forum

osu!catch ScoreV2 Discussion

posted
Total Posts
83
Topic Starter
smoogipoo
Hey all,

This might be a little bit too early (a whole 9 months), but here's goes nothing!

I've had a few people interested in discussing osu!catch ScoreV2 so I'm opening up a thread to collect the community's feedback. I'm interested in hearing your opinions about what you like/dislike about the current system and what you want out of the system so we can get it perfected before the next CWC.

THIS IS NOT FINAL
Please, do not discuss Star Rating and PP here.

Here are a few facts/talking points about the current osu!catch ScoreV2 system:
  1. >90% of the score comes from combo, the <10% comes from tiny slider ticks.
  2. The percentage worth of tiny slider ticks is scaled by the ratio of tiny ticks to all other objects in the map.
  3. Bananas are worth a flat 350 points each.
  4. The mod multipliers are as follows: HR - 1.12x, DT - 1.06x, FL - 1.12x, HD - 1.06x
Basically, I'm willing to experiment trying different stuff so throw out any ideas you have. I'll be checking this thread periodically but don't expect me to reply to every comment or release any updates any time soon in an effort to not impact existing ongoing tournaments - however I'll try and get some sort of way to test it working.
And please _please_ don't spam me with PMs telling me to read a comment in here ;___;.
Riari
I'm a very silly person.
Topic Starter
smoogipoo
Your ScoreV2 multipliers are for osu!std. osu!catch are as I listed.
Tonairu

Riari wrote:

I'm a very silly person.
same tbh
Riari
What do people think about scaling banana values similar to small droplet values?

Inside osu!catch spinning is not a mandatory action, the score is a bonus on top of what you get for playing the rest of the map.
Currently in ScoreV2, droplets scale up to 10% of a maps value and would naturally diminish their value in longer maps. Bananas on the other hand have a constant score value of 350 giving them a significant edge in the longer run.

I think that whilst high-end spinner plays are impressive, droplets are a core part of osu!catch scoring and should not be so heavily reduced in comparison to bananas.
I think that a score cap that is similar to what small droplets face right now for bananas would be helpful for competitive scores as it would allow accuracy to make a bigger impact on maps that contain spinners.



I also saw something over in the standard discussion thread where score could possibly be multiplied by your accuracy. I think this would work much better over here and give misses and droplets a much more competitive edge.
Krah
Too lazy for a full post but just for the mods multiplier.
HD and DT shouldn't be a fixed multiplier
- HD become easier with high AR and actually a bunch of people can't even play high ar without hd (or even on all cwc maps aka ar8+). Just look for that the rules of the last cwc where people were allowed to pick hd on their dt/hr picks.
- DT is a pain on ar9+ and shouldn't be lower than hr in those cases.
Riari

Krah wrote:

Too lazy for a full post but just for the mods multiplier.
HD and DT shouldn't be a fixed multiplier
- HD become easier with high AR and actually a bunch of people can't even play high ar without hd (or even on all cwc maps aka ar8+). Just look for that the rules of the last cwc where people were allowed to pick hd on their dt/hr picks.
- DT is a pain on ar9+ and shouldn't be lower than hr in those cases.
I think that HD is a crutch mod. The majority of people who play with the mod play with it as a requirement to play the game and are heavily rewarded for it because of the 1.06x multiplier. I personally like the way that visualization mods are treated in mania where they receive no multiplier. However, I think that Flashlight as a visualization mod is not a crutch and should be rewarded correctly.

DT and HR are polar opposites in terms of difficulty in their level. At lower levels, HR is the harder mod whilst DT is easier. As the difficulty ramps up, DT becomes the harder mod (AR has a influence on this, mainly due to the mapping meta of ~AR9) and HR is comparatively easier. Between the two I would rate DT higher personally as towards the higher end of play the scores become slightly more warped and see more interest from the community, though I do like the idea of a scaling multiplier as it would make things more accurate.
iiyo
later on in this thread i'm going to get together with some top players and make a giant wall of information and how things should be, may take some time tho, I'll write up a giant paper when i get around to it.


DT SHOULD BE WORTH MORE THAN HR

I can say this without even needing to consult any player.
koliron

Krah wrote:

Too lazy for a full post but just for the mods multiplier.
HD and DT shouldn't be a fixed multiplier
- HD become easier with high AR and actually a bunch of people can't even play high ar without hd (or even on all cwc maps aka ar8+). Just look for that the rules of the last cwc where people were allowed to pick hd on their dt/hr picks.
- DT is a pain on ar9+ and shouldn't be lower than hr in those cases.

Fantasy wrote:

DT SHOULD BE WORTH MORE THAN HR
Totally agree, specially the last one, unfortunately in ar 5 and less DT is so much easier than HR, so is hard to say which multipier would be fine
Imo both sjould be +0.12, i mean i dont think HD HR deserves the same as DT HR D:

Btw FL should be +0.16 /run
Digitalfear117
I also agree that DT should give more score than HR, mainly since I believe that scoring should reflect PP, where in most cases DT gives more PP than HR

But I don't think a simple switch is enough, here is what I personally think the mod multipliers should be:

HD 1.04
HR 1.08
DT 1.12
FL 1.12

Now I'm not sure what should be done on maps where the star rating is increased more by HR than DT, maybe there can be an exception and it can be switched on maps where this happens?

As for slider ticks, I think they should be worth a little more, since we currently have scores on the leaderboard that are beating SS's with C's (using the same mod combinations) because of better spins

Edit: I see we are not supposed to talk about star ratings and PP but I feel like it's hard to make my point without at least mentioning why I think it should be this way...
Nelly

Fantasy wrote:

DT SHOULD BE WORTH MORE THAN HR
Totally agree, not because of the score but the PP system either. The HR trains AR and low CS while DT will train AR and reflection. This is even better than other mods, for hidden im keep that in silence. Fantasy, Amlink, where are you guys?!

P.S. Hidden is OVERFLOWING the CtB system
P.S.S. Im not talking about the PP system here
PakaChan
DT 1.14x
FL 1.1x
HR 1.06x
HD 1.03x

DT > FLHD > FL > HRHD >HR > HD
I don't think it should be any other way as long as the multipliers are static and don't care about CS and AR. Maybe 1.12 DT if you want FLHD > DT
Riari

PakaChan wrote:

DT 1.14x
FL 1.1x
HR 1.06x
HD 1.03x

DT > FLHD > FL > HRHD >HR > HD
I don't think it should be any other way as long as the multipliers are static and don't care about CS and AR. Maybe 1.12 DT if you want FLHD > DT
FLHD>DT in your case as I think the multi-mod thing would still apply so FLHD would be 1.41x or whatever it is?

I may be confused but I'm sure that's a thing.
PakaChan

Riari wrote:

PakaChan wrote:

DT 1.14x
FL 1.1x
HR 1.06x
HD 1.03x

DT > FLHD > FL > HRHD >HR > HD
I don't think it should be any other way as long as the multipliers are static and don't care about CS and AR. Maybe 1.12 DT if you want FLHD > DT
FLHD>DT in your case as I think the multi-mod thing would still apply so FLHD would be 1.41x or whatever it is?

I may be confused but I'm sure that's a thing.

1.1x1.03 = 1.133 < 1.14
Riari

PakaChan wrote:

Riari wrote:

FLHD>DT in your case as I think the multi-mod thing would still apply so FLHD would be 1.41x or whatever it is?

I may be confused but I'm sure that's a thing.

1.1x1.03 = 1.133 < 1.14
Mods are additive not multiplicative?
PakaChan

Riari wrote:

Mods are additive not multiplicative?
Note that when multiple mods are applied, the product of the score multipliers is given instead of the sum of them.
- https://osu.ppy.sh/wiki/Game_Modifiers

The numbers themselves shouldn't matter much, it's how much mods and mod combinations give compared to each other that matters.
Saying "DT > FLHD > FL > HRHD >HR > HD" is more usefull than giving some numbers imo.
Riari

PakaChan wrote:

Riari wrote:

Mods are additive not multiplicative?
Note that when multiple mods are applied, the product of the score multipliers is given instead of the sum of them.
- https://osu.ppy.sh/wiki/Game_Modifiers

The numbers themselves shouldn't matter much, it's how much mods and mod combinations give compared to each other that matters.
Saying "DT > FLHD > FL > HRHD >HR > HD" is more usefull than giving some numbers imo.
Ah, I always took at as the game value and thought it was slightly higher because of double mods.
PakaChan

Riari wrote:

What do people think about scaling banana values similar to small droplet values?

Inside osu!catch spinning is not a mandatory action, the score is a bonus on top of what you get for playing the rest of the map.
Currently in ScoreV2, droplets scale up to 10% of a maps value and would naturally diminish their value in longer maps. Bananas on the other hand have a constant score value of 350 giving them a significant edge in the longer run.

I think that whilst high-end spinner plays are impressive, droplets are a core part of osu!catch scoring and should not be so heavily reduced in comparison to bananas.
I think that a score cap that is similar to what small droplets face right now for bananas would be helpful for competitive scores as it would allow accuracy to make a bigger impact on maps that contain spinners.



I also saw something over in the standard discussion thread where score could possibly be multiplied by your accuracy. I think this would work much better over here and give misses and droplets a much more competitive edge.
You could also consider giving droplets a static score like how it's now except make it so it actually matters when compared to that of a banana. Right now 1 banana = 110 droplets, you could always just change things up to make it 1 droplet = 1 banana or 1 droplet = 3 bananas or 3 droplets = 2 bananas. Giving both a multiplier sounds messy and giving just one a multiplier makes it so we have 98% FC > SS on the leaderboard or +1 droplet, -20 bananas giving more score.
Riari

PakaChan wrote:

Riari wrote:

What do people think about scaling banana values similar to small droplet values?

Inside osu!catch spinning is not a mandatory action, the score is a bonus on top of what you get for playing the rest of the map.
Currently in ScoreV2, droplets scale up to 10% of a maps value and would naturally diminish their value in longer maps. Bananas on the other hand have a constant score value of 350 giving them a significant edge in the longer run.

I think that whilst high-end spinner plays are impressive, droplets are a core part of osu!catch scoring and should not be so heavily reduced in comparison to bananas.
I think that a score cap that is similar to what small droplets face right now for bananas would be helpful for competitive scores as it would allow accuracy to make a bigger impact on maps that contain spinners.



I also saw something over in the standard discussion thread where score could possibly be multiplied by your accuracy. I think this would work much better over here and give misses and droplets a much more competitive edge.
You could also consider giving droplets a static score like how it's now except make it so it actually matters when compared to that of a banana. Right now 1 banana = 110 droplets, you could always just change things up to make it 1 droplet = 1 banana or 1 droplet = 3 bananas or 3 droplets = 2 bananas. Giving both a multiplier sounds messy and giving just one a multiplier makes it so we have 98% FC > SS on the leaderboard or +1 droplet, -20 bananas giving more score.
All modes seem to be going towards a capped score of 1,000,000 (Spinners as bonus) hence why the droplets can make up to 10% of that million and no more. It's more about limiting the values of bananas so they do not outstrip acc and make high acc slightly lower spins competitive with lower acc slightly higher spins.
Amlink
DT 1.12 FL 1.12 HR 1.06 HD 1.06 simple as that
RAMPAGE88
DT>HR
Bunnrei
i think it should be dt > hr > hd or something (1.12x > (something inbetween) > 1.06x)

since in most cases dt is harder (getting used to higher platter speeds, chance of getting beyond ar10) compared to hr (it does make the map harder by increasing map stats and messing up the notes but it's not as fast-paced and straining as dt).

though in some cases hr ends up being harder than dt (like on cs7+ maps) but that's uncommon

and hd is a commonly used mod and is the easiest to learn in comparison to the other difficulty-increasing mods (imo) so it should stay as it is

(i didnt read most posts so idk if i just repeated what others said)
rostld
where is ctb spunout
Bunnrei

Jimmy Rustler wrote:

where is ctb spunout
tru
MBomb
I think people are heavily oversimplifying stuff in a way which is kinda inaccurate. At AR8.5+, DT will be harder than HR most of the time, I agree, but in cases where high CS and low AR are used (Or even in cases where it's just high CS), HR can end up being a lot harder, especially to get accuracy on, which would be a lot bigger part in this scoring system. However, a lot of people commenting on this thread are high end players, so these maps end up being irrelevant to them, which is why these points are being ignored, as high end players play high CS maps a lot less, for the most part.
He Ang CTB
The current scores system is alright :O There are few short-comings but the new score system will also have short-coming but from the opposite end.
[Kami]
You sould add this condition to the score

if AR>=9 then DT=1.13x else DT=1.06x

DT should be worth more than HR if the AR is really high like 10.3.
nya10
I'm not really related to this, but a multiplier of a mod without a FreeMod bracket in the tournament (assuming this will be used only for o!CWC, or similiarly condition o!CWC tournament), the multiplier won't be do much, since you'll only be able to use 1 mod per bracket (+HD in DT and HR bracket), and I don't think in this case, HR and DT could be comparable. So the only thing to do is to adjust HD (judging from some people that has a better reading in HD than NoMod) into the correct multiplier without really debating how HR and DT compares.
DeletedUser_500696

nya10 wrote:

I'm not really related to this, but a multiplier of a mod without a FreeMod bracket in the tournament (assuming this will be used only for o!CWC, or similiarly condition o!CWC tournament), the multiplier won't be do much, since you'll only be able to use 1 mod per bracket (+HD in DT and HR bracket), and I don't think in this case, HR and DT could be comparable. So the only thing to do is to adjust HD (judging from some people that has a better reading in HD than NoMod) into the correct multiplier without really debating how HR and DT compares.
This is exactly what I was thinking, you can't use both DT and HR in the same pool so there is no reason to adjust those values. The only points of concern should be how much weighting HD gives with HR and DT. Personally I think HD shouldn't even be allowed in the DT and HR pools since it makes it easier for some people (like me)
iiyo
give easy mod 1.06
Dianthus
Talking about tournament settings here.

1. Make HD give 0.00x on non-HD maps on the tournament. Probably the easiest way would be to add Tournament HD (naming sense=0) as the second HD pick (just as NC is 2nd DT), THD gives 0.00x.
2. HD picks should be maps that actually get harder with it.
3. Bananas should have set value, as it currently is. This way banana gods will be rewarded as well.

FL: 1.12x
DT: 1.12x
HR: 1.06x
HD: 1.06x
iiyo
Krah

-Kurisu- wrote:

This is exactly what I was thinking, you can't use both DT and HR in the same pool so there is no reason to adjust those values.
Quite true but scoring should be directly valid in all possible tournaments so having DT and HR at the same time could be a thing cuz idk reasons or troll pool or reasons or just a dt pool where hr could be allowed in freemode. Not saying that it will happen in the cwc but it could happen in another random tournament that would like to use the v2 scoring.
I still think that dt should increase with ar and hd decrease and not just trying to randomly sort them and thinking there is only hard maps and the cwc with its ar8+ meta. Maybe something like :
DT : 1.06+(0.5*AR)
HD : 1.08-(0.5*AR)
With the AR being the one without any mods to avoid some useless complications.



Hedgeturtle wrote:

1. Make HD give 0.00x on non-HD maps on the tournament. Probably the easiest way would be to add Tournament HD (naming sense=0) as the second HD pick (just as NC is 2nd DT), THD gives 0.00x.
I actually asked for that the day the 2015 cwc rules were released and it wasn't possible at this time to simply calculate it. Hope it could be a thing now. (btw THD should give 1.00* not 0.00*)
Jasmine

PakaChan wrote:

DT 1.14x
FL 1.1x
HR 1.06x
HD 1.03x

DT > FLHD > FL > HRHD >HR > HD
I don't think it should be any other way as long as the multipliers are static and don't care about CS and AR. Maybe 1.12 DT if you want FLHD > DT
i'm not agree with you for DT=FL+HD, i thinks FLHD is a LOT more harder than DT only

I thinks score multipliers need to care about CS and AR because for example HD on low AR (like AR6); AR6 HD is really more harder than AR6 HR/DT
(Don't forget EZ+HD is ONLY 0.53x)

And about EZ multipliers i thinks 1.00x is good because the map is not more easy, ofc AR and the CS is reduce with EZ but the most of time it's a lot MORE harder than no mods. ( EZ+DT 0.53x hahaha )
Laharl

Fantasy wrote:

give easy mod 1.06
Oh, come on.
iiyo

Laharl wrote:

Oh, come on.
sorry for the spook
Laharl

Fantasy wrote:

Laharl wrote:

Oh, come on.
sorry for the spook
I hope so. Almost had a heart attack when I read EZ > NM when it defies all sorts of logic.
iiyo
Heard from a little birdie that these changes will be on tournaments only and not on the actual client, if that's the case then CTB truly is dead because HR > DT makes literally no sense half of the people who are deadlocked into HR scores cannot DT the overdoses they HR'd, which makes a lot of people lose motivation and not really care to play.
Zak
ScoreV2 is apparently wanted to be on public someday but that also requires wiping every score set ever so there's reasons why it's not going in anytime soon.
I'm pretty sure it wouldn't even be as simple as just recalculating every play because it just isn't that simple especially on any scores which have broken replays or are not top 250, also stop saying CTB is dead it's a stupid meme and it's getting old
DeletedUser_500696
dt still gives way more pp than hr lol. It's obvious that dt is more difficult than hr, that's part of the reason why less people can dt overdoses. Fl is too but people don't really do that on ar9+ with the exception of a few, because people know it's difficult af. I do think more people including me would like to be rewarded for playing those mods by seeing their hard earned scores at the top of the scoreboard. But ctb definitely isn't dead just because those scores aren't on the scoreboard, there are people out there that train dt and fl. Also the important thing isn't whether it'll be easy to make the change or not, leave it to the underestimated coders. It's how much the community wants to see the change.

tl;dr I agree with parts of fantasy and zak's logic and disagree with other parts.

Also @Krah ty for clarifying that, I like both of those ideas. Ar needs to be taken into account for dt and hd ye, i think fl too. Hr is fine without any adjustment imo
Allrounder
FL = 1.15x
DT = 1.12x
HR = 1.08x
HD = 1.05x

That's how I would do it. Cause you need to know the whole map for FL | DT'ing a 4* map for example is harder than HR'ing it | HR'ing is harder than HD'ing.
But the main point is, that DT should have a higher multiplier than HR, because otherwise you'll have problems if you HR'd a lot of maps and later you are able to DT these maps.

For example: You SS'd a 4* map with HDHR. One year later or something you are a better player and you are theoretically able to DT this map. If you SS it with HDDT you won't be able to beat your current score even though it's harder (in most cases) and you would gain more pp. (Or at least create something, which always gives you the highest pp out of your scores, doesn't matter if you beat your current score or not)


Ofc that's just my opinion. I can agree if somebody says HD is harder than HR for example, but still DT should have a higher multiplier than HR just to "balance" the pp system.
Zak

Allrounder wrote:

Ofc that's just my opinion. I can agree if somebody says HD is harder than HR for example, but still DT should have a higher multiplier than HR just to "balance" the pp system.
The first post of the thread says not to discuss pp here as ScoreV2 is strictly for CWC.

Also a thought that just came to me is that I don't really see much need to weigh DT vs HR for a tournament as you're never able to see DT vs HR in a tournament setting, even in a free mod pool, it's impossible to have DT on unless it's on for everyone, so there's no need to worry about HR giving more score.
LegoTonix
FL = 1.14x
DT AR<8 = 1.08x DT AR>9 = 1.12x
HR = 1.10x
HD = 1.03x AR>9 1.08x AR<7

I think DT worth more HR is bad on easy maps <2.7* or under AR 8.5
1 because if you check scoreboreder on AR 7-8 maps 2 * if you check you will see in most case 15 DTHD score, and 2-3 HRHD score, on maps under 2.7* HR is ways harder DT isn't so hard on this maps
2 CS 5+ is very very hard in HR if you look CS 5+ AR 7-8 have only few score in HR or even no score at all i see a lot maps in CS 5+ have 0 score in HR
and for CS 6-7 is HR is almost impossble unless be a god like exgon or other pro player
3 this will broken score for AR 7-8 maps this will be spam DTHD spins to win for get #1 ranking and HRHD will worthless and get bad rank, even if is way harder

but i'm not saying DT is easier than HR i think DT is much way harder in AR 9+ maps
if you look scoreboard in AR 9 maps you will see a lot more HRHD than DT or DTHD even on 0 DT on almost AR 9 maps so yeah DT worth more than HR on AR 9 is a good idea but for CS5/AR8 maps HR is ways harder often because pixel jump and convert maps.

EDIT: I read this is only for CWC so yeah, scoreV1 is fine for CWC because i dont think will have game like DT vs HR, maybe just update HD to x1,03
I don't think FL will be playable at CWC. also for HRDTFL no one will chose that unless on very easy maps under AR 8 because 1 FL is very easy to miss in frist try so unplayable for cwc and DT will are not playable for AR 9 at frist try i think 99 % player will chose nomod and FC instead DT do 15 miss and HR is playable but is harder than nomod but eaiser than DT i dont think will have DT vs HR in AR 9 in cwc
Digitalfear117

LegoTonix wrote:

I think DT worth more HR is bad on easy maps <2.7* or under AR 8.5
1 because if you check scoreboreder on AR 7-8 maps 2 * if you check you will see in most case 15 DTHD score, and 2-3 HRHD score, on maps under 2.7* HR is ways harder DT isn't so hard on this maps
2 CS 5+ is very very hard in HR if you look CS 5+ AR 7-8 have only few score in HR or even no score at all i see a lot maps in CS 5+ have 0 score in HR
and for CS 6-7 is HR is almost impossble unless be a god like exgon or other pro player
3 this will broken score for AR 7-8 maps this will be spam DTHD spins to win for get #1 ranking and HRHD will worthless and get bad rank, even if is way harder
I personally feel like this won't actually solve the problem of what is harder, this is just a blanket for the problem, because there are plenty of difficult AR8 DT maps, but I do agree DT is not always more difficult than HR, and you also mention this. I (still) think it should be determined by the star difficulty after the mods are applied

So for example, lets say a map is 2 stars.

You apply HR and it increases to 2.54 stars
You apply DT and it increases to 2.86 stars

DT should in this case be given a higher score multiplier than HR, but if HR were to increase the star rating more it should have the higher multiplier.

Less plays with certain mods doesn't necessarily make a map harder, what if less people just bothered to FC it with that mod combination? I see where your coming with that, but for example take this map here, https://osu.ppy.sh/s/130504 in standard this is considered an easy DT map, with many DT fc's, but there is only one HDHR score even submitted, that doesn't mean its harder with HR, but just that no one wanted to do it with those mods. I am aware standard and CtB are completely different, but its just an example (and i couldn't think of a CtB map off the top of my head lol)

Zak wrote:

The first post of the thread says not to discuss pp here as ScoreV2 is strictly for CWC.
I personally interpreted this being meant in regards to suggestions for changes in the pp system, I could be wrong (and if i am i will edit this post) but just talking about the system as it is should be allowed.

Zak wrote:

Also a thought that just came to me is that I don't really see much need to weigh DT vs HR for a tournament as you're never able to see DT vs HR in a tournament setting, even in a free mod pool, it's impossible to have DT on unless it's on for everyone, so there's no need to worry about HR giving more score.
True... but I think they eventually want to make scorev2 roll out to more than just tournaments, so I think we should come up with a score system that will be better for both.
2zz
If look objectively, the maps with DT is no more difficult than nomod-maps. Its are simply 1.5 times faster, nothing more. The patterns remain the same, their location too. Yes, to do anything half faster is very difficult, if you can't even on normal speed. But in general, DT is not "hard" mod. This is a "fast" mod. However, HR is "difficult" mode, cause the position of the patterns making them little bit harder to pass. But the speed of the taps remains the same. Yes, AR is increasing, but so what? If change AR you'll see that the only thing that has changed is the distance between the notes. If you do the same with DT nothing will change because the speed of taps on the notes will remain accelerated in 1,5 times, isn't it?

Replacing of the multiplier from the DT to HR and back on different difficulties to balance would be technically difficult, and it is not necessary for the CWC so that would be to equate it.

FL - 1.12
HR - 1.12
DT - 1.12
HD - 1.04
PakaChan
Holy shit the last guy is just too much.
So many people that have no idea what they're talking about...
RAMPAGE88
Yeah too much but it's funny.
Also, HR doesn't change location so much in CtB maps tbh, It affects convert maps more and most of the time old converts (old mapping)
I didn't say that HR is easy but keep in mind that HR sometimes also changes pixel jump to hyper and make some jumps (pattern) easier to catch.
So sometimes it helps in some patterns but never for DT (I think)
Well It's just an opinion.
iiyo

2zz wrote:

FL - 1.12
HR - 1.12
DT - 1.12
HD - 1.04

i hope ur joking
rostld
how about everything 1.1
Secre

Jimmy Rustler wrote:

how about everything 1.1
no 1.0
show more
Please sign in to reply.

New reply