Airship Guide

[Arcticharm]

Be careful drawing conclusions from small datasets. RNG is RNG - some people could just be getting unlucky. Looking at the data on the spreadsheet, which is still a small dataset, I'm not seeing anything that clearly influences sector discovery. Yes, some people did do a stat upgrade/downgrade and then immediately find a sector. Others have done the same and not found a new sector. Some have kept their stats the same, doing the same route, and eventually found the new sector.


Btw, the only strong correlation I've found so far from mining the data is that higher favor stat correlates with higher per-sector ratings. That could be because another stat is also high, like survey, and not be related directly to favor at all, but that correlation was clear. I also did a check for rank vs ratings and I didn't see a strong correlation there at all. The ratings were kinda all over the place with various rank ranges. Still, this is with a relatively small dataset (a few hundred voyages).

Also, our data for Retrieval scores may be off =\ I hadn't realized until this weekend that there were two more possibilities - "sub-normal" and "low" - so those weren't options in the chart until now...

Yeah, I'm with you on not jumping to conclusions. I'd like to optimally sift through the data to find trips with and without discovering sectors and then compare their stats but unfortunately all of the data is confounded by other stats, rank and changing ship parts. Because of this, no two flights are really ever the same and so they can't be directly compared. I realize that some people are going to get the short end of the stick with RNG but it'd be nice to know that it's only RNG and that there's nothing else they could do. It's also somewhat convoluted in that sectors can only be discovered once which limits the total number of data points we can really get for it.

[Nyalia]

It's also somewhat convoluted in that sectors can only be discovered once which limits the total number of data points we can really get for it.

Yeah... Since we have the current sector unlocks in a standard format, we could put the unlocked sectors into a standard format as well in a new column, and then run a formula that only looked at unlock possibilities, but I don't even know what qualities to test against...