I recently started running 70 star and my split layout included subsplits for each individual star. This was firstly to help me remember the star order and secondly to make it so I can easily see which stars I lose time on vs my pb. However, I sometimes missed tricks (e.g cannonless) and did a different star as a backup, so I had to 'skip' those splits to avoid getting any unrealistic golds. However, livesplit still counts skipped splits towards to sum of best (I think), so now I am quite confused as my sum of best is different to the total of my golds. Is there any way I can fix this and if not, should I change my layout to one split per level to avoid the confusion.

sometimes livesplit can be really, really weird with your sob. one time in a category my bpt would go up by 20s randomly in one split, even if it was super close to gold. to fix it, i ended up just making new splits.

but from what your problem looks like, i’d say just make it one split per level. you can always have the order on a doc or something just in case.

Not an expert on this one but I think I know what could be happening here. Livesplit uses a system where, if you skip but the total time for two stars is faster than the sum of those two golds, that time will still be used for the SoB calculations. Say, for example, that your canonless and wild blue splits are consecutive and each have a gold of 30 seconds, but you do a run where you skip the cannonless split but the time for wild blue is 55 seconds, livesplit will use that 55 seconds in place of the two golds. While this may not be what is happening in your case, this happened to me in a different game so that could be the issue. If you do not wish to have this happen, there is a setting somewhere in livesplit called “simple sum of best calculation” that you should turn on. Best of luck with your runs.

Yeah, I thought that was happening. I did try pressing simple sum of best calculation but that didn't completely work. It changed my SoB from a 53 to a 55, but the actual sum of my golds is a 58.