Tier Point Overhaul (Concept)

[rushyrulz]

Introduction
I think it's no secret that tier points have lost their way in FFR over the years. The last revision to them was nearly a decade ago, and while the files do still represent a broad range of skill sets, it still seems a bit clunky and arbitrary. The title may be a bit misleading, as I am not actually looking to overhaul the tier point system, but rather replace it entirely, using some existing elements from it as a baseline.

One might ask, what is the problem with the existing tier points system?
First off, I would say that the number of tier points a player has doesn't really tell you much about their play style, what they are unusually good at, and what they need to improve on.
Secondly, tier points being awarded for percentage of AAAs introduces "completionist creep" similar to how the percentage-based achievement tokens like the original Here We Go requirement got harder and harder as new songs were added to the game.

While completionist content is not bad for the game, having it be a component to tier points doesn't fit, in my opinion, when the FC / AAA status bars on the user profile should be sufficient.

The Concept
This isn't the first time something like this will have been proposed for FFR. The reason is that I believe a lot of players feel the game lacks on the statistics side of things: a graph showcasing the strengths and weaknesses of the player. A graph, like the one displayed below, allows someone to easily determine what their own strengths and weaknesses are, and those of the other players around them. We can use the existing tier point system as a baseline by assigning values 1-6 to each existing TP file for each of the categories listed below.



The Categories
Jackhammers:
Repeated notes in quick succession, includes jacks, trills, and to a lesser extent, anchors, running-men.

Speed:
High BPM, high NPS

Stamina:
containing long-lasting sections of intensity and/or lengthy songs in general (I would consider consistency to be a part of stamina)

Technique:
Unusual patterning, grace notes, swing, bpm changes, time signature changes, jazzy drum fills, etc. Anything that breaks the mold from standard 4th note on the beat patterning.

Rolls:
fast, repetitive, may be manip'd. Doesn't need to fit the "dictionary" definition of rolls in the sense that it's 12341234 or 43214321

Jumps:
Gluts, hands, quads, chordjacks, jump density in general.

How do we do it?
I think we can start by using the existing tier point files and assigning whole-number maximum values for each. For example:

Magnum
Jackhammers: 4
Speed: 2
Stamina: 2
Technique: 3
Rolls: 0
Jumps: 4

A AAA on Magnum would award these maximum values which would contribute to the overall total points a player can earn for each category, and the points awarded would scale down to some 0-point (we could use the 1 TP value here for example, beyond which no points would be awarded for the file). The graph would be constructed as a percentage of the maximum, meaning scoring well one one file and simply not playing the rest won't give you a very good graph.

For instance, let's say there are 50 hand-picked files which are chosen to be encompassing of these categories and, when added up, the points totals look something like this:
Jackhammers: 248
Speed: 237
Stamina: 255
Technique: 229
Rolls: 182
Jumps: 261

For a player to reach the edge of the graph in any of these categories, they would need to obtain one of these maximums. Alternatively, if the player does not have any points in a given category, it would be reflected as 0 or nothing on their graph.

The existing Tier points page could even be used to track these macro values, like so:


Who will maintain this?
The ones who should be responsible for maintaining this system would be the difficulty consultants, in my opinion, as they are already the most familiar with what constitutes difficulty in a file, and what is considered challenging for the player base as a whole. Feedback systems for challenging these concepts already exist for the community at large, and could be used for this system as well.

As with tier points, there shouldn't be a need to be constantly maintaining this system. Once a selection of files are chosen, it really doesn't need to be updated unless new files are introduced which do a better job of testing the attributes that a previous one was aiming to do. Let's consider an example of when something might need to be replaced using the existing TP system:

Currently, Club is a tier point file which aims to test primarily jacks, with a little bit of Jumps and Technique mixed in, however it has the downside of being easily mirrorable which can skew results. If a new file were to be released which has similar length, difficulty, and challenging parts, without the capability to mirror for improved performance, it might be a good candidate to replace Club.

Conclusion
Tier points are a relic of FFR that don't seem to tell much about a player's skill level. The song list for tier points has gone unattended for a long time, and there is a looming atrophy of AAA percentage which negatively impacts tier points over time as new songs are added. Overhauling the system and converting it into a graph-based system would resolve most of these issues, while giving tier points a unique place in FFR, rather than its current state of being a secondary skill rating.

In terms of this pitch, nothing is set in stone, not even the categories, graph shape, etc. I do think that this needs to be kept to a finite list of hand-selected files, as a programmatic assignment of points to files would be a huge undertaking, and wouldn't necessarily be accurate. Manual assignment for every file is probably not possible given the sheer volume of content in the game.

Thanks for taking the time to read this. Your feedback would be greatly appreciated!

[rushyrulz]

Rolls/Bursts is probably more appropriate, I'd also like to see this system be accessible to a broader range of players if possible, since the current system caters mostly to D3+

[xXOpkillerXx]

I love that someone is stepping up with a redesign for it again, but in detail. Will read this soon carefully and give my thoughts.

That being said, do note that Velo is incredibly busy recently; there's the enormous MP redo, and more stuff planned too. Still worth discussing it though

[SK8R43]

I see nothing wrong with updating the outdated tier point system. I like the idea and points that you gave. That way we can more easily see a players strengths and weaknesses and it might feel less clunky.
It seems like half of the tier points are moreso pretty random easier files. I remember when there were no(or way less?) easier files in the tier point system as well.

[rushyrulz]

Originally Posted by xXOpkillerXx

I love that someone is stepping up with a redesign for it again, but in detail. Will read this soon carefully and give my thoughts.

That being said, do note that Velo is incredibly busy recently; there's the enormous MP redo, and more stuff planned too. Still worth discussing it though

For sure, I'm mostly just aiming to spark conversation at this point rather than REHAUL TPS NOW! Lol I'm super excited for the MP changes too.

[xXOpkillerXx]

Ok so. My first reaction to this was the following:

- This looks like a system that partially wants to mimic Etterna's "skill buckets". I have my opinion on that, but this is just an observation.

- It's always nice when someone starts a new idea, but often times when I ask them to provide a non-negligible example subset, they don't. Are you willing to try to do that (and talk to some people to confirm your example) before asking the difficulty consultants to do it? One song is far too little of an example to properly discuss/debate; and it might not present your thoughts optimally.

- This last point gets closer to the diff calc concepts I've debated extensively with some others (Wirry primarily), but the main question is: while you said the categories aren't set in stone, do you believe that there actually is a strong set of similar high-level categories that can describe charts accurately enough? Why?

[XelNya]

What it sounds like you actually want to do, is assemble a system similar to Etterna's where it tells you what skill sets are more likely required (Etterna isn't perfect) rather than overhaul the useless system that should have been deleted years ago

But my stance is cut the arbitrary file count, and do it for all of them, don't make it some circle jerk awards system that's gonna be outdated on arrival again

But thanks for finally realizing the tier point system is bad. Glad to have you in the camp.

[rushyrulz]

Originally Posted by xXOpkillerXx

Ok so. My first reaction to this was the following:

- This looks like a system that partially wants to mimic Etterna's "skill buckets". I have my opinion on that, but this is just an observation.

I've only played Etterna a few times, but yes I think this is very similar.

Originally Posted by xXOpkillerXx

- It's always nice when someone starts a new idea, but often times when I ask them to provide a non-negligible example subset, they don't. Are you willing to try to do that (and talk to some people to confirm your example) before asking the difficulty consultants to do it? One song is far too little of an example to properly discuss/debate; and it might not present your thoughts optimally.

I probably could, however it would be limited to my direct experience with the game, which is why I think a collaborative effort by those who have already taken the time to analyze this stuff would be best. But as a proof of concept, I can give it a shot at some point after more people have a chance to provide their feedback.

Originally Posted by xXOpkillerXx

- This last point gets closer to the diff calc concepts I've debated extensively with some others (Wirry primarily), but the main question is: while you said the categories aren't set in stone, do you believe that there actually is a strong set of similar high-level categories that can describe charts accurately enough? Why?

These are simply 6 (an arbitrary number) aspects of the game that I think are unique enough to warrant their own branch. There may be some things I've forgotten or didn't think of, and some that I mentioned may not be broad enough. Again, this is intended to be more of an open concept that others can provide input for. As for why I believe it's possible, the fact that we've seen similar concepts in StepMania and Etterna speaks to the feasibility of the concept.

Originally Posted by XelNya

But my stance is cut the arbitrary file count, and do it for all of them, don't make it some circle jerk awards system that's gonna be outdated on arrival again

In a perfect world, the whole universe of FFR files, including ones released in the future would be included, then you would just get a mapping based off your top 50 or 100 scores. However, until there's an automated way to assign these values, and do it accurately, it's just not a realistic goal.

[xXOpkillerXx]

Fair enough. There's a lot to discuss but I won't flood this thread; I'm also curious to see what responses you get. Feel free to DM if you wanna chat about it.

That being said, the difficulty consultants don't really have much more qualifications than you to provide a sample set; I'd even argue you could very well do a better job (or not) than them, simply because this is not really comparable to what they do and also because you've played this game for a very long time too.

[Zyxel]

This is so sick, I love the concept of this

[kmay]

I think if we are to make a skill map like that we should be utilizing a files difficulty to map the points.. what make the file a 92... a 46... it will make things a lot harder to first implement but definitely gives a more accurate value to a skillset

A skill map will need to be inclusive of a player no matter the skill and also should aim to be set at some max level... which could just be something like

song has been played: this is the skill map
update total skill points: previous + new song skill


So your skill map is only based off songs you've actually played rather than the current system of here are songs you have 0 TPs on and might never even get any on


something like this...

If you AAA Magnum
Magnum - 80
Jackhammers: 30
Speed: 10
Stamina: 10
Technique: 15
Rolls: 0
Jumps: 15

If you don't Magnum it would take that AAA equiv and multiply by the max points to give you a skill map


Magnum - 80
Jackhammers: 30 * equiv
Speed: 10 * equiv
Stamina: 10 * equiv
Technique: 15 * equiv
Rolls: 0 * equiv
Jumps: 15 * equiv




previous skill total: 37000

Jackhammers: 3800
Speed: 6000
Stamina: 5500
Technique: 4800
Rolls: 7000
Jumps: 3600

new skill total: 30780

Jackhammers: 3830
Speed: 6010
Stamina: 5510
Technique: 4815
Rolls: 7000
Jumps: 3615

this means we would also have to track the possible max skill points they could have and then make the map off of that so maybe


Max skill Points: 39300

Jackhammers: 6500
Speed: 8000
Stamina: 6300
Technique: 5500
Rolls: 7500
Jumps: 3615


So each little peak of their skill map will reflect a percentage of the total points

[Black_Shield]

Originally Posted by Zyxel

This is so sick, I love the concept of this

i was about to say, you already made graphics for this exact thing

[ULTIMEGA]

I never actually thought about the Tier Points system. I always looked at it as something to work on as a player. But I think you make a very good point in the system needing to be overhauled and revised from the ground up. It may actually make a huge change to the game in the future.

[gold stinger]

I honestly think the only reason Tier Points hasn't been touched in as long as it has been, is because qqwref & Sprite of all people were the last ones to touch the code regarding it back in 2011 and it's baked into the website through levelranks.

Why mention levelranks? Because there's more stuff broken in levelranks than just Tier Points. None of it fixed because it either needs to be revamped or overhauled to begin with in order to fix, since it's written in long-lost unintelligible website code, and neither coders are around to reference it.