# Mathematically impossible most common BPM? [confirmed] [resolved]

Might be because your BPMs are weird. >150.001 >199.995. It could be messing up the maths.

(BTW THIS HAS NOTHING TO DO WITH THIS PROBLEM BUT: You only need 1 timing section, use green lines for speedups. The BPM does not actually change to 200bpm)

no, none of these maps have broken timing or anything. they are perfectly normal multiBPM songs.
you broke math, m8
Some maps have the correct AVG BPM, example :-
http://osu.ppy.sh/b/41865

BPMs are : 140 , 145 , 145.3 , 125 , 134 , 144
AVG = 138.8833333333333333333......
In game = 139

But some of maps have the negative average BPM.

And Finally :-
Go home Math
You are drunk.
I'm not sure I should be confirming a Testbuild issue, but peppy can unconfirm it if it's not supposed to be confirmed.
This is a public build issue, not test.
This should be fixed in the next public release.
Average BPM is still not correctly calculated on latest test build (2012-12-28T09:27:48Z).
[20:22:05] <@peppy> then report it
http://osu.ppy.sh/b/85520
In game: 152-192 (~140)
Correct: 192

The following maps' average BPM in game lies within its BPM range, but it's widely inaccurate.

http://osu.ppy.sh/b/85355
In game: 152-192 (~169)
Correct: 188.062 (geometric mean, may be slightly different depending on how it is actually calculated)

http://osu.ppy.sh/b/170608
In game: 81-96 (~92)
Correct: 95.937

http://osu.ppy.sh/b/61124
In game: 75-168 (~105)
Correct: 164.907

http://osu.ppy.sh/b/188562
In game: 75-174 (~147)
Correct: 167.769

EDIT: Fixed explanation
Do the above cases still incur any error with the new display?

#### peppy wrote:

I've changed this, since I too do not agree with the current implementation. Is up on test build if you would like to test it out.
Would median or 75th percentile (or some other percentile) make more sense? (eg. on maps like ΔMAX)

Anyway it's broken on 2012-12-28T16:19:30Z test build

Edit: x% percentile := y bpm so that ≤x% of draining time is in <y bpm sections and ≤(100-x)% of draining time is in >y bpm sections
I haven't seen anything mathematically impossible yet in the new version (although I've only done a bit of looking), but there are still some very questionable results, like Flutter Girl ("75-168 (75)"... 75 BPM is easily the least representative BPM in the song, it's used for 2 very short transitional retards between the 166 Waltz/168 Common sections).

As for using median/percentile, I don't think it really makes much difference except in weird exceptional cases like medleys and deltaMax, so it's acceptable if things don't work too well for them (range is probably good enough for them). For most maps, the mode and the median are going to be the same because one BPM will dominate almost all of the song. Median is just a bit more prone to catching a value that represents very little of the map (although it requires some coincidences to happen), whereas mode guarantees that you represent as much of the map as possible with a single value.