User blog comment:Sirebel/Mathematician Required - Apply Within/@comment-27170954-20170425003411

My interpretation of your math problem is you are assuming a displayed 0.3 PR upgrade is really an upgrade with a PR between 0.2500 and 0.3499 (ir rounded) or between 0.3000 and 0.3999 (if truncated) and you would like a method to figure out what the real value is.

I'm not sure you can arrive at the exact value, but you might get close. I think It will also be different for each car.

There might be a way to estimate using statistical regression. If you have a dataset that included several (but not necessarily all) final PR levels you could run a regression against the incremental PR for each upgrade and get an estimate of the fractional PR (i.e. the part after the first decimal).

I would have to think about what data you'd really need, but my first impression is you would need a few data points for each upgrade where the change in final PR isolates the contribution of that individual upgrade. In otherwords, if you want to know the true value of engine 2, then you will need a handful of incremental PR values (both actual and expected increments) for engine 2 from different starting points.

In case you are wondering - I am in a graduate program that involves statististics (only about 20% complete) and this sort of question intrigues me. I'd be interested in playing with it, but given my workload for the next several weeks it will take me a while to formulate the problem and data needs.