forum

[confirmed] Same Map, Different SR

posted
Total Posts
8
Topic Starter
Ciel
Map in question is https://osu.ppy.sh/s/312935

The 2 (current) difficulties "burst tests" and "Unraveling Destruction" contain the exact same notes. (verified by a program i wrote). However, the notes in the 2nd diff were copy and pasted from the first diff, which resulted in a massive increase in SR (at least on the client, from 8.22* -> 8.37*). I suspect this is caused by the fact that the notes, while exactly the same, are ordered differently in between the 2 files, thus somehow breaking the SR algorithm (which it shouldn't).

osu! version: 20150924.4beta
Topic Starter
Ciel
badump
Bara-
Weird
Perhaps it's not completely snapped, thus times between jacks are different
Topic Starter
Ciel
This happens with other (ranked) maps too though.

For example, (taking some of the hardest ranked maps, as those differ more clearly):
D(ABE3) - MANIERA [Masterpiece]: 8.04* -> 8.03*
Yuyoyuppe - AiAe [Wafles' SHD]: 6.77* -> 6.76*
orangentle / Yu_Asahina - HAELEQUIN (Extended ver.) [Jinjin's Dissociative 7K Identity Disorder]: 6.48* -> 6.47*
Yuuna Sasara feat. Tai no Kobone - Imperishable Night 2006 [7K Lunatic]: 7.13* -> 7.20*
sakuzyo - AXION [Ex-ray]: 6.48* -> 6.60*

I've run both the original maps and copied maps through a program, that double checks that every note is matched in between the 2 difficulties. The only difference is the ordering in the .osu. It would be nice if someone else could confirm this maybe? (Though it would take a bit of coding on the side)
Bara-
What do you mean with ordering in the .osu?
Topic Starter
Ciel
http://puu.sh/kFprP/e2e954d49f.txt - 8.22*
http://puu.sh/kFpo3/5a2d3c0981.txt - 8.37*

Just manually inspect the .osu yourself. (HitObjects)

EDIT: There is a change between the things i just posted and what's uploaded, in that I increased the HP of all diffs to 8.5. In case that explains any inconsistencies.
Bara-
Hmm
This is weird
It's the same, except that certain objects at the same time are switched
TheVileOne
I'm working on a solution for this.

Edit: I spoke with tom and he said it's better to rewrite the code rather than patch the existing code. This will need to wait until he has the time to do it.
Please sign in to reply.

New reply