The qualifications are:
1) We only save full data for the last 1hr of stats, so if we don't find 2 blocks within the same hour, we don't have an easy way of seeing how many shares on that particular coin were mined. That really only happens on either litecoin, or dogecoin if we skip mining it for a decent bit of time, almost all the rest of the blocks have luck tracked.
Basically, I am interested in a snapshot periodically of the shares we have mined at every difficulty on every coin. And if we found blocks at that difficulty/coin (many rows would be 0 for a KGW coin, presumably).
2) The luck is based on the found block difficulty, which is not necessarily the difficulty the shares were at. For example, if we mine dogecoin as difficulty is on the rise (say some shares at 1k, and then some shares at 1.1k), then stop mining because diff is too high, wait a bit until it drops to 800, and immediately find a block. The luck is based on all of the shares being submitted against a 800 difficulty blocks (which is incorrect for luck calculations, but good enough for my purposes).
I totally understand. And I know the data I am won't look like much on its own. I am interested in calculating basically, "expected profitability" based on shares mined at each difficulty.
I just worry with showing these stats that people will freak out. Better option is for me to properly track luck values (just another thing on the list), but that has some (smaller) subtleties as well, since we buffer inserts into our database for share submissions, so data will be off based on when the block comes in compared to when shares get inserted (although very slightly). Maybe I'll work on this in the coming days, luck would be nice to have publicly available.
I think people generally like more stats =). IMO, that's a big part of the reason why WafflePool tends to attract hashpower like it does. Luck would definitely be nice to have public. What I am asking for would require a new row every difficulty change that recorded how many shares were accepted at that difficulty and how many blocks were found (if any). If you felt like throwing that in JSON or something, I could probably build some calculations and a front-end.