My name is (perhaps unsurprisingly) Daniel. I am a mathematician from Britain, specialising in algebraic number theory, but I have always had a love for speedrunning and TASing. I grew up (in a loose sense of the phrase) playing the Super Mario series and other platforming games, so these are the games that I tend to focus on.

Another key fact about me is that I like rambling, puns and horizontal rules.

I started TASing in 2013 with NES Super Mario Bros. and various romhacks. This was mostly a learning experience to get to grips with TASing, and likely most of what I created during this time was poorly optimised. Nevertheless, this was a fun low-stakes way to learn.

Since then, I have taken the daring leap to other NES platforming games, and even recently to a platforming game on the SNES. TASers marvel at my breadth of talent. More seriously, I tend to search for platforming games without TASes that look like they would be cool to play around with, and where the available TAS tools for the console are not painful to use. In theory, I am interested in the all platforms challenge, but you can see what happens in practice.

I am open to suggestions for new games to TAS, but it is good to keep in mind the above, as well as my current projects (see below).

Jelly Boy is a SNES platforming game where you play as a jelly baby who has come to life and wants to escape the factory he was created in. There are many powerups throughout the game that transform Jelly Boy into various objects with different attacks and powers, and many puns and British idioms throughout - what more could I want.

I have created a rough version of the TAS, and an optimised version of the first three worlds of six. The latest optimised WIP (world 3) can be found here, and some of the larger skips can be found in this playlist. I am hoping to finish the TAS by the end of September 2024.

Street Fighter 2010 is a NES platforming and beat-'em-up game, unrelated to the fighting game series of the same name. It has one of the oldest non-obsoleted publications on the site, and as a result, it is one of the few remaining targets for the First 500 project:

I originally worked on improving this with negative_seven in 2014 and 2015. However, it is a very challenging game to optimise, so we didn't progress very far into the game. In 2023, I have started collaborating with eien86 and Challenger on this. We are currently at the end of the 2nd planet of 5, and are around 25 seconds ahead of the published TAS. A WIP of the first planet can be found here.

Darkman is a NES platforming game loosely based on a film of the same name. As someone will tell me every time in the YouTube comments, its gameplay is somewhat similar to Robocop 3, created by the same company. The rough plot is that Darkman disguises himself as the boss of each level, does some sick parkour to the end and wins. Something like that, anyway.

At the time of writing, I have created all three published movies of Darkman:

I started work on the first of these in November 2014, inspired by a cancelled submission from Niamek that I thought I could improve significantly. I ended up saving 25 seconds, mostly through the discovery of a helpful glitch that allows you to attack enemies repeatedly without having to stop to fight them. This also makes the run much more fast-paced and interesting to watch. It eventually got published in March 2015.

In the YouTube comments for my first movie, I received a comment bringing to my attention a glitch where knockback is applied to enemies in the direction you are facing when they get hit. This allows you to influence which way the knockback is applied during the parts of the game where your character has a projectile attack. Using this in two places, I improved my original TAS by 2.63 seconds, and this was published in December 2015.

In November 2021, when showing off my second TAS to a friend, I noticed that the improvement to the Carnival boss (Spike) was likely not optimal. In fact, I had tried to make this precise jump work when working on the first movie, without success, but this time I managed to make it work, saving over a second. However, I was very busy with academic work at this time, so I put off redoing the TAS. Finally, for much of August 2022, I worked on finding several more improvements, some very large, and some only a handful of frames. With some significant analysis of how the game handles objects and frame rules, with some debugging help from negative_seven, I ended up saving 23.38 seconds on the second TAS. I would expect that this is the final iteration of the Darkman TAS (at least from me), but you never know :)

Superfast Mario Bros. is a fairly well-known romhack of NES Super Mario Bros., created by acmlm, in which everything travels very fast. It is probably most well-known due to the somewhat comical graphical glitches appearing in the background, as the console can barely keep up with the speed at which it needs to load the levels.

I have been involved in two of the publications for this romhack:

The first of these started as an April Fools' day submission in 2014, since negative_seven and I knew that any submission of the hack would immediately get rejected under the site rules of the time, due to the quality of the hack itself. Nevertheless, the TAS itself was fairly well-received, earning a place on the gruefood delight page. Nine years later, the site rules had advanced sufficiently to allow runs of many more romhacks, and the hack was brought to attention again by despoa's warpless TAS, allowing it to be published with some small improvements chatterbox had found in the meantime.

A short time later, eien86 found a new trick involving vines using his JaffarBot, and I found a few small optimisations in 4-2, 8-3 and three of the rooms in 8-4, leading to the second publication. I think this is a good example of how botting and human TASing work well together - JaffarBot was excellent at optimising away the lag frames that plague this romhack, but missed some of the more conceptual timesaves that I added in.

Super Mario Bros. Deluxe is a remake of NES Super Mario. Bros for the GameBoy Colour. It features many additional gameplay modes on top of both the original levels and the 'lost levels' from SMB2J. I have one publication for this game:

I was doing some testing for a potential warpless run of the main quest, as a GBC run for the all platforms challenge (see below). The only up-to-date TAS of the earlier levels was negative_seven's warps run, in which I noticed that a large amount of time was spent waiting for piranha plants. As with the bridge animation in warpless runs, this can be avoided with a fire flower, and after sufficiently optimising the item grabs, this route ended up 12 frames ahead, despite not being able to find any optimisations outside of the route change.

negative_seven and I also joined teams together for the so-called 'Dream Team Contest' on multiple occasions. In DTC5, we initially teamed up with NitroGenesis for team TNT, but he was inactive for the first half of the contest. Having suffered the same fate with their teammates, we eventually merged with Truncated's team halfway through, to form team TTT (can you figure out why we called it that?) In the end, we finished 4th of 9 teams. Luckily, we managed to do one of the rooms 1 frame faster than anyone else, thus booking our place in the eventual publication:

Metal Force was a very fun game to TAS, and I greatly enjoyed the contest. I can't say the same for Zook Man ZX in DTC6, where we added Aglar to our team ¯\_(ツ)_/¯, although if I recall he was not hugely active. However, we ended up finishing 3rd of 8 teams this time around. The contests page tells me that for DTC7 we added DrD2k9 to ¯\_(ツ)_/¯ instead, and we finished 3rd of 5 teams. I don't remember a thing about this contest, but the discord logs tell me that I was significantly more active than I remember.

For DTC11, I teamed up with Moozoh, invariel and FatRatKnight, finishing 3rd of 6th teams. Invariel and I also made it into the final publication:

It was interesting to branch out into shoot-'em-up games, and although much of Crime City was not too tricky to optimise, it was fun to find strategies for the more technical bossfights.

Although not strictly TASing, in August 2016, negative_seven and I debugged Super Meat Boy to figure out what was going on behind the scenes in the so-called 'wrong warp' glitch. It turns out it is not so much a warp as using memory corruption to alter which chapters of the game are currently unlocked. We used the understanding to improve the speedrun route by about 10 seconds, and also came up with a route to beat the chapters in reverse order, which we called 'reverse boss order'.