Hi everyone!
I recently started playing iGP Manager and decided that what I really needed was a statistical model that could help me decide which tyre compounds to use for each of my sprint races. It kind of got out of hand and what resulted is a model that is trained on lap times, push levels, tyre compounds, stint lengths, temperature, and wear, as recorded from over twenty sprint races.
The worst part is, now that the model is fully trained, it suggests a 1-stop Hard-Hard strategy for every single sprint race (with the 2-stop Hard-Hard-Hard often in a close second). It seems like some players had already surmised this, but apparently that's statistically the best way to minimize the total race time. Anyway, had anyone else noticed this? Does it seem true based on your experience playing? It may also be an artifact of only having relatively little data still at this point, or of how DRS comes into effect almost randomly as I just let the race play out without boost or anything.
Let me know if you're interested in the model and I can try to pass other knowledge along. Since cars are relatively equalized for sprint races, this part of the model was easier--the league model seems more complicated than I want to tackle right now.
