamses. It looks fairly sloppy here, but the fact that the Williamses decided to bunch up at the end means I didn't lose much if any time. The last fight is a painful one for RNG, the Geisha's behavior is random and they can really throw a wrench in my plans. If too many enemies are stacked on the Abobo while I hit him with the tazer, it can actually increase the animation time so much that the Abobo hits me with one of his counterattacks. I finally get a knee drop here, which I equip for use until stage 10. I was actually really worried about this, because after mission 4, the Williamses tend to not be the basic ones, and non-basic Williamses no longer drop it, and it is extremely unusual to not get one before the end of stage 2. Sometimes the money collection at the end of the stage triggers incorrectly, which lets me save ~5 seconds or so out of nowhere.

Mission 4: Airlocks. Airlocks are tricky because while I have a plan, the nature of the airlocks means that if the plan goes wrong at all, the timing is completely jacked and I have to improvise. That said, the first and third airlocks went pretty much perfectly, as did the fights between airlocks. The second airlock is without a doubt the hardest to get right, and unfortunately I wasn't able to manipulate the Lindas to throw their grenades and attack as fast as I'd like. Overall, I'd say I could finish that fight about 1-1.5 cycles of the airlock faster, but given how tricky it can be I'm very happy with this fight. Concerning the tapeworm: Run canceling jabs at the right rhythm allows me to hit him as many times as I do here. The first Abobo here gives me a power gambit, making 3.

Mecha Biker gives a decent pattern. The best possible pattern is him doing the flame shot on the first cycle, since it lets me kill him during his second pass and him dying immediately on the third pass. Flame shot on the third cycle is actually the second best pattern since it means I will still be gleamed when I karate finish him after he falls off his bike. The knife then does its work, although I only need 3 hits thanks to the gleam.

Mission 5: Countryside. In 2p, Jimmy says "Terraffirmative!" at the beginning of the stage. Losing this line is probably the most unfortunate thing about running this game single player. I missed the falling Williams by a lot. Shuns make their first appearance here, which is important because Shuns are EXTREMELY obnoxious if you let them live. For some reason they have massive amounts of invulnerability on very random actions such as jumping. Fortunately, run canceling kind of breaks things against them; normally it is not possible to land two strong attacks on a Shun when they jump after their initial attack. The first fight is especially nasty, there are a lot of Shuns here, and missing any of my planned attacks almost automatically means 5 or more seconds lost. I absolutely must have level 9 power gambit after coming out of this stage, and preferably level 10. Usually I end up with only 9 and try to use the scythe to force a drop from the Abobo, but in this run I didn't need to as I already had level 10 (with enough money to spare for another level if I needed it!). The end of this stage is pretty stressful when I get a good run like this going, as sometimes the game will crash for no apparent reason after killing the Abobo. Also; you are completely invulnerable during the post-stage celebration.

Mission 6: Assault Zone. Getting to this stage usually means really good things for the run. Not having to worry about power gambit drops or common game crashes means it's mostly execution from here on out. This stage also allows for mad sick dancing skills. The fight after dancing is actually extremely important. While it looks simply like an optimized fight, if you have not finished all the enemies by the time the helicopter door has closed, it triggers a 35 second long animation. This is the reason it is so critical to have at least level 9 power gambit, as that's what is needed to kill shuns in one gleamed strong attack. At level 9, there is a possibility of rolling a low number for damage and it not working, but at 10 that risk is gone. Anyways, the last Abobo trolls me a bit by selecting an attack I can't gleam, and the last Shun kill is sloppy, but overall this was a strong stage.

Giant Tank: I get pretty bad luck on the tank overall. I take two pellet hits, and am forced to stop to dodge one of the mine things the tank drops in the first phase, and I get to the smoke vents at a really bad time during the second phase. I take a hit in the third phase, but I get to use the invisible barrel glitch to save picking up the barrel, so it works out ok. I would actually say the giant tank is one of the worst overall sections in this run, but really it boils down to getting really bad luck, so I don't mind it too much. Plus, even with all the bad luck I had here, it costs less time than a single more common occurrence of bad luck at any number of other places.

Mission 7: Some kind of lab. "AW MAN! More like science Unfair!" is another fantastic line. I actually gain time compared to a 2p run in the beginning of the stage because Billy and Jimmy have a 10-15 second unskippable conversation at the start of the stage. This mission introduces Roxy, an enemy most people will hate having a jetpack, but in reality they are extremely predictable (possibly the most predictable enemy in the game). The very first Williams trolls me pretty hard by not doing a cartwheel attack, but the positioning I end up in as a result helps make it not cost much. The last wave also gets messed up a bit by the explosive Hoverbizzles getting caught by a stray attack I didn't expect them to. The silhouette section here is the last place I've seen the game crash on a NG run, so getting past that point is nice even if it is an uncommon occurrence. I get a bit of bad timing on arriving the claws (which are from Double Dragon 2, not Contra), and the last Roxy decided she desperately wanted to live. Bimmy and Jammy participate nicely in project ‘Knee drop is OP'. I have mixed feelings overall about this stage, because on one hand a lot of things went wrong, and on the other hand even the total of those things didn't really cost all that much time.

Mission 8: Genetic Research. The elevator here is another section that kind of goes badly. I miss a gleam on a Geisha, which results in me being out of position for the next fight. I feel like I made the right decisions to minimize time lost, but it's definitely a section that did not go as planned. The last Williams on the second fight really confused me, I was convinced he would attack immediately because of how far he had walked, but he decided to delay anyways. The 3 Shun fight between claws lags really badly if you don't end it quickly for some reason. The Roxy on the falling bridges is one of only two times in the game I slow down to attack an enemy that isn't required to stop and fight, as just running results in her dive-bombing you. Last fight has potential to be a lot faster if you are cornered to gleam knee the second wave, but it's only a few seconds lost to do it this way.

Marian II goes alright. Her initial pattern is extremely important in getting a good time, a bad pattern can cost 10+ seconds on the spot. The ideal pattern is for both pods to attempt a horizontal attack while being on the same line as each other, and to correctly guess the opposite side from which one attacks first (in other words, almost completely impossible). This is also the last fight I had been using spin kick for, but I discovered during a run where I somehow didn't pick up a spin kick that knee drop allows you to attack again so much faster that it didn't even matter that it’s second hit wouldn't hit the pods. I actually could have gotten a really excellent time if I had finished off the shark head before allowing it to get off screen, but this is still a really above average fight. I even managed to gleam off a down strike, which is basically a wild guess as there's no real cue for it that I've found. Also knee drop after the stage ends.

Mission 9: Tube Ride (ignore what the game says). Tube ride is where most people get really annoyed and give up on the game. It actually works out to be a pretty cool section in a speed run, though, since you can gleam a lot and enemies just drop onto you. I make several errors here, unfortunately costing me a good 5-8 seconds. This is also one of the sections where errors can cost on the order of 10+ seconds, though (specifically if you get knocked off the platform), so I'll take it.
Mission 9.5: Haunted forest. Haunted forest is all about optimizing wand usage and some tight use of shoulder charges to kill a lot of enemies at once. This run had almost flawless wand use, the second fight in particular usually takes a lot longer, but because I caught both skeleton Shuns with the last wand shot it went very fast. I bop my head on one of the spike things and feel stupid. Oh well. Almost everything else in this stage went flawlessly, if I had gotten the money skip this would have easily been a PB on the stage.

Mission 10: Neon Fortress. This stage is really long, and it's extremely important to do a few specific things in it. It also has a few counter-intuitive ways to save time, such as gleaming off the same Williams twice without attacking him to have a gleam later in the fight, and taking a hit from a Roxy for the same reason, as well as some uses for some otherwise not useful techniques, including pushing a Williams down a pit, and later using the high five to drop a sword to conserve its durability. The first fight can be a real pain; it introduces Bao Boshis, who have few attacks that you can actually gleam, and often the other enemies in the fight don't like to attack. It went almost perfectly here. I specifically wait to kill the last Linda in the silhouette section, since her dead body hits the Abobo with just enough force to finish him off after the 2x4 hit. I get bad luck on the first claws, but knee drop my way through it. The elevator has a ton of enemy manipulation to get Williamses to cartwheel into pits on their own, and to force the Abobos to whiff their reactionary attacks. I did, however, forget that I had lightning already equipped and went into the menu to check. This doesn’t actually cost any time, but it looks really sloppy and annoys me more than it probably should. With the exception of the second to last fight, everything went exactly as planned for this stage, and the plan for that fight involves a ton of luck with how things work out. This was actually a PB for this stage by about 5 seconds; it's difficult to imagine it going any better

Evil Marian actually gives me the worst possible pattern for this fight. It still only takes 6 seconds from when I can move to when the fight ends. I don't think there's much more to say about that. Also butt-pipe.

Mission 10.5: The future or something. All the enemies take damage cap here always, meaning lightning is OP. Ideally I would punch the Abobo before the roll to save a lighting cast but w/e. DEATH BOOOOOOOST!

Giga Skullmageddon goes fairly well. During his second set of swings he threw me for a loop, usually he goes top-bottom-center, but decides not to do that. That said, I still killed him without getting to his long invulnerability period, so after a run this good I am extremely happy with getting this. He then takes the long fall to the fantastic credits sequence and a showing of the time, a wonderful 38:01.

I'd like to thank everyone who checked out my stream while I was making attempts, you definitely helped keep me sane while resetting for power gambit drops. This goes doubly to the Way Forward staff, who not only made this wonderful game, but also gave me the privilege of their presence in my stream, and even answered some questions I had on game mechanics and such. I can honestly say I never thought I'd be chatting with the people who made a game while I beat it down; it was an extremely unique experience I will not forget. Next is RaneofSOTN, who did a lot of the initial footwork and experimentation with skills. Not much of that initial data may have made it to the final product, but it certainly saved a lot of time finding the right direction. A definite thanks to all the SDA staff for all the hard work you do so people like me can post things like this. And lastly I'd like to thank all the people who checked out this final produced run, I hope you enjoy watching it as much as I enjoyed getting it.

Oh, and don't think this is the end for Neon, it's just the beginning.