Wednesday, August 7, 2013

How My Starters' List is Stacking Up (Not Well)

So a while back, I did a post ranking the Opening Day starters for each team this year. I explained my FOOL-PROOF, SUPER PREDICTIVE methodology, and produced a top-notch list. Now, with July in the rear-view mirror, it's time to take a look back at that list and see how I did.

Below is a list, 1-30, and two names appear at each number. The first name, in boldface, is the list of Opening Day starters based on their performance this year (I'm basically using FIP and WAR from Fangraphs to determine this). The next name is where I placed that name on my list on April 5th. Let's see just how wrong one can be 4 months later.

30. Vance Worley/Jhoulys Chacin
29. Matt Harrison/Edinson Volquez
28. Brett Anderson/Ricky Nolasco
27. Jason Hammel/Matt Harrison
26. Johnny Cueto/A.J. Burnett
25. Jon Niese/Justin Masterson
24. Edinson Volquez/Vance Worley
23. Ian Kennedy/Tim Hudson
22. Yovani Gallardo/Bud Norris
21. R.A. Dickey/Jason Hammel
20. Matt Cain/Jon Niese
19. Ricky Nolasco/Jeff Samardzija
18. C.C. Sabathia/Yovani Gallardo
17. Tim Hudson/Jon Lester
16. David Price/Jered Weaver
15. Jered Weaver/Johnny Cueto
14. Bud Norris/Adam Wainwright
13. Jon Lester/Ian Kennedy
12. Jeff Samardzija/Matt Cain
11. James Shields/R.A. Dickey
10. Stephen Strasburg/James Shields
9. Cole Hamels/C.C. Sabathia
8. Justin Masterson/Brett Anderson
7. A.J. Burnett/Chris Sale
6. Jhoulys Chacin/Clayton Kershaw
5. Justin Verlander/Stephen Strasburg
4. Chris Sale/Cole Hamels
3. Clayton Kershaw/Justin Verlander
2. Felix Hernandez/Felix Hernandez (Hey!)
1. Adam Wainwright/David Price

Wow. Ehhhh not great. So let's test my predictive model. I used SIERA, xFIP, BABIP, WAR, age, and games started (in that order of importance) to determine my ranking. Let's figure out just how far away I was.

An easy way to do this is to take each pitcher's difference between where I ranked them and where they actually are now, and average those out. That'll be the average amount by which I was wrong. Then let's take that figure and divide it by the overall items on the list (30 in this case), and get the percentage by which I was wrong. Let's call it my WRONGNESS FACTOR.

My Wrongness Factor, as of August 7, 2013: 26.

Wow. So essentially, on average, my picks were about 8 spots off (that means about 26% off) from where they actually ended up as of today. I nailed King Felix, sure, but I have a mammoth 24 spots in between real Jhoulys Chacin and predicted Jhoulys Chacin, not to mention 20 spots for Brett Anderson and 19 for A.J. Burnett.

So it just goes to show you: never listen to me. I'd be interested to see other lists' Wrongness Factors on this and other sports websites. Maybe 26 ain't so bad in context! Maybe?

No comments:

Post a Comment