James Stewart and the rest of 2016 Supercross
Word on the street is that James Stewart is going to be back on his bike in Oakland for Week 4, ready to put up some points on the 2016 450SX leaderboard. While we're all on the edge of our seats, waiting to see if he does ride and if he still has it, let's do some estimating with numbers.
So you don't think Stewart can still win the 2016 450SX crown?
Well, you're right. At least, unless one of Fox Mulder's secret-government-aliens abducts Dungey sometime soon. Nevertheless, let's see what it would take to beat Dungey without any X-Files conspiracies. Right now, Stewart is 71 points behind Dungey (it's 72 to 1), which Stewart would have to make up in 14 races. That works out to over 5 points per race. So if Stewart could somehow win every remaining race, AND Dungey finished 3rd in every remaining race, Stewart would still finish 1 point behind Dungey (not to mention the other riders, currently led by Chad Reed with 59 points).
OK, then what CAN Stewart do with the rest of the season?
He can add to his career total for Supercross wins -- he has the 2nd most, behind Jeremy McGrath (Stewart trails McGrath by 22 450SX wins in their respective careers). No, Stewart probably can't ever catch McGrath, considering Stewart's age, but Stewart could put some distance between himself and Reed (trails by 6) and Dungey (trails by 27). And who knows -- if Stewart can punch out a few wins this season, maybe that overall record starts to look vaguely plausible.
"Punch out a few wins" - is that even realistic for this season?
Now there's something we can analyze. Let's look at Stewart's recent 450SX win percentages, then project forward for this year -- using our standard weighting, 5:4:3:2 going back year-by-year (where the most recent year gets a weight of 5, the prior year a weight of 4, etc., which has become a standard method (from "Marcel" in baseball analytics) for bare-bones estimating performance for an upcoming season).
First, Stewart's career win %'s in 450SX:
In all, he's won 42% of the races he's entered, which is pretty crazy. But what we're focusing on here is the last four years, where his numbers have not been as good, 28% (one of those reasons was Villopoto, who's no longer competing; but Stewart does have to face an improved Dungey, so we'll call it a wash). If we apply the Marcel weighting from above, that gives us a projected winning percentage for 2016 of 20.6%. On top of that, we should apply an aging adjustment, which in Stewart's case is approximately a 4% penalty (being 30 years and 1 month), taking the projection down to 19.8%. So, with 14 races left and an expected win % of just under 20%, Stewart would be expected to win 2.8 races, so somewhere between 2 and 3 times finishing #1.
There are some major assumptions being made here, though. The first is that Stewart returns healthy and stays healthy. There's certainly no guarantee that he won't be slowed by after-effects of his concussion from Week 1. Also no guarantee that he doesn't get hurt again along the way. Second, we're assuming that his age penalty is similar to the average. But, some riders fall off more quickly, and some -- such as Chad Reed, it seems -- don't age nearly as quickly (as mentioned before, one goal is to soon look at how the top level riders age compared to the typical rider).
If we accept these assumptions and that Stewart is looking at something like a 20% win percentage for the rest of 2016, what are his chances at winning at least one race? First, his probability of winning any number of races, from 1 win to 14 wins:
More to the point, I think, is his probability of winning AT LEAST some number of races:
That's actually a much stronger statement than I expected - if he stays healthy, a 95% chance that James Stewart wins at least one race during the rest of 2016.
(Edit: based on a comment from RichieW13, I'm going to modify the "Aging" penalty to account for the fact that the ages we're comparing to Stewart's 30-year-old season go as low as 26. As such, the appropriate penalty, weighted the same way as the previous season win %, would be closer to 10%. That would change the expected win %, but doesn't change the likelihood of winning 1, 2, or 3 races by that much:
This change drops his % of winning at least 1 race by 1 % point, 2 races by 3.5 % points, and 3 races by 5 % points. So, that's something, but I think the overall point still stands that winning 1 or 2 races seems pretty likely based on these numbers and assumptions.)
Let's go back to the assumptions: that he's fully healthy and will stay that way, and that he's more or less the same rider who we last saw in 2011 through 2014 -- after accounting for a slight decline because of aging. Now let's get pessimistic -- he actually might take a week or two to acclimate from his head injury, he may miss a race or two along the way with a minor injury, and he's aging or rusty or out of shape from missing 2014. So let's cut his win % in half, just to see what happens:
That's actually a little more like I would have guessed. Probably you too. If so, then our guts might be underestimating what James Stewart can do, compared to the numbers. But then again, the numbers we're looking at can't see his current and future injuries. So for Stewart's sake, let's hope he really is back to 100% and can keep it up, meaning our guts are wrong and the higher numbers prove to be right.
Posted by: SagehenMacGyver47 ::: As always – Feedback welcomed
Previous: Week 3 – By-The-Numbers Recap: Anaheim 2 | Back to Blog | Next: 2016 Supercross Predictions - 450SX Update after week 3 at A2