Supercross "This Week" Predictions -- Salt Lake City 2017
As usual, the updated 2017 Supercross Prediction puts the whole season in perspective, while the "This Week" rankings look at the riders' perceived ability (aka "true talent") as well as taking into account injuries and the riders' history at the upcoming track.
**For those not familiar, a brief explanation of the components that factor into the "This Week" rankings:
The standard MotoXGraphs prediction of rider performance, based on the current and prior seasons, weighing recent results more heavily. It differs from the "Full Season" prediction in that it is not worrying about the rider's current point standings (which obviously are a crucial part of the full season prediction), only about their current ability.
Attempts to account for riders competing at less than 100% due to injury. Full research behind the Injury Adjustment can be found in The MX Book (or eBook). While this adjustment is only an estimate (what predictions aren't?), the purpose of the Injury Adjustment is less about trying to make an exact prediction and is more about indicating that a rider is recovering from injury so the viewer can take that into account at whatever scale he or she chooses.
Provides an increase or decrease in expected points for the riders based on their historical performance (over the past 4 years, where applicable) in the upcoming city/stadium. Also validated by research in The MX Book ( or eBook). The track/location effect has a slightly muted nature in Supercross compared to outdoors since the tracks are not exactly the same in Supercross sites from year to year (or even within the same year for sites like Angel Stadium or Petco where there are often multiple races in one season).
Here are the predictions for the 450 class this weekend.
Again, it might seem like a surprise that Ryan Dungey is predicted ahead of Eli Tomac, but that is due to Dungey being much better than Tomac in each of the past 3 seasons. The Track Adjustment (minor) narrows the gap very slightly and is probably not meaningful considering it's been a few years since a race at SLC.