Purpose of thread:

The purpose of this thread is to facilitate discussion about note locking so that a viable solution can be agreed upon and presented to Peppy as a possible implementation.

Guideline:

Posts that comment on possible solutions, give new solution, or point out errors are allowed.

Think of these factors when posting (not exclusive):

1. does the solution abide with the intention of the mapper?
2. does the solution provide player appropriate feedback?
3. can the solution be taken advantage of?
4. what are the possible effects of solution on performance point weighting?

About note locking:

Right now if each note was 60 ms away and the hit window was 140 ms wide, then each note has an overlapping window of 45 ms. If you do not hit the 1st note, then the 2nd note would only have 80 ms hit window. Even if the 2nd note is hit at exact time, subsequent notes are not considered to be hit, resulting in a series of misses.

As such note locking penalizes the player with a possible fail of the whole map when only one note has been missed. Note locking also prevents feedback given to the player. When a series of notes are missed due to note locking, the player cannot determine which of the notes have been pressed or missed.

Current list of solutions:

Solution 1

Get rid of overlaps, and make score 300 window relatively smaller.

The problem with this solution is that note density of the map can affect actual OD.
Even if the map had OD10, notes that use this solution would have a smaller score 300 hit window. Mappers may take advantage of this to artificially inflate OD leading to higher pp maps. This also means streams will be more weighted than before.

Solution 2

Get rid of overlaps, and make score 300 window absolute value. For example, OD10 score 300 window is 39 ms. This will stay the same even with the solution implemented. The score 50 window or score 100 window will be cut off.

The problem with this solution is that it increases proportion of 300 window. While OD is not affected, streams are less likely to have 100s and 50s than before. This may make streams more underweighted than before.

Solution 3

If you have note 1, 2, and 3, hitting note 2 before note 1 is valid (not miss or combobreak). However, hitting the notes in wrong order creates an accuracy penalty that allows score 100 at maximum.

The problem with this solution is the possibility that a player can take advantage of this mechanism to make the map easier. On a map that has jumps back and forth, the intention is to make the cursor go back and forth. Instead, the player can click notes two times on each side having to move the cursor only once (like hitting two doubles).

Solution 4

(from Endaris. refer to p/5434427 for full explanation)

Make hit window for each OD smaller in general. Hit window for 100 and 50 is very lenient, decreasing very little compared to hit window for 300. Making hit window for 100 and 50 smaller on high OD would reduce note locking considerably.

One problem with this solution is that eventually note locking becomes an issue again at higher BPM. Reducing 50s area from current 100 ms to 70 ms would eliminate note locking of streams up to 190 BPM. However, higher BPM streams would be affected. It is likely that this solution will be presented in conjunction with one of the other solutions.

Solution 5

(from Fleimi)

Notelocking only occurs from combo to combo. Example scenario:

Stream = 1 2 3 4 1 2 3 4

You can hit notes in both combos in whichever order you prefer eg. 2 1 3 4 and 3 2 1 4, provided that the hitwindows of the notes are active. However, you can't hit a note in a new combo until you have cleared the previous combo or their hitwindows have expired, so an order of 1 2 3 1 4 2 3 4 isn't allowed. This way combos have a gameplay element as well. Normal accuracy rules still apply. Given the way streams are usually mapped, with 4, 8 or 16 note combos, this gives plenty of occasions to recover even if you misaim a stream.

Solution 6

(from Full Tablet)

1. When the player hits a key within the timing window (including the "early miss" timing window) of the circle, but the cursor is not aiming the current circle, instead of making it as if the circle wasn't interacted with at all, make it instantly miss, unlocking the next note.
2. A variant of this would be adding a "misaim miss" area surrounding circles that gives you a miss in the circle regardless of timing as long as you hit them in their timing window. This area should be big enough to include the distance between circles in a jumpstream.

Most of the time, getting a miss in the middle of a stream is caused because of a misaim rather than not pressing the note at all in time. This change would make note locking not be so much on an issue in this case.

In the case where a note is missed because the player completely ignored hitting the key, and the timing window of the ignored note overlaps the perfect timing time of the next note, the next note would be registered as being hit late (which is not a problem, since that occurring means the player effectively ignored the first note and then pressed the next note late).

Mistiming can be solved using solution 1, 2, or 3.

I didn't see that post. It's a very informative post and I will put that as solution #4: make hit window for each OD smaller. However, that thread is for score v2 in general while this thread is just for note locking. It is good to have a separate thread for this so that people can focus on note locking.

I'll chip in with something more from the same stream:

Solution:

Notelocking only occurs from combo to combo. Example scenario:

Stream = 1 2 3 4 1 2 3 4

You can hit notes in both combos in whichever order you prefer eg. 2 1 3 4 and 3 2 1 4, provided that the hitwindows of the notes are active. However, you can't hit a note in a new combo until you have cleared the previous combo or their hitwindows have expired, so an order of 1 2 3 1 4 2 3 4 isn't allowed. This way combos have a gameplay element as well. Given the way streams are usually mapped, with 4, 8 or 16 note combos, this gives plenty of occasions to recover even if you misaim a stream.

That wouldn't work because in many cases new combos are used in conjunction with streamjumps and if you jumped too early to the 1 you would miss both the 4 AND the 1 because the previous combo is still active, so you'd still get overly punished.
And NCs have their use in terms of making a map readable so changing the way you put NCs is not an option.

It still reduces the occurrence of fails because of notelock. I'm comfortable with missing streamjumps in the way you described punishing the player - after all, they're meant to be a difficulty element. I am aware though that with 1/3 snapped notes that are often mapped as a series of consecutive new combos (a stream of 1s) this would still make for an issue. However, this still reduces the amount of punish by at least an order of magnitude if we look at the map as a whole.

I also don't think this change would require a change in how new combos are mapped.

Fleimi wrote:

I'll chip in with something more from the same stream:

Solution:

Notelocking only occurs from combo to combo. Example scenario:

Stream = 1 2 3 4 1 2 3 4

You can hit notes in both combos in whichever order you prefer eg. 2 1 3 4 and 3 2 1 4, provided that the hitwindows of the notes are active. However, you can't hit a note in a new combo until you have cleared the previous combo or their hitwindows have expired, so an order of 1 2 3 1 4 2 3 4 isn't allowed. This way combos have a gameplay element as well. Given the way streams are usually mapped, with 4, 8 or 16 note combos, this gives plenty of occasions to recover even if you misaim a stream.

Interesting idea. I didn't fully understand what Rohulk was saying in stream. One problem I immediately see though is that some players may take advantage of this (same problem as solution 3). Also, could you clarify how note locking doesn't occur within the combo? I'm guessing players still need to hit the notes on time, which means solution 1 or 2 should be considered. Finally, what is the point of giving note lock to end and start of NC if either solution 1 or 2 is implemented?

Endaris wrote:

That wouldn't work because in many cases new combos are used in conjunction with streamjumps and if you jumped too early to the 1 you would miss both the 4 AND the 1 because the previous combo is still active, so you'd still get overly punished.
And NCs have their use in terms of making a map readable so changing the way you put NCs is not an option.

Under the solution Fleimi brought, jumping too early to 1 would cause you to miss both note 4 and note 1. However, that is the case right now as well since note locking is present for every note. Correct me if I'm not understanding you.

I agree with you that NC has their role as well. It might be too complicated to give another role to NC.

Fleimi wrote:

It still reduces the occurrence of fails because of notelock. I'm comfortable with missing streamjumps in the way you described punishing the player - after all, they're meant to be a difficulty element. I am aware though that with 1/3 snapped notes that are often mapped as a series of consecutive new combos (a stream of 1s) this would still make for an issue. However, this still reduces the amount of punish by at least an order of magnitude if we look at the map as a whole.

I also don't think this change would require a change in how new combos are mapped.

However, 1/3 notes have less issue with note locking anyways. I believe mapping style will not be affected much for 1/3 notes.

Circle wrote:

Interesting idea. I didn't fully understand what Rohulk was saying in stream. One problem I immediately see though is that some players may take advantage of this (same problem as solution 3). Also, could you clarify how note locking doesn't occur within the combo?

As in, notelocking just doesn't exist. At all. Every active note is free game, apart from the transitions between combos. Of course normal accuracy rules apply so it wouldn't really be feasible to play the notes in that order, you'll just get a bunch of 50s.

I think we also need to leave out the player abuse factor here. If notelocking is removed in any way, of course there's a possibility for abuse. But that abuse comes at a price of accuracy (and if the hitwindows get reduced as per Endaris's suggestion coupled with a low enough BPM, it might not even be possible to play the notes in a wrong order). In any case, if something like this is implemented, it's feasible to first have a test period to possibly revert the changes if it ends up being something game breaking.

What about this change?

1. When the player hits a key within the timing window (including the "early miss" timing window) of the circle, but the cursor is not aiming the current circle, instead of making it as if the circle wasn't interacted with at all, make it instantly miss, unlocking the next note.
2. A variant of this would be adding a "misaim miss" area surrounding circles that gives you a miss in the circle regardless of timing as long as you hit them in their timing window. This area should be big enough to include the distance between circles in a jumpstream.

Most of the time, getting a miss in the middle of a stream is caused because of a misaim rather than not pressing the note at all in time. This change would make note locking not be so much on an issue in this case.

In the case where a note is missed because the player completely ignored hitting the key, and the timing window of the ignored note overlaps the perfect timing time of the next note, the next note would be registered as being hit late (which is not a problem, since that occurring means the player effectively ignored the first note and then pressed the next note late).

The case where the player completely ignored the first circle, but then hit the following notes correctly is more problematic. The first note would be considered as being quite late (a 50 in most cases), and every note afterwards would be considered as quite late too, if the player hits them in time. A possible solution for this is making the timing window for late 50s be considerably smaller for high OD. In osu!mania, there is no timing window for late 50s at all.

I think the case where a note is missed, not because the player ignored the note, but because they hit the key way too late, is quite rare; this case would most likely make the player fail or get a bad performance in their play, since this means the player can't keep up with the notes in the map.

Fleimi wrote:

Circle wrote:

Interesting idea. I didn't fully understand what Rohulk was saying in stream. One problem I immediately see though is that some players may take advantage of this (same problem as solution 3). Also, could you clarify how note locking doesn't occur within the combo?

As in, notelocking just doesn't exist. At all. Every active note is free game, apart from the transitions between combos. Of course normal accuracy rules apply so it wouldn't really be feasible to play the notes in that order, you'll just get a bunch of 50s.

I think we also need to leave out the player abuse factor here. If notelocking is removed in any way, of course there's a possibility for abuse. But that abuse comes at a price of accuracy (and if the hitwindows get reduced as per Endaris's suggestion coupled with a low enough BPM, it might not even be possible to play the notes in a wrong order). In any case, if something like this is implemented, it's feasible to first have a test period to possibly revert the changes if it ends up being something game breaking.

Sounds good. I'll edit the original post.

Full Tablet wrote:

What about this change?

1. When the player hits a key within the timing window (including the "early miss" timing window) of the circle, but the cursor is not aiming the current circle, instead of making it as if the circle wasn't interacted with at all, make it instantly miss, unlocking the next note.
2. A variant of this would be adding a "misaim miss" area surrounding circles that gives you a miss in the circle regardless of timing as long as you hit them in their timing window. This area should be big enough to include the distance between circles in a jumpstream.

Most of the time, getting a miss in the middle of a stream is caused because of a misaim rather than not pressing the note at all in time. This change would make note locking not be so much on an issue in this case.

In the case where a note is missed because the player completely ignored hitting the key, and the timing window of the ignored note overlaps the perfect timing time of the next note, the next note would be registered as being hit late (which is not a problem, since that occurring means the player effectively ignored the first note and then pressed the next note late).

The case where the player completely ignored the first circle, but then hit the following notes correctly is more problematic. The first note would be considered as being quite late (a 50 in most cases), and every note afterwards would be considered as quite late too, if the player hits them in time. A possible solution for this is making the timing window for late 50s be considerably smaller for high OD. In osu!mania, there is no timing window for late 50s at all.

I think the case where a note is missed, not because the player ignored the note, but because they hit the key way too late, is quite rare; this case would most likely make the player fail or get a bad performance in their play, since this means the player can't keep up with the notes in the map.

I can see where you are coming from. Missing a circle can be because of misaiming or mistiming. If we are talking about misaim your idea would get rid of misaimed note affecting hitwindow of the next ones. This still doesn't consider missing because of mistiming. Implementing this idea along with solution 1 or 2 or 3 would work to cover both. This cannot be done with solution 5 (all notes available within a combo) since whether the cursor was aimed wrong or intended for another note would be hard to determine.

Full Tablet wrote:

What about this change?

1. When the player hits a key within the timing window (including the "early miss" timing window) of the circle, but the cursor is not aiming the current circle, instead of making it as if the circle wasn't interacted with at all, make it instantly miss, unlocking the next note.
2. A variant of this would be adding a "misaim miss" area surrounding circles that gives you a miss in the circle regardless of timing as long as you hit them in their timing window. This area should be big enough to include the distance between circles in a jumpstream.

This would still cause an annoying issue with streams: Suppose you have a stream, and you misaim the first note (by starting the stream from the second note by accident). Now, the first click only causes a miss for the first note, but you still don't hit the second note, on which your cursor currently is. You would need to somehow instantly react to that and doubletap to register the second note. But because people can't react to stuff like that, you'll make the motion for the stream, missing everything because now the issue duplicates to all subsequent notes.

This solution would work perfectly though if one click would register the first note as a miss as well as register a hit on the note you are currently on (if you are on a note at all of course)

note locking is absolute cringe i hate it... no reason to penalize the player for hitting a note holy sh1t

MidnightBounty wrote:

note locking is absolute cringe i hate it... no reason to penalize the player for hitting a note holy sh1t

skill issue + ratio + L + low iq

5 years, not bad kid.

MidnightBounty wrote:

note locking is absolute cringe i hate it... no reason to penalize the player for hitting a note holy sh1t

why the hell did you comment on a 5 fucking year old post? how did you even find this shit?

Note locking still bad after 5 years..