This is really interesting, thanks for posting it. It was super accurate for me at Year 1, I did weigh very close to the prediction. Year 2 is where I started to diverge, in reality I kept losing but the prediction was essentially that I would stay the same. I switched it to Lap RnY to see how it would affect the prediction - my projected weight loss went up by 15 pounds and my probability of complications went up 10% (to 20% total). With open RnY, the projected weight loss didn't change but the probability of any complications was 30%. I would love to see the model this is based on, it's pretty fascinating.