I think it's a bug

@Eduard Gerber, thanks for posting! Interestingly enough, I'm also level 31 on Reverb Wizard but my percentile is 90.5% which is only 1% more than your level 16. Strange.
I've been trying to develop evenly in all the games based on percentile, not level. In other words, who cares what level I'm on as long as my skills are tracking with my peers. So now I'm not so sure that's been a good strategy because I've been pushing levels in some of these games and not seeing the percentile go up.

Odd yes, I'm 83.49 on level 9. I just finished today's round and it kicked my butt, so....

I'm on Level 26 for 87.56%

Hi, there might be a bug with this game stats. We will look at it.

There seems to be a lot of users reporting game stats bugs. Perhaps SoundGym should review this whole area in its code.

Agree, Ross - I'm not too hung up on the stats and points, but on the other hand, SoundGym hasn't really done any revealing at all regarding how this stuff is calculated/tracked, either.... maybe some insights from them would be helpful other than "we'll look into it". Not grumbling, just observing.... feels kinda like working with Apple....

A principle of debugging code is that when you find a bug, you don't just fix the bug, you go looking for all its brothers and sisters and uncles and aunts. You identify the class of error that the bug represents and try to eliminate the whole class in the codebase. SoundGym has received a signal that there might be bugs in its calculations, so it shouldn't just fix a reported bug; it should review that whole area of the codebase.

Fixing the bug which is competitive in nature is fine. I would rather the developers spend more time making additional gym games (i.e. Bass Detective). I don't have much use for EQ Cheetah but still play it now and then for a warm up. The whole idea behind sound gym is to strengthen our ability to hear very subtle differences. Most of the heavy lifting for mixing and mastering in a home studio is to improve/finish the creative process associated with the music. The "average Joe" doesn't begin to hear what we hear.

I'm on level 30 with a 98.73% rating. What are your top scores? My highest is 4,894,400.

My top score at level 31 is 6,510,932. Honestly, I don't really care about being in the high percentile. I was just using percentile as a metric to help guide which games I put extra time on.
Since the difficulty between games is so different it doesn't seem to make sense to worry about which level I'm on. In other words, level 31 on Balance Memory isn't necessarily the same difficulty as level 31 on Peak Master.
But if I'm sitting in the same percentile range across all the games, that tells me my skills are developing more equally. So if the percentiles can be fixed, that would be cool.

wow. just checked -- I'm at level 3, as it's been included in my trainings only twice for half a year.
seems like the algorithm should really be improved.

Jason, I'm on level 22 in the 97th percentile - sounds like one or both of us is "mis-percentiled."

Thanks for the update. It seems that one of the background processes stopped running. We fixed it and it is now back on, so everything should run smoothly again. If you experience any issues please contact support.

SoundGym: Use systemd !!! Or, if systemd isn't available, create a process that does nothing but watch to see if any of the other processes that are meant to exist have disappeared and start a new one. Did you know that Netflix has a process called "Chaos Monkey" that goes around killing random processes in the Netflix cloud so as to force the engineers to engineer robustness into everything?!

SoundGym: I still have level 22 in the 95th percentile, which seems wrong. compared to the other Balance Memory percentiles declared in this thread. Is level 22/Percentile 95, in fact, correct?

Awesome. My stats page looks completely different now! Thanks for fixing it. @John Miller, yeah, I think that's right because I'm now showing 98.92% at level 31.

Thanks, Jason. @Jason Romney