Tagged: projections

2014’s SP projections: the best and the worst

Maybe this is absurd, but I’ve never honestly checked the accuracy of my projections. It’s partly because I have placed a lot of trust in a computer that runs regressions with reliable data I have supplied, but it’s mostly because I originally started doing this for my own sake. I used to rely on ESPN’s projections, but the former journalist in me started to realize: it has a customer to please, and the customer may not be pleased, for example, if he sees Corey Kluber ranked in the Top 60 starting pitchers for 2014. (At this point, I am giving ESPN an out, given that everyone at FanGraphs and elsewhere knew the kind of upside he possessed.) Kluber is not the issue, however; the issue is that although ESPN (probably) wants to do its best, it also does not want to alienate its readers who, given its enormous audience, are more likely to be less statistically-inclined than FanGraphs’ faction of die-hards.

In sum: I started doing this because I no longer trusted projections put forth by popular media outlets.

So I didn’t really care how every single projection turned out. I wanted to find the players I thought were undervalued. For three years, it has largely worked in my rotisserie league. (Honestly, I am a complete mess when I enter a snake draft.)

Anyway. All of that is no longer. I quickly sampled 2014’s qualified pitchers — 88 in all — to investigate who panned out and who didn’t. I will ignore wins because they are pretty difficult to project with accuracy; I’m more concerned about ERA, WHIP and K’s.

Here is a nifty table that quickly summarizes what would have been tedious to transcribe. You will see a lot of repeat offenders, which should come as no surprise. At least there is some semblance of a pattern for the misses: I underestimated unknown quantities (and aces, who all decided to set the world ablaze in 2014) and overestimated guys in their decline. There isn’t much of a pattern to the guys I got right. Just thank mathematics and intuition for that.

Here would be a shortlist of my most accurate projections from last year, measured by me using the eye test:

Name: 2014 projected stats (actual stats)

Nathan Eovaldi: 5 W, 3.82 ERA, 1.32 WHIP, 6.4 K/9 (6 W, 4.37 ERA, 1.33 WHIP, 6/4 K/9)
R.A. Dickey: 11 W, 3.84 ERA, 1.24 WHIP, 7.2 K/9 (14 W, 3.71 ERA, 1.23 WHIP, 7.2 K/9)
Alex Cobb: 12 W, 3.49 ERA, 1.17 WHIP, 8.1 K/9 (10 W, 2.87 ERA, 1.14 WHIP, 8.1 K/9)
Hiroki Kuroda: 11 W, 3.60 ERA, 1.18 WHIP, 6.7 K/9 (11 W, 3.71 ERA, 1.14 WHIP, 6.6 K/9)
John Lackey: 10 W, 3.67 ERA, 1.25 WHIP, 7.5 K/9 (14 W, 3.82 ERA, 1.28 WHIP, 7.5 K/9)
Kyle Lohse: 9 W, 3.60 ERA, 1.17 WHIP, 6.1 K/9 (13 W, 3.54 ERA, 1.15 WHIP, 6.4 K/9)

If it brings consolation to the reader, I have since tightened the part of the projection system that predicts win totals. I’m not gonna lie, it was pretty primitive last year because I thought it’s already a crapshoot to begin with. Obviously, it shows, even in the small sample above. It’s still difficult given the volatility inherent in the category, but the formulas are now precise.

Advertisements

Updated SP rankings

Click here for updated SP rankings.

I have updated the starting pitcher rankings to reflect offseason signings, rotation battles and spring training injuries — and holy cow, have there been a lot of spring training injuries.

I also truncated the list to the top 90 pitchers. I will write about my favorite pitchers outside the top 90 because a lot of them are really good; they simply won’t get enough get enough starts or pitch enough innings for them to crack the top 90 in value. In terms of stuff, though, there are plenty of diamonds to find in the rough.

Stock up: James Paxton, Justin Verlander, Ervin Santana

Stock down: Cole Hamels (shoulder), Hisashi Iwakuma (finger), Kris Medlen (elbow), Patrick Corbin (elbow), Jarrod Parker (elbow), A.J. Griffin (elbow), Brandon Beachy (elbow), Miguel Alfredo Gonzalez

2014 Rankings: Closers

Rankings based on standard 5×5 rotisserie format.

Name – Saves / ERA / WHIP / K’s

  1. Craig Kimbrel – 47 / 2.32 / 0.65 / 106
  2. Aroldis Chapman – 41 / 2.42 / 0.83 / 114
  3. Kenley Jansen – 39 / 2.48 / 0.87 / 103
  4. Greg Holland – 42 / 2.21 / 0.99 / 97
  5. Trevor Rosenthal – 39 / 2.41 / 1.00 / 90
  6. Koji Uehara – 34 / 2.42 / 0.69 / 81
  7. Joe Nathan – 40 / 3.15 / 0.95 / 72
  8. David Robertson – 38 / 3.13 / 1.05 / 82
  9. Jason Grilli – 34 / 2.80 / 1.14 / 78
  10. Sergio Romo – 36 / 2.93 / 0.99 / 67
  11. Grant Balfour – 43 / 3.46 / 1.11 / 74
  12. Glen Perkins – 34 / 2.93 / 0.98 / 68
  13. Ernesto Frieri – 36 / 3.74 / 1.14 / 91
  14. Steve Cishek – 31 / 2.92 / 1.14 / 70
  15. Casey Janssen – 34 / 2.91 / 1.01 / 54
  16. Addison Reed – 32 / 3.19 / 1.18 / 71
  17. Jonathan Papelbon – 33 / 3.30 / 1.14 / 66
  18. Jim Henderson – 32 / 3.76 / 1.18 / 80
  19. Fernando Rodney – 32 / 3.26 / 1.32 / 74
  20. Bobby Parnell – 32 / 2.76 / 1.16 / 48
  21. Nate Jones – 30 / 2.64 / 1.22 / 52
  22. Jesse Crain – 27 / 3.13 / 1.09 / 61
  23. Huston Street – 29 / 2.52 / 1.15 / 47
  24. Jose Veras – 33 / 3.62 / 1.22 / 69
  25. Rafael Soriano – 43 / 3.85 / 1.25 / 52
  26. Jim Johnson – 36 / 3.42 / 1.17 / 41
  27. John Axford – 35 / 4.36 / 1.33 / 80
  28. Neftali Feliz – 29 / 4.13 / 1.19 / 43
  29. Rex Brothers or LaTroy Hawkins
  30. Chad Qualls – pending

Thoughts:

  • All ERAs are inflated a little bit. Closers (and relievers in general) tend to strand more runners than starters and, thus, prevent runs from scoring as often. My model fails to capture this nuance, but the difference isn’t a huge one, as a 2.32 ERA from Kimbrel is still really, really good. But for a guy with a career 1.38 ERA, it makes sense to expect even better from him.
  • The top 5 are pretty much consensus picks. I think Uehara is worth considering as part of a potential “Top 6” elite tier of closers, and he is absolutely better than Nathan. Are you aware that Uehara has posted a 0.702 WHIP in 219-1/3 innings since 2009? Are you serious? And he still strikes out double-digit batters per nine innings.
  • Johnson is absolutely overrated. The Baltimore Orioles generated 113 save situations the past two years. The Oakland Athletics, Johnson’s new employer, generated only 83. That’s two-thirds the opportunities he used to get. If you’re expecting 50 saves again, you’re crazy. He also strikes almost no one out. Try to catch lightning in a bottle if you want, but I think he is one of the worst investments in the game for saves.
  • Henderson and Crain are really underrated (compared to ESPN), but they also don’t have the job security. That leaves Frieri as the last true bargain. He walks too many batters, but at least he strikes out twice as many as Johnson does. Also, if the Angels bounce back in a big way, he will be the beneficiary of greater workload.
  • Sorry, I was too lazy to project Brothers or Hawkins. I just don’t think Hawkins will last long, but it’s tough to say exactly how long, and it’s not worth guessing. Just get him on the cheap, handcuff Brothers to him and be ready to jump ship.

2014 Rankings: Catcher

Rankings based on standard 5×5 rotisserie league.

Name – R / RBI / HR / SB / BA

  1. Buster Posey – 69 / 85 / 20 / 2 / .299
  2. Wilin Rosario – 67 / 78 / 27 / 4 / .267
  3. Jonathan Lucroy – 60 / 78 / 17 / 6 / .293
  4. Yadier Molina – 59 / 78 / 16 / 4 / .296
  5. Brian McCann – 58 / 82 / 20 / 3 / .281
  6. Joe Mauer – 72 / 66 / 10 / 5 / .297
  7. Wilson Ramos – 54 / 73 / 26 / 0 / .283
  8. Carlos Santana – 76 / 75 / 20 / 5 / .253
  9. Matt Wieters – 60 / 73 / 23 / 2 / .256
  10. Evan Gattis – 52 / 76 / 24 / 0 / .264
  11. Miguel Montero – 65 / 75 / 15 / 0 / .260
  12. A.J. Pierzynski – 58 / 64 / 15 / 1 / .277
  13. Jason Castro – 74 / 58 / 17 / 1 / .245
  14. Salvador Perez – 54 / 71 / 12 / 0 / .270
  15. Yan Gomes – 56 / 53 / 16 / 1 / .278

Thoughts:

  • McCann? At Yankee Stadium? Yes, please.
  • Gomes is Cleveland’s starting catcher; my projection doesn’t account for that. Give him a full year of at-bats and he’s easily a top-10 catcher who threatens the top 5.
  • Perez is “due” for a breakout, as everyone says. Try him for a bargain pick, but I think there are adequate substitutes you can still get for cheap without risking the lack of counting stats.
  • Castro may be safer than Gattis, but even an underwhelming Gattis may still rival Castro’s numbers fueled by an anemic team and spacious ballpark.
  • I love Lucroy but, like yesterday’s shortstops, he’s not definitively better than Molina. Just look at their projections — they’re almost identical. Go with your gut.

2014 Rankings: Shortstop

Rankings based on standard 5×5 rotisserie league.

Name – R / RBI / HR / SB / BA

  1. Hanley Ramirez – 96 / 88 / 22 / 19 / .306
  2. Ian Desmond – 83 / 82 / 25 / 21 / .279
  3. Troy Tulowitzki – 84 / 85 / 24 / 3 / .305
  4. Jose Reyes – 95 / 56 / 10 /26 / .295
  5. Elvis Andrus – 102 / 62 / 4 / 35 / .275
  6. Jimmy Rollins – 96 / 94 / 15 / 20 / .262
  7. Brad Miller – 80 / 70 / 16 / 11 / .283
  8. Jean Segura – 77 / 52 / 9 / 34 / .279
  9. Everth Cabrera – 77 / 45 / 4 / 50 / .268
  10. Alexei Ramirez – 79 / 75 / 9 / 22 / .266
  11. Starlin Castro – 81 / 66 / 11 / 15 / .275
  12. Andrelton Simmons – 77 / 64 / 16 / 9 / .270
  13. Erick Aybar – 73 / 53 / 11 / 17 / .285
  14. J.J. Hardy – 75 / 72 / 23 / 1 / .255
  15. Asdrubal Cabrera – 71 / 66 / 13 / 10 / .263
  16. Jonathan Villar – 68 / 42 / 3 / 46 / .234

Thoughts:

  • Need cheap steals? Villar is your guy. He strikes out too much to be a reliable hitter, but he could end up being one of those high-BAbip guys because of his speed. He needs to shore up his plate discipline to be anything but a one-trick pony. He’s only 22, though, so he has plenty of time to figure it out. In the meantime, ride the wind.
  • Miller is one of my big sleepers for the year and will be the topic of a forthcoming bold prediction.
  • Rollins’ RBI number above is absurd. At the risk of discrediting the projection, it will occasionally spit out a statistical anomaly. His RBI are certainly the elephant in this room.
  • Not a lot to see here, although the spotlight is on Castro right now. I would say this projection is closer to his floor — that is, he’s a No. 11 shortstop with upside. In all honesty, though, the Nos. 7 through 11 shortstops are all really, really close. I’d be better off calling them all a tie, but that’s taking the easy way out.
  • Desmond has hit 20 home runs two years straight. His power-speed combo up the middle is unmatched — except by Ramirez, of course. If only Tulowitzki would run more, let alone stay healthy…