<p>Sam - It’s an imperfect approximation, done on-the-fly -> feel free to improve it to get closer to the truth! </p>
<p>(the only assumption I made was the 18% OOS weighting - the differential was purely built with excel where I applied various combinations of in-state and OOS scores to arrive at the weighted average that was posted on collegeboard.com - this one seemed plausible to me, but obviously there is room for improvement)</p>