Time to Advance Our Prediction Models
September 15, 2010Chiefs Will Be “Paying Attention” to Jerome Harrison
September 15, 2010Note: this piece is not about Cleveland sports per se, though it is about the responsibility of one Cleveland sportswriter in particular. Consider yourself warned.
I’ve been debating about whether to respond to this piece in last weekend’s Plain Dealer, but considering that Paul Hoynes writes to a massive and impressionable audience, I think it’s probably worth the time and energy.
In the linked piece, Hoynes suggests that the only stat that matters for a pitcher is wins, and therefore, the Cy Young award should be given to the pitcher with the most wins in his respective league. He cites last year’s winners (Zack Greinke and Tim Lincecum, with 16 and 15 wins respectively) as evidence that “[t]he value of the win has been devalued by members of the Baseball Writers Association of America who favor statistics meant to remove all influences on the pitcher except himself.” He thinks this is bad.
Needless to say, I disagree. And I’m a little offended. So offended, in fact, that I tore the piece apart, line-by-line, in firejoemorgan.com style. But that’s not going to do anybody any good. So instead, I’ll try to refute Hoynes’ point without being a snarky jerk.
His point, as best I can tell, is that the game is about winning, so why shouldn’t we ultimately award the pitcher with the most wins—after all, that’s the point of baseball.
I actually agree with this — at least the part about the supremacy of winning. It’s why so many baseball fans try to figure out how teams win games. That’s what ERA is for, after all: pitchers with low ERAs tend to help their teams win more. That’s why we measure home runs and RBI—players with a lot of each help their team win games. That’s why some really smart people have tried to estimate how many wins certain players add to their teams (WAR). It’s why we spend time trying to separate defense from pitching (FIP and UZR). We do all this because we know that good players help their teams win games. It’s not to torture people with math or to self-aggrandize ourselves with important-sounding acronyms. We do it because we love baseball and thinking about its intricacies. Some of us even write about these newer ways of thinking, with the hope that other people will see that they’re not so scary—they can even be fun.
But there’s a more insidious message in Hoynes’ piece that needs refuting: the “pitcher-win” is NOT the same thing as a “team win.” A pitcher win is a decision made by the official scorer to give credit (often undue) to one player for the accomplishment of an entire team. It’s a silly and archaic statistic. Don’t believe me? Check this link out, and tell me with a straight face that the most basic statistic in baseball is the pitcher-win. A team win is much simpler: who scored more runs? They’re NOT NOT NOT NOT NOT the same thing. Hoynes is conflating two very different things, because it makes his argument easier.
Furthermore, just like a lot of other statistics, the pitcher-win is subject to the vagaries of luck, defense, run support and a thousand other things a pitcher doesn’t control. Because of this problem, people have started thinking of better ways to evaluate how well a pitcher really performed. ERA plays into this. Strikeouts play into this. Innings pitched plays into this. And for some, FIP, xFIP, K/BB, and WAR (that’s WINS above replacement) play into this evaluation. These are good things, in my mind. They try to give us a better picture of the things a pitcher can control.
And this isn’t a revolution of bloggers in their mothers’ basements, either. Many newspaper men have adopted the more nuanced ways of thinking about players: Joe Posnanski (formerly of the Kansas City Star), Bernie Miklasz (St. Louis Post-Dispatch), and Dejan Kovacevic (Pittsburgh Post-Gazette), just to name a few, have incorporated statistical advances into their work, and their institutions are better for it. They learned about new evaluations because it is their job to think seriously about baseball. It is their duty to learn about the game and vote for awards after considering all the evidence. It is supposed to be their life’s work. But Hoynes casually dismisses them, because they don’t agree with him in the sanctity of a statistic that is so obviously flawed.
Last season, Greinke led the league in ERA. By a lot. He also led the league in FIP. He also led the league in WAR. He was really, really good—the best pitcher in baseball. Unfortunately, he played for the Royals. Who were really, really bad. In his career, Zack Greinke has made 165 starts. In those starts he has a career ERA of 3.77—better than all but one AL team this season. The Royals have gone 67-98 in those starts (.406). They stink. The fact that Greinke was drafted by the Royals and not the Yankees doesn’t make him any less good.
Same with Tim Lincecum: he led the league in WAR last year; led the league in FIP; led the league in strikeouts; was second in ERA.  The Giants had the lowest collective on-base percentage in baseball last season (.305) and managed to score only 655 runs—nearly 300 fewer runs than the Yankees; that paltry offense certainly cost Lincecum some pitcher-wins. But was that his fault?
And this season, voters will be confronted again with a choice between a pitcher with a lot of wins (CC Sabathia) and a pitcher who is statistically better in every way (Felix Hernandez).  That’s not even counting the performances of Cliff Lee or Francisco Liriano or the other half-dozen pitchers who have outpitched Sabathia but have had the crummy luck not to play for the only team who will score over 900 runs this season.
Just look:
Sabathia | Hernandez | |
IP | 217.0 | 225.2 |
ERA | 3.03 | 2.39 |
WHIP | 1.18 | 1.09 |
FIP | 3.56 | 3.00 |
xFIP | 3.81 | 3.25 |
K | 179 | 214 |
K/9 | 7.42 | 8.53 |
BB/9 | 2.82 | 2.51 |
K/BB | 2.63 | 3.40 |
WAR | 4.3 | 5.9 |
Hernandez has pitched more innings with a higher strikeout rate, fewer baserunners (WHIP), a lower ERA, a lower walk-rate, and a higher value (WAR). Hernandez is not arguably better than Sabathia. He IS better.
But Paul Hoynes thinks that even though CC Sabathia is a demonstrably worse pitcher, that he deserves the Cy Young because the Yankees have a great offense.
It is his job to know better.
32 Comments
Between Livingston, Hoynes, Grossi, and Shaw, I think we have by far the worst writers of any newspaper around…the only ones worth reading are Pluto and Lesmerises…the other 4 are just 4 grumpy old men that are skewed to hate the teams they cover rather than giving a concise reason why other than their opinion…
I actually cancelled my PD subscription because I was tired of those 4 blowhards…I’d rather just read the other 2 on the web where I dont have to even waste my time reading the other 4 names in the paper.
I believe you forgot to mention Mr Windhorst in your list of PD guys worth reading. Windy is the best in the business, IMO.
Well-written article and I agree with pretty much all of the defense of “new stats”. The main complaint I frequently hear about new stats boils down to “I don’t like what they say about me/my favorite player/my team”.
One very, very tiny nit: though: RBIs are just as much of a unicorn stat as pitcher-wins. But I suspect you already knew that when you mentioned ’em.
Yeah, I definitely don’t mean to besmirch the PD here. Simply put, as a sports fan, I hope the paper does well because it makes my life easier as a fan.
I just disagreed with Paul’s take, and wanted to point out why. That’s all.
And Tapin, I agree. I was trying to start small by mentioning RBI.
Good point on using more than just wins to evaluate how successful a pitcher is. I’m surprised there are still writers out there that cling to that one stat as the measuring stick.
Actually, since we’re talking about baseball here, I guess I’m really not that surprised.
In any event, good work!
Paul Hoynes’ job is to help the PD sell as many of its overpriced issues as he can. How can he do this? First of all, inform the reader. This has become less and less relevant thanks to the instant gratification of the online media. Thus, Hoynes relies on his second option: offer opinions that inspire lively debate. His take on the Cy Young has obviously done just that. It is presumptuous and simplistic to deem Hoynes foolish, as we actually have no idea what his personal thoughts are, only what he’s written to help move papers off the shelves.
@ #1
Ive never been a fan of Tony Grossi either, in fact I find him just about unreadable. Always have. Maybe its just me, but Ive always believed that he was a closet Denver Broncos fan. I hardly ever read PD articles anymore since I no longer live in the area, but Grossi always went out of his way to praise John Elway and Mike Shanahan, and rarely ever picked against Denver.
agreed on all points from Jon here.
also, an interesting note, for all the people that whine and complain about how the Indians traded CC and Lee…
Seattle, to start the season with a $100mil payroll and a combination of pitchers better than CC and Lee (Felix and Lee) still are going to finish with a similar record as the Indians this year.
Jon – Can you post or email me the firejomorgan.com version? That’s the one I really want to read.
Needless to say, good work. It must be hard not to bang your head against the keyboard all day when you read the Hoynes type of arguments
@SxDx – agreed…he doesnt seem to be much of a Cleveland fan, nor does it seem like he enjoys his job…maybe he’s just stuck here because the PD was dumb enough to hire him…
@Andrew – I agree…to an extent…I think he was too high on LBJ’s bandwagon at times, to where he didnt see the South Beach move coming…when you works with and cover someone repeatedly day in and day out, you should have some sort of idea what that person is like…but overall, i do like Windy…
*** work with…not works…
(good thing I’m not a writer, eh?)
Hey now…conflate is a big word. No fair!
@ Leftys72/6:
Mostly, I agree with you: when all else fails, be sensationalist!
But Hoynes actually gets to vote for the Cy Young from time to time–he’s a member of the BBWAA. We may find out soon (I don’t know if he has a Cy vote this season or whether he’d disclose it) whether he was baiting people or not.
Nice article. While he wasn’t a world beater or anything, I thought Kevin Milwood was really underrated that one year in Cleveland. He finished 9-11 with a 2.86 ERA in 2005, while Cliff Lee finished at 18-5 with a 3.79 ERA. I swear Millwood would lose a game 1-2 and the next day Lee would win a game 15-13.
And, the way that local media personalities appear to shun facts and logical discourse in favor of sensationalist commentary drives me absolutely nuts. It’s as frustrating as hell because they want you to send e-mails or call the station complaining. I wish the local AM stations would come up with something worth listening to.
I don’t agree with the Hoynes article and you bring up a lot of great points, but if I had a vote today for the Cy Young my list would be:
1. CC
2. Felix
3. David Price
I think wins for me isn’t the biggest stat, but it does have more weight in this case being that CC has 19 and Felix has 11. Out of CC’s 19 wins 18 have been quality starts. Yes Felix has 3 more quality starts overall compared to CC (27-24), but if Felix doesn’t win 15 games it’s going to be tough for him if CC ends up winning 21+.
I was happy to see Greinke win the Cy Young last year, because of everything he’s been through as a baseball player and a person and the fact that he had a ridiculous year. But it can be argued that Felix deserved the Cy Young just as much. Greinke did lead the AL in ERA, but it wasn’t by “a lot.” Greinke’s 2.16 ERA and Felix’s 2.49 ERA means that Greinke had and ERA of 1/3 of a run better per 9.
With that being said Jon, would your vote be for Felix over CC hands down because of the overall statistics and not quality wins and wins overall?
I think this race is still yet to be determined with the rest of the month to go. I also think it’s interesting and ridiculous that ESPN’s Cy Predictor doesn’t have Felix in their top 10, but that was a formula derived by Bill James and Rob Neyer.
http://espn.go.com/mlb/features/cyyoung
Jon, what do you think of Baseball Reference making changes to WAR and pushing CC to the top of the WAR list? http://www.baseball-reference.com/blog/archives/8264 Does this shift the argument a little toward CC or you? Personally it seems hard to believe CC is that much an improvement over Felix, if they switched teams.
@ Mike L:
I don’t love b-r’s WAR, but I have spent some time familiarizing myself with it lately. The changes they made here are of particular interest.
Basically, they said that that Yankee’s stadium is more hitter-friendly THIS YEAR, and therefore adjusted the stats of those players downward. Similarly, the found that SafeCo was more pitcher-friendly THIS YEAR than it has been previously. The problem (to me and others) is that parks aren’t generally different physical places from year to year. So you’re adding more randomness into the equation rather than removing it.
If you’re interested, read this: http://bit.ly/d0TJpU
But let me say one more thing before I become a CC-basher: he’s had an incredible year. he’s definitely one of the five best pitchers in the AL this season, and he could be top two. Hey, he could be the best. But it’s not because of pitcher-wins.
the adjustment for Yankee Stadium would make more sense than the adjustment for Safeco. I would think the sample size for YS is still pretty small overall to determine what it’s true mean should be.
Does anyone even bother to read guys like Livingston and Shaw anymore? While waiting for my sandwich to be made today at lunch, I begrudgingly read Shaw’s article on the Browns today. Ugh…Mary Worth or Judge Parker would’ve been more interesting. Here’s an excerpt:
“If I’m the head coach of a team coming off a 5-11 season, if I have an error-prone QB and a boss who hasn’t ruled out coaching again, I might be tempted to cue up Sinatra’s “My Way” after a couple more games like Sunday’s.
If I’m gonna die, I’d be tempted to do so with my infantry boots on.”
Hoynes is bad, but the rest of the jokers (except lesm, pluto, and BW) are worse.
Actually, BRANSON was the worst, but that goes without saying.
Wins and Losses are quite possibly the DUMBEST stat in baseball. You can pitch a 2-hit, 12 K shutout, but if one of your fielders drops an easy fly ball or throws a ground ball away, you can easily lose the game 2 to 0 while having a 0.00 ERA. An anemic offense and a poor defense are not any fault of the pitcher. And if you pitch amazingly through 6 innings, get taken out of the game, and then all your offense comes in the 7th and 8th, you don’t get a lick of credit, even if the relief pitcher gives up a couple runs, he’s gonna get the win. Foolish.
i always feel badly ripping the old school print guys.
take grossi. he learned from the best in the business (at the time), chuck heaton. consider: the browns have had two beat writers since 1954. so grossi may feel he owes more to maintenance of the sanctity of the position versus adapting to new reporting. i mean, would you rather have a woody paige or jay marriotti? i always feel badly for him (and mkc) when shefter or mort or @ravensinsider break browns news. i suspect that the arm’s length reporting method that was taught to him by heaton hurts in this day and age.
same applies to hoynes. he followed russ schnieder who’d been doing the tribe since the 50s (there was someone in between in the early 80s.. forget who). so hoynsie has been doing this for like 25 years and for 20 years his mandate was go be russ schnieder or sheldon ocker. now everything has changed. but aren’t we glad the PD funds a guy to report on-site at every game? it won’t be long till we just get AP news feeds and we’ll miss him then.
back to your point jon, i’m slower to dismiss his opinion. ‘wins’ are often determined by intangible grittiness. ie, who bears down with two on and none out in the 7th inning? guys like jack morris or orel hershiser or dennis martinez. only orel won a cy, but the point is that these are examples of guys whose WAR might not have looked great… but in situations like the above you’re not calling the bullpen to get ricardo rincon warmed up. know what i mean?
lol.. AND ANOTHER THING!! 🙂
haven’t moneyball and bill james been discredited? do you like the soulless redsox team that nerd/douche theo epstein assembled with an unlimited payroll budget? dump johnny damon, pick up jd drew; dump orlando cabrera, pick up edgar rentaria; bring in mike cameron, piss off jacoby ellsbury — all in the name of OBP and VORP. ultimately it is still a _team_ game.. not strat-o-matic.
i wonder who walt jocketty would vote for in AL Cy Young?
ok, more one thing and this is on the state of cleveland talk radio.
just as print media faces unprecedented competition from the web… so too will it be for brick/mortar radio. as wfny demonstrated last week, they can stream just like knr, just like cleve.com podcasts, just like espn. when everyone can stream, the game changes and the best content wins.
i for one would rather listen to rick/scott/craig et al over roda in a heartbeat. who wouldn’t? for that matter, i’d rather hear harv21, mr.c, stinfist, bode, robbie, FROWNS… over the content i pick up on wknr.
here’s hoping this comes to be. i bet within two years wfny will be giving wknr a run for its money.
@23
Schneider covered the Browns for a time in the late ’70s before Grossi came to the PD. And it was Pluto who covered the Indians in the early ’80s.
jimkanicki – your example of who bears down with 2 on, none out in the is a fallacy. We use selective memory when discussing that.
On Jack Morris – http://www.baseballprospectus.com/article.php?articleid=1815
And hopefully your next post is in complete jest. The first 7 years of the Epstein/James Red Sox have won an average of 94 games. The previous 7, the team averaged 87 wins. Epstein’s Sox have finished below 95 wins just once, a mark that the Red Sox have eclipsed just once in the previous 24 years.
RE: Baseball-Reference’s adjusted WAR and Hoynes’ theory
I agree that the idea that adjusting a park’s factor year-to-yea is an absolute joke. What it boils down to is B-R is allowing the home team’s collective +/- performance to effect the park factor which then in turn affects the player’s WAR score, which is supposed to allow them to stand apart from their great or horrible squad for measure/comparison. In doing these recalculations, it appears they’ve tossed out their saber baby with the Hoynes-bath water.
As has been pointed out umpteen times by eleventy-thousand blogs: the Yankees offense this year is GREAT. They score a TON of runs. Be they in “New Yankee” or away. And remember, they are guaranteed to be 1 of the teams playing in that bandbox in the Bronx every single game!
And Safeco and Felix is “losing out” — only in the opposite way. The Mariners are a horrible horrible offensive team this year. I mean, “Russell the Muscle” can only hit so many HRs — he’s got to get those strikeouts in.
The most glaring support to this train of thought is what’s happened to the field in St. Pete for the Rays. How exactly can Tropicana Field’s BPF and PFP change? It’s a freaking dome! What’s happened is the Rays’ pitching staff got good. Like Yankees’ hitters good. So now their individual merits will be undervalued because collectively they all got better.
It’s not hard to apply common sense to these mathemati-calisthenics . . . but you have to sometimes. Just glad I’m not an agent — I’d start trying to get my guys to avoid better teams to assure a bigger contract. Oh wait — never-mind! That might mean they’d go play for the Indians!
Guys,
I’m voting for AL MVP this year, not Cy Young. If I did have a Cy Young vote, Hernandez would not get my first-place vote. I understand the statistical arguments in Hernandez’s favor — at least most of them.
But no matter how poorly a team supports a starting pitcher, there is usually a time in a game where it comes down to making one pitch. If the pitcher makes it, the game may spin in a different direction. If he doesn’t, the outcome is predictable.
I still think wins are a big part of a starting pitcher’s resume.
paul hoynes
“But no matter how poorly a team supports a starting pitcher, there is usually a time in a game where it comes down to making one pitch. If the pitcher makes it, the game may spin in a different direction. If he doesn’t, the outcome is predictable.”
My goodness. There are concrete indicators of pitching performance. GSDD (Games Spun in Different Direction) is not one of them.
Your argument could easily be ever so slightly be re-worked into an argument for Bert Blyleven’s inclusion to the HOF. He had 14 seasons where he was among the best pitchers in baseball but sadly he didnt play for the Yankees or Red Sox, he was with the Twins, Pirates & Indians….teams without the big media outlets that hit nationally.
I mean WAR for pitchers has Blyleven as the 13th best pitcher of all time. OF ALL TIME. Its completely disgusting he’s being kept out of the HOF because his offense that he had no part of didnt score enough runs when he pitched. Think about how dumb it is that Paul Hoynes is using that as a logic stream to form his opinion. Its absurdist, its akin to saying that grass is purple not green. Its literally saying “well, because Chone Figgins was terrible at the plate this year, we cant honor Felix Hernandez pitching”. Its letting the team effect a singular player honor. Using logic like that is an insult to anyone with a splinter of intelligence.
Paul, times have changed man. Pitcher wins made sense before relief pitchers were the norm and they went 9 innings most of the time. Nowdays that is not the case. You simply cannot base a single-player award off what other players on his team did, and that’s exactly what you are doing.
Tyler Clippard was the 1st pitcher to 7 wins this year. You’d really give him the Cy if the season ended on that day? That’s ridiculously stupid.