I have not gotten to share much of the rules I have developed at my current job, but I do have something I have been playing with that I have found useful. I made a series of toy statistical functions I thought I would share. I am still working on them a little bit (I might try overloading them to simplify them). They use string lists to compute statistics which is clearly not ideal, but it does have its uses (largely using string lists to potentially control shapes, adjust splits, etc). They are not perfect or optimized, but they work.
Let me know if some other statistic might be helpful for your work.