If it's not obvious, I'm a picky mapper. I think we all should be honestly. Higher standards are a good thing. I believe in song representation first and foremost, because it's the whole point of rhythm games. My primary enjoyment of maps comes from the satisfaction of notes/colors syncing up to songs I like, and any pattern that defies what the music is doing unironically sounds like nails on a chalkboard to me. This is how I gauge most of a map's quality in every mode, but most of what I'm saying applies specifically to taiko, as it's the only mode I'm skilled at mapping in. Also I mainly play metal/stream maps, so keep that in mind too. I believe a good map is defined by two factors:

Everything should reflect the music in some way
If your map can be slapped over any song of that BPM and fit about the same, it's bad. I've seen people say "taiko mapping is like playing on an improvised drumset" which I find absurd. What's the point of mapping a song if you're just gonna make stuff up? Being able to put anything to any song means no map can be bad, which in turn means no map can be good either. Anyway, what part(s) of the music you choose to follow and where is subjective, but whatever you choose to do with that...

The map should follow a consistent pattern
Every note placed should be a cog in a machine. If you keep changing the rules of what your map is supposed to do, players could become confused and have their immersion ruined, potentially as much as if it didn't follow the song in the first place. This means you should map identical parts of a song the exact same, opting to put variety within the repeating parts themselves rather than forcing it by mapping each one differently. If you think it can't be done and you have to mix things up for the sake of engagement, chances are you're not trying hard enough or are just afraid people will criticize any repetition. As for collabs, I think all repeating sections should be done by the same person to retain consistency, but I do understand the philosophy that each person in the collab should be able to express their own interpretation of those parts, so I won't say you're wrong for doing that.

There is one mapping philosophy that's most important of all. Whatever you do, do NOT base your mapping off of external factors. SR/pp, as well as players' skill/perspectives, will change. Music will not change. Don't force unwarranted variety into your map because "simplicity/repetition bad" despite that literally being what songs do. Don't force your map to be easier because people are butthurt they can't play it. They should just get better, and if more hard maps get ranked, players will be forced to improve and it won't be a problem anymore. Don't force your map to be hard/weird to spite people who make their maps too easy/simple, because that's not much better. Don't force your map to have stuff that inflates pp, because a rework will come to nerf it, and it will be forgotten. Don't map for ranked count or playcount, because these are arbitrary factors which hold no actual value. Strong music-based mapping is timeless, like a well-written story. You will always have a leg to stand on, no matter what changes or negativity you face. Don't conflate this with me thinking it's bad to want a gimmick for your map. I think that's all well and good, as long as its usage has musical relevance.

idk where to put this section, but undermapping is fine if not overdone, to represent lack of musical intensity or to give emphasis to the parts that remain. Slight overmapping (EG short bursts where the drums are only singles and triplets) is acceptable to differentiate between varying levels of intensity, or represent longer sounds like held vocal notes, but blatant overmapping for the sake of difficulty is not. I dislike double BPM greatly. If a song isn't fast enough to be a hard map, then I'd rather it stay easy. I understand why people like double BPM, and I'm not gonna call a map bad just because of that, but don't be surprised if I do because the sheer amount of baseless filler patterns needed to make dbpm maps interesting on some songs violates my first rule above. I'm also not gonna use dbpm in my maps regardless, because it doesn't feel as fulfilling knowing the song isn't actually doing as much as the map is. Don't even get me started on dump though. I can't stand it, and think it should be reserved for the loved category. There, maps can be anything as long as enough people like them, hence the name. Yes, this means I think some songs just aren't viable for some map types. I know some would disagree and insist that any map should be able to exist of any song, but I do not.

Something I don't like is when someone ranks a map that features a skillset they're new to and haven't refined yet, like if my stream one trick self decided to try tech mapping and push the first half decent thing he finished. A tech map "larper," as the kids would say. I'll admit I have pretty high standards for what's considered deserving of ranked, so this is a pretty subjective point. It's possible people want a map ranked simply as a milestone to show where they were in their mapping journey at a certain point in time. Not something I relate to since I've always strived to make sure my ranked maps are my best possible effort for the time, but ofc I don't have to relate to something for it to be a valid perspective. I personally believe mappers should rank things that stand the test of time, rather than things they'll look back at mere months later and go "Yeah that wasn't very good." You can't predict the future, so there can always be maps you'll come to think that way about regardless how much you try in the present (Under the Veil of Madness being that map for me). Even still, I think you should retain the principle of trying to make a futureproof map, which means giving yourself the time to fully understand a mapping skillset before you rank it.

Taiko pp is (and always has been) in an objectively poor spot. With each rework that claims to and fails to fix glaring issues which frankly never should have existed in the first place, my honeymoon phase gets shorter and shorter. I love the concept of pp. I always have ever since I learned what it was in 2019. As someone who is a huge fan of numbers and algorithms, and loves when games use them to rate/compare things, it's easily the most important part of osu! to me and I want to see it improve. However, it has also completely ruined the game. The worst part is that it doesn't have to be this way. You're probably expecting me to rip into farm mappers and circlejerking BNs next for being the ones ruining the system by abusing the hell out of it, right? Well, you would be half correct. I do loathe both of those groups, as well as the NATs to an extent for doing almost nothing about them, but there's a fourth group who I think always goes unnoticed when it comes to their responsibility for pp being in the state it is: the developers.

pp devs are the first line of defense when it comes to how abusable a rework is. If they make it well, fewer maps will be able to break it. Obviously I don't expect them to be perfect. Mistakes happen, farm mappers are annoyingly persistent, and players constantly evolve. I understand that, but I can safely say there's been a great deal of bad handling over the years. The system is held together with duct tape and has many serious bugs. I get time crunches, but if you create something fundamentally flawed, fail to be transparent about how it works, ignore glaring issues and only fix things with band-aids instead of accepting they're bad and just redoing them, how much is it really others' fault that things get exploited? I'm not naming anyone in particular with this. Multiple people have made these same mistakes over several years. What's important isn't who's guilty, but that people stop doing it.

Deploys take time to get right. They need to be made conceptually strong, have informative documentation that can be understood by people who don't know programming or calculus, and have people purposefully try to break them as hard as they can to ensure stability. Any possible exploits should be considered an inevitable meta. If it's not ready, it shouldn't be deployed at all, regardless how long it's been since the last changes. As someone who has been a taiko pp committee member since February 8, 2025, we shouldn't pretend like we can't do wrong and it's all the evil mappers' and BNs' fault. It is their fault, because we let it be. I sympathize with the amount of work that goes into making these algorithms. I've created a complex one of my own in recent years, so I understand the pain of spending so long on something just to realize it doesn't work and I have to start over, but that shouldn't be an excuse to push something bad anyway. As for the community wanting fixes desperately, be careful what you wish for, and be skeptical of the ones happening instead of celebrating just because they kill the current meta. I've been guilty of that myself, which I'll admit may have even contributed to things still being as bad as they are.

I think the impact of the aforementioned paragraphs on the taiko community is highly overlooked. In fact, I would go as far as to say the 2022 rework has ruined taiko mapping. While I still consider the 2020 rework to be the worst in the game's history from an objective standpoint, this one made the system's issues abusable by ANY maps, not just certain kinds, which led to farm mapping being more accessible than ever. Everyone started doing it. New mappers looked to them as reference for how to map, so they started doing it too thinking that's just how it worked. BNs had the same problem, but with modding/nominations instead. Generations of mappers later and here we are. It's normal for maps to be made with sterile patterning optimized for the most comfortable gameplay, regardless of music.

It's not all doom and gloom though. I'm trying (when motivation and coding skills permit) to fix stamina/colour myself. std pp committee member tsunyoku noticed taiko pp code is on fire, and will likely start proper reviews to fix any long standing bugs. Myself and the goat GamersDecision are constantly testing algorithmic issues, revealing anything we find to the team, and experimenting with ways to fix them. Another goat rloseise has a really good reading rework in progress that has potential to finally reward some historically underweight maps (if peppy would just listen to our tiny community and fix gameplay exploits). Things are looking promising, but until then, rip in piss drum pepy point

Now that my general feelings are out of the way, feel free to move on to the next section. The rest of this one will be about the specifics of why I think pp is in a bad spot, so more nerdy and technical. I'm gonna break this up into parts, and keep it brief because this is already long enough as it is.

General
The difference in pp between accs is too small, because acc pp takes way too long to taper off. 200 UR scores are still worth 30-40 acc pp even though noobs can get this UR easily. That's why low stars give so much pp compared to other modes. Making acc pp taper off sooner would fix that, and increase the distance between accs. Also, the SR component of acc pp does basically nothing until 8*, and above that serves as a means to keep high end speed scores from being worthless, because high end speed is underweight to begin with. I can get behind the principle that acc gets harder with map difficulty, but I think it's better off gone from pp for the sake of balancing.

Converts have several blanket nerfs because of TL, instead of a proper TL nerf being implemented in the first place. Despite converts almost always being harder than their normal map counterparts due to repetitive straining patterns, ones that aren't TL are often multiple stars underweight, and are only worth anything if you can get really good acc on high OD ones.

Length bonus is way too weak, and favors acc too much. Why it's too weak is because of something explained under stamina. In the future I hope it goes into SR like std does with its harmonic sum. I like how much less misses are nerfed on long maps though.

Stamina
Stamina SR rises too quickly, making short streams overweight. It also caps too quickly, making very long streams underweight.

It's rigged against monos for reasons related to TL nerf attempts. Monos literally get less of it, so a bandaid buff was put in to directly buff monos, but it's not good either because it immediately resets the moment the color changes, which is not how it should work.

There's something called strain length bonus that basically buffs the SR of anything long enough, so long as there isn't a diffspike, which is why my entire top ten is even as high as it is to begin with, deserved or not. This is not the actual length bonus. It was a very lazy attempt to buff stamina maps.

Colour
This skill makes absolutely no sense and shouldn't directly impact SR to begin with. It arbitrarily buffs certain patterns with no rhyme or reason and ignores their actual mechanical difficulty. Patterns and stamina go hand in hand, so they should just be handled there.

It's also laughably abusable, where spamming kkdkkkddkddd will flood your map with pp.

Good utility skill though.

Rhythm
idek where to begin. It's completely broken, to the point where I'm convinced that it doesn't function at all like intended. Some of the easiest rhythms are the highest SR while some of the hardest literally get none.

pp from rhythm goes down as acc drops, which is good in principle, but rhythm had to be made way too influential to compensate, which makes the maps that abuse it immensely overweight. Again, I think this mechanic is better off gone for system stability, and OD should be changed to matter for rhythm SR instead (it technically does matter now, but to the sum of none due to shoddy implementation).

High end speed was also relying on rhythm being broken until recent patches nerfed it. This is because power of 2 delta time ratios got rhythm for some reason??? Right after a 1/3 or 1/6 is one thing, but most of these maps don't have any divisors that aren't powers of 2.

Unfortunately I lack the skill or experience necessary to have a solution for rhythm.

It's dumb, mainly for two reasons:

BN issues
There are BNs who don't care about map quality and just push anything that follows RC, for a variety of reasons. Popularity, nomination count, friend status, blind mapper trust, Mappers' Guild points, or of course pp, which is how we ended up with the epidemic of bad pp maps, keyword being BAD pp maps. If a map happens to be overweighted because pp is bad, but it's still good, then let it get ranked.

There's a problem of BNs being nowhere near good enough at the game to judge a map. We have people not even capable of passing what they're modding, and the quality of their mods shows it. I understand skill level doesn't always equal level of mapping knowledge, but it's true way more often than it's not. Someone else told me that having BNs for specific SR levels may be a good solution. I'm not completely sure myself, but some consideration for a BN's ability to play a map they're wanting to nom could be good.

Other BNs, as well as NATs, need to put their feet down more and not be as afraid to get their hands on a map that's bad. I think nominating bad maps should come with consequences, but honestly, the more efficient solution to both that and the skill issue would be to make the entry bar for BNs way higher. It's awful how basically anyone can become a BN if they pass a test that's only tangentially related to creating high quality beatmaps.

Ranking criteria issues
Some of the ranking criteria is horrible. Look no further than spread rules. Both having to map multiple diffs just because a song is short, and having to follow arbitrary diff specific rules. My heart goes out to the five digits who have to play nothing but mono triplet spam in their muzus. I understand why spread rule is a thing. In fact, I had lower diffs on A Greater Purpose at one point, but now I think mappers need to come first in the ranking system. Please stop making them cater to noobs if they don't want to. Skill issues are temporary, ranking restrictions are eternal unless repealed. There are already thousands of easy ranked maps for noobs to train up on, and they can play inners within months if they gun for it. This would reduce the amount of modding BNs would have to do on sets, making ranking queues flow faster, which would encourage more BNs to join. That's good, because lack of BNs is a huge problem right now. You wanna rank something? Well, there are only three BNs open, and they only like electronic or vocaloid music. Sucks to be you. I think implementing some kind of optional reward for mapping lower diffs would be a nice middle ground, since noobs being able to enjoy new songs at their level is a good thing, but the diff specific rules would need to be improved. Namely the abolition of rest beats, not all songs support a forced break every so often. If I played almost nothing but stream converts until rank 2k and turned out fine, others can handle a few extra notes. Credit where it's due, though: Both spread and rest beat rules have been nerfed, which is a step in the right direction, but I would still like to say goodbye to them.

As for the mindset that not mapping multiple diffs for a short song is "lazy," I find that to be very idiotic. Amount of total time mapped does NOT equate to amount of effort spent. There can be one TV size length diff that's less lazy than a full five minute spread, simply because the longer one did only the bare minimum to get ranked and didn't focus on quality at all. Not to mention the mapper can map only one diff and get GDers to do the rest anyway. The mindset that easy only sets are inherently lazy is the same story. They're not. Every map, regardless of length or density, has the potential to be well thought out in its composition. So yeah, I do think single diffs should be rankable at any difficulty or length, and you can fight me on that. If you want an example, check out my Viva Evolution Introduction map. I spent multiple hours on it, ensuring that every single choice I made had a pattern and reason. I even explained my thought process in the description. I consider this diff to be very high effort, and would much rather see it in ranked by itself than a set where the mapper churned out 3+ uninspired lower diffs because the RC told them to, and they can't really be that interesting anyway because RC handholds noobs so aggressively. You will be able to tell through the quality of those diffs whether the mapper wanted to make them.

If I have motivation, I'll try to fix it. If I don't, I'll leave it alone or clarify that it sucks in desc if I feel I should. TLDR, if it's mapped before 2025 it's probably not that great, and shouldn't be considered up to my standard of quality.

Voting sucks. The fact that we need a separate site like OMDB is lame (though I like being able to type a review out on there Nvm RIP OMDB).

There should be ratings per diff. Individual diffs can be great or terrible, when the rest of the set isn't, especially on multi mode sets.

Voting needs to be disabled for NF. Too many whiny idiots will NFDT a map just to rate it 1 when they're completely incapable of passing. I agree that people can have valid opinions on a map they can't play, but I think they're a worthy sacrifice to filter out the noise, especially considering EZ and HT exist (and even easier alternatives if you play lazer).

Votes need to be changeable and removable. Right now they're permanent, and there are votes I regret giving long ago.

Votes need to be givable from anywhere, not just on the result screen after finishing a play, as long as you have a pass. I think it's stupid how I have to replay the entire map to vote just because I got carried away by my score and forgot.