Some short spinners can be left unplayed and still be completed

[pengyllis]

Type

Game behaviour

Bug description

Some short spinners don't require to interact with them to keep an SS on your play.

When testing on local maps, I could only reproduce this issue on spinners with a length of 400ms or below. However, on this map, it occurs with all 500ms spinners and some, but not all, 667ms spinners.

Screenshots or videos

While the video shows an autoplay of the previously mentioned map, the spinner behaviour is the same when playing. Note that the first and second spinners shown here are the same length of 667ms but only the first one has proper behaviour.

2025-06-28.20-53-00.online-video-cutter.com.mp4

Version

2025.607.0-lazer

Logs

compressed-logs.zip

[Joehuu]

This is because of:

osu/osu.Game.Rulesets.Osu/Replays/OsuAutoGenerator.cs

Lines 154 to 156 in 5c89644

	// spinners with 0 spins required will auto-complete - don't bother
	if (spinner.SpinsRequired == 0)
	return;

Not sure if this should be left open.

[frenzibyte]

Does it match stable? It doesn't feel so judging by the map design alone, but I haven't tested yet.

[peppy]

There's an ongoing proposal to change the ranking criteria around spinner completion. I've raised this thread with those involved.

[pengyllis]

Does it match stable? It doesn't feel so judging by the map design alone, but I haven't tested yet.

I have done some testing on stable and... kind of, actually? It seems that this specific map has odd behaviour with both clients, which I didn't expect at all when making this issue.

From what i saw on local test maps and other existing maps, it generally seems that in stable, unplayed spinners under 200ms will give a 300, and those over 200, 400, and 800ms will give a 100, a 50, and a miss.
While in lazer, spinners under 400ms give a 300 and the others give a miss.
(Which looks intentional? idk, it's not documented anywhere I know of, like the gameplay differences wiki page.)

Anyway, what's interesting is, in the Tsurupettan map, the exact same unplayed spinners that give a 100 on stable give a 300 on lazer, and those that give a 50 on stable give a miss on lazer, down to the variations in the 667ms spinners, despite these judgements not making sense in either client considering the spinner durations.

I apologize if this makes the issue irrelevant, or if I perhaps missed something about this map that makes that behaviour intended.

[AJT127]

https://osu.ppy.sh/wiki/en/Gameplay/Judgement/osu!#spinners

Taking this formula, the behaviour on stable checks out: 667ms OD0 spinner requires 0.375ish spins, which gets rounded down to 0 because internal spinner calculations are rounded down to the nearest half, so 0 spins for a 50. I'm not well equipped enough with lazer to clarify here - I'm not sure why you have to spin on lazer client

[peppy]

@smoogipoo any thoughts / memories of this code?

[smoogipoo]

Unfortunately not...

[peppy]

Turns out the duration_error is unrelated, so that's ok.

The crux of this issue looks to be what is described in the OP of #24932. Which is to say, the cutoff required for passing has not changed, but the cutoff for clear has changed intentionally in that PR.

Which means

• At face value, the response to the title of this issue is a "yes, that's an intentional mechanic and matches stable"
• Autoplay is not spinning hard enough on these spinners (to get a 300), which could be considered an actual bug scratch that, it does. arguably it should look like it's attempting to spin just to look cool?
• Subjectively, the minimum spin cutoff has changed, but not in a way which should be hugely detrimental to users, so I don't believe we need to modify that.

[peppy]

@smoogipoo would appreciate your brief thoughts on the above and if you don't see an issue with this discrepancy let's close this issue as "intentional" and I'll look to have this documented on the wiki so people are aware of the difference.