I use stats and such intensively, though not so much in SQL Server. We're a manufacturing company and I've designed and built from the bottom up custom SPC (statistical process control) applications for all aspects -- raw materials to packaged product. We also have a sensory lab, and I designed and built the app for experimental design, flights, data collection and analysis. Some of this was done using math libraries imported into python, but others we just rolled our own. I originally used some heavy artillery type programs like SPSS and SAS (later JMP, which for our purposes was more than enough), but found that rolling my own was 1) easier to give the end user what they needed and 2) way more fun.

We also use a data viz product where it is a bit easier to analyze data, but as with all the other projects it works best when done in close collaboration with the database. By that I mean being able to craft a custom, tight dataset makes the final analysis MUCH easier. I kind of grew these skills side by side, so I've come to appreciate both analysis and db skills even more. This board is mostly sql pros, but if you can learn that you can learn stats.