All underrated because of the 1key sections, not the sliders afaik.uberpancake wrote:
Maybe maps like scary rose, big black, adult's toy and talent shredder? I'm sure there are better examples thoughjesus1412 wrote:
A lot of people say this, got any example maps?
Community
Forum
Performance Points feedback and suggestions (Standard)
posted
Total Posts
2,647
of those examples, the only ones with streams are big black and adult's toy and they are extremely easy streams, so the fact that most of the maps difficulty comes from 1key parts is a big factor, i think jesse is right
well except maybe towards the end of bb, but id say that map is still 6.5 stars so..
well except maybe towards the end of bb, but id say that map is still 6.5 stars so..
Re-posting my original suggestion because people keep seeing the UR thing Riince talk about after this.
Woobowiz wrote:
This may be a stupid suggestion, but is it within reason to "buff" lower OD by weighing the accuracy values of 100's and 50's?
So for OD 6, 100's would be worth 2/5 or 1/2 of a 300 rather than the usual 1/3rd of a 300. Low OD is weak enough for giving less pp even for an SS, why not give it a little push
that makes no sense since it basically rewards poor performance (low OD low acc) and would make low OD high acc further underrated.
No one replied because it was a stupid idea, don't just keep posting what you said until someone acknowledges it.
As for examples, I think you mentioned Neuroncia as one, this would be another example. But I think it's already pretty evident that sliders are underrated, I don't think there is one slider map that gives good pp, every good pp map is almost all circles. They don't give anywhere enough pp for the extra aim and reading you have to do, and of course they give no accuracy pp which is a big deal.
What do you mean by '1key section'? I think it's safe to say that those maps would be a lot easier if the sliders were circles, I actually tried changing some sliders to circles on Scarlet Rose and the star rating actually increased.jesus1412 wrote:
All underrated because of the 1key sections, not the sliders afaik.uberpancake wrote:
Maybe maps like scary rose, big black, adult's toy and talent shredder? I'm sure there are better examples though
As for examples, I think you mentioned Neuroncia as one, this would be another example. But I think it's already pretty evident that sliders are underrated, I don't think there is one slider map that gives good pp, every good pp map is almost all circles. They don't give anywhere enough pp for the extra aim and reading you have to do, and of course they give no accuracy pp which is a big deal.
1key sections as in... the entirety of the maps (if you 1key). Very spaced slow streams if that's how you prefer to see them.B1rd wrote:
No one replied because it was a stupid idea, don't just keep posting what you said until someone acknowledges it.What do you mean by '1key section'? I think it's safe to say that those maps would be a lot easier if the sliders were circles, I actually tried changing some sliders to circles on Scarlet Rose and the star rating actually increased.jesus1412 wrote:
All underrated because of the 1key sections, not the sliders afaik.
As for examples, I think you mentioned Neuroncia as one, this would be another example. But I think it's already pretty evident that sliders are underrated, I don't think there is one slider map that gives good pp, every good pp map is almost all circles. They don't give anywhere enough pp for the extra aim and reading you have to do, and of course they give no accuracy pp which is a big deal.
There are only two ways that I can figure this out, one possibility is that jump distance is taken from the ends of slider, in this case, when a player doesn't follow a short slider (as in, they treat it like a normal hit circle) the effective jump distance is calculated with the wrong assumptions. The player presses the slider start and then moves to the next note/slider as if it was a hitcircle but the star algorithm assumes that they follow the slider to the end then treats the jump as the shorter distance between the slider end and the next note.
The other possibility that could possibly mitigate the above scenario is that the jump distance is taken from the point furthest from the end of a slider that would yield a 300. In this scenario the previously mentioned scenario can still occur but it should be dampened unless the sliders get especially long. This solution however has it's own issues, shorter sliders will be assumed to work as standard hitcircles.
The one thing I can't understand is how long slider maps are underrated, in my eyes it just doesn't make any sense and I'd have to guess that it's underrated for other reasons.
I'll admit this is an issue and that it's probably one of the more fixable ones at this point in time (compared to heavy 1key sections/slow spaced streams).
to give you a better example of underrated sliders: https://osu.ppy.sh/s/128645 or https://osu.ppy.sh/s/102282
i personally think pp doesnt look enough at technical difficulty, it's pretty much only physical ability. https://osu.ppy.sh/s/290040 are easily 5+ stars in terms of difficulty but only 4.32 stars because it's a slow map with pretty much no spacing
i personally think pp doesnt look enough at technical difficulty, it's pretty much only physical ability. https://osu.ppy.sh/s/290040 are easily 5+ stars in terms of difficulty but only 4.32 stars because it's a slow map with pretty much no spacing
lol I didn't know Terminal was so underratedsilmarilen wrote:
to give you a better example of underrated sliders: https://osu.ppy.sh/s/128645 or https://osu.ppy.sh/s/102282
i personally think pp doesnt look enough at technical difficulty, it's pretty much only physical ability. https://osu.ppy.sh/s/290040 are easily 5+ stars in terms of difficulty but only 4.32 stars because it's a slow map with pretty much no spacing
Not sure if this has been suggested before.
Maps could analyzed for difficulty on a combo timeline to determine what combo guarantees you fcd the hard parts. Find the lowest difficulty section, that x combo passes for some amount of combos (25%, 50%, 75%, 90%, 100% combo for example).
You could make a simple table for every map to weight the pp/combo% they give and buff nonfcs on maps with hard parts in the middle, lower the pp nonfcs in maps with hard parts in the beginning/end give.
Maps could analyzed for difficulty on a combo timeline to determine what combo guarantees you fcd the hard parts. Find the lowest difficulty section, that x combo passes for some amount of combos (25%, 50%, 75%, 90%, 100% combo for example).
You could make a simple table for every map to weight the pp/combo% they give and buff nonfcs on maps with hard parts in the middle, lower the pp nonfcs in maps with hard parts in the beginning/end give.
Suggested but I still agree with the idea.jaaakb wrote:
Not sure if this has been suggested before.
Maps could analyzed for difficulty on a combo timeline to determine what combo guarantees you fcd the hard parts. Find the lowest difficulty section, that x combo passes for some amount of combos (25%, 50%, 75%, 90%, 100% combo for example).
You could make a simple table for every map to weight the pp/combo% they give and buff nonfcs on maps with hard parts in the middle, lower the pp nonfcs in maps with hard parts in the beginning/end give.
factoring in UR doesnt sound reasonable
first of all you'd have to replace it with average aberration and then again:
If you can get a bonus to a 300 why wouldn't you give extra pp for hitting circles closer to the middle?
And if there's already a function to calculate boni why wouldnt we replace every circle with a target like this?
http://www.clipartbest.com/cliparts/pc5 ... e9pkxi.png
would kill any DT-players with non-perfect aim and move the focus to smaller circle sizes when ppl git gud at it
slight irony because it would make the game weird imo but didn't want to keep it for myself
first of all you'd have to replace it with average aberration and then again:
If you can get a bonus to a 300 why wouldn't you give extra pp for hitting circles closer to the middle?
And if there's already a function to calculate boni why wouldnt we replace every circle with a target like this?
http://www.clipartbest.com/cliparts/pc5 ... e9pkxi.png
would kill any DT-players with non-perfect aim and move the focus to smaller circle sizes when ppl git gud at it
slight irony because it would make the game weird imo but didn't want to keep it for myself
UR isnt about where you hit the notes, it's about how accurate you hit them. it's a better version of accuracy
I think he knows that, but it measures your absolute accuracy, whereas hitting notes somewhere within the hitarea does not, only whether it's a hit or not based on CS, while this aim aspect could also be measured in an absolute way (regardless of CS, just like UR doesn't care for OD) if we wanted to go down that direction: hit distance from 1x1 centre of hitcircle, averaged similarly to unstable rate (and similarly to hit timing, more than X aim error = miss, where X depends on CS/OD).
It seems fair just like basing acc on UR, but I don't think it'd "feel good" to play like this (even though it'd be good to have mods that allow players to push their "aim accuracy" too by allowing CS7 and stuff like that on ranked maps). Probably measuring absolute error instead of categories wouldn't feel that harsh when applied to accuracy, maybe because that's already divided into 4 outcomes vs hit/miss of aim and we already have OD10-11 that require you to be rather precise with your acc, while CS7< is pretty rare.
It seems fair just like basing acc on UR, but I don't think it'd "feel good" to play like this (even though it'd be good to have mods that allow players to push their "aim accuracy" too by allowing CS7 and stuff like that on ranked maps). Probably measuring absolute error instead of categories wouldn't feel that harsh when applied to accuracy, maybe because that's already divided into 4 outcomes vs hit/miss of aim and we already have OD10-11 that require you to be rather precise with your acc, while CS7< is pretty rare.
UR weighting would be really unfair for those who have cheap setups, I don't think it's a good idea.
Introduce rainbow 300's to standard that give acc weightings of 150%.
huaehaueahu
huaehaueahu
^Drezi wrote:
I think he knows that
Additionally UR is defined as:
Which the part with "average aberration" referred to as UR doesnt measure accuracy as mentioned in the definition.osu! Wiki wrote:
This value represents how consistently you time your hits, with lower numbers being better (top players often score below 150). Note that this measures consistency, and not accuracy, so if you're consistent in hitting 15ms early, you'll get similar results to if you're consistent in hitting on time. The formula is essentially the standard deviation of your hit errors (in milliseconds) multiplied by 10.
You'd need a different value that calculates the deviation to the perfect timing in a sensible way.
The part with the target thing was meant as a fitting analogy that shows how awkward it would feel if something like weighted accuracy outside of OD would be added to the pp-calculation. Because what it essentially would be is that you're able to modify the OD of the map at will(to exactly what you're capable of) and the map still counting for your pp afterwards.
Coming from that, you could claim that possibility for any of the metadata: O look, I can SS this map at Hp4 and Hp9, gimme more pp for the latter.
and
My personal UR went down by around 30 INSTANTLY just because i switched to good input-devices before I even got used to them. When cheap setups already have a harder time it'd be unfair to give them less reward for something they already have to put more effort in.Default wrote:
UR weighting would be really unfair for those who have cheap setups, I don't think it's a good idea.
PS: I feel seriously trolled by scoring system:
https://osu.ppy.sh/s/13768 this map has insanely hard sliders too and gives ~180pp (WWW is the only one who has an S on it) I'm unsure if it's possible to fix it with the current system though.
pp: [Tom94] Reduce value of the Easy mod to counterbalance difficulty calculation bugfix.
Fix fast sliders and fast streams not worth anything atm
This has probably already been mentioned (and is probably a bad idea for one reason or another), but oh well, prepare for opinions.
I think it's a bit odd that max combo% has such a large impact on the amount of PP that a non-FC play is worth; I have had multiple cases where a play with multiple misses at the end of the song is worth considerably more PP than a play with one miss in the middle of the song.
In my opinion, the number of misses should have a bigger impact on how much PP a non-FC play is worth, and combo% should have a much smaller impact (or no impact at all). To account for slider breaks, perhaps there could be a PP bonus for getting a full combo (or a penalty for not getting a full combo).
I think it's a bit odd that max combo% has such a large impact on the amount of PP that a non-FC play is worth; I have had multiple cases where a play with multiple misses at the end of the song is worth considerably more PP than a play with one miss in the middle of the song.
In my opinion, the number of misses should have a bigger impact on how much PP a non-FC play is worth, and combo% should have a much smaller impact (or no impact at all). To account for slider breaks, perhaps there could be a PP bonus for getting a full combo (or a penalty for not getting a full combo).
I think this is a good idea too.jaaakb wrote:
Not sure if this has been suggested before.
Maps could analyzed for difficulty on a combo timeline to determine what combo guarantees you fcd the hard parts. Find the lowest difficulty section, that x combo passes for some amount of combos (25%, 50%, 75%, 90%, 100% combo for example).
You could make a simple table for every map to weight the pp/combo% they give and buff nonfcs on maps with hard parts in the middle, lower the pp nonfcs in maps with hard parts in the beginning/end give.