Cliff Lee: Complete Games, Shutouts, And Cy Youngs
In his final start of August, Cliff Lee went 8 and 2/3 innings without allowing a run. Lee plunked Miguel Cairo with pitch number 117, cuing Ryan Madson to get the game’s final out. Last night against the Atlanta Braves, Lee finished where he started, using only 100 pitches en route to his 200th strikeout and 6th complete game and 6th shutout of the season.
Lee is now at 106 batters faced without allowing a run — 29 and 2/3 straight scoreless innings across 4 starts. Wow.
Despite striking out an uncharacteristically low number of batters (6), Lee instead trolled the Braves hitters by inducing 14 ground balls (second only to his present season high of 17 in his complete game against the Cardinals) and allowing nary a walk.
Last night’s shutout makes complete game number 6 for ol’ Cliff Lee, pushing his statistics down to: 2.47 ERA, 2.64 FIP, 2.76 xFIP, and a 2.67 SIERA.
Lee ranks 3rd in ERA, 3rd in FIP, 2nd in xFIP, and 3rd in SIERA. And he now leads the majors in shutouts with 6, ahead of James Shields (4) and Derek Holland (4). In the NL, it’s not even close:
When it comes to the 2011 NL Cy Young race, it presently comes down to just three fellas: Cliff Lee, Roy Halladay, and Clayton Kershaw. Yes, one could make the case for the likes of Madison Bumgarner, Matt Cain, Cole Hamels, and even Daniel Hudson and Matt Garza, but the Big Three are presently sporting Cy Young statistics, residing on a plateau of their own Manly Awesomeness.
Of course, anything can and will happen over the last few weeks, and a few crazy-good games could make this Cy Young race even closer, but for now, I imagine the voters are most closely watching these gents.
And frankly, among these twirlers, Doc Halladay stands the strongest. Not only does he lead the league in FIP (2.12), xFIP (2.61), tERA (2.39), and SIERA (2.54)*, but he has been the league’s most consistent starter.
*NOTE: Johnny Cueto leads the league in ERA… HAHAHAHAHAHAHAHA… HAHAHAHA, oh ERA, you’re so funny. *wipes away a tear*
How has Halladay, with not but even one shutout, been more consistent than Cliff Lee? Well, first of all, every one of his meaningful statistics have been lower. As we noted above, Lee is second and third in the league in all categories except shutouts, while Halladay and Kershaw typically fill the first and second position respectively.
Add to that truth bomb the fact that Roy Halladay has the lowest variability in ERA and FIP among the Top Three. Whereas Lee’s FIP has a standard deviation of 1.78 and his ERA a 2.07 standard deviation, Halladay stays cool with a 1.69 FIP standard deviation and a crazy 1.41 ERA st. dev.
Kershaw owns an impressive FIP 1.70 st. dev. and 1.999 ERA st. dev., but like everything else, he’s just woefully second to Halladay.
Visually, we can illustrate this distinction like so:
Looking at earned runs over the season, we can see how Halladay has been sneakily good (assuming there’s anything sneaky about utterly dominating a whole league) — Doc has managed to pitch super deep into games, but because he’s allowed just 1 or 2 runs, he’s missed any and all shutout chances.
In real terms, this is more valuable. Yeah, Lee pitching 6 shutouts guarantees a win in each of those games, but the Phillies won 9 to 0 last night. It’s more valuable across the season to have consistently fewer runs than being somewhat of a binge elite pitcher like Lee (which, in truth, may exaggerate his ability when he’s “off”; he’s only allowed 4 or more runs 7 times this year; Halladay, by contrast though, has only been tagged for such numbers 4 times).
Lee has continued to impress in 2011, but there’s no surprises in the NL Cy Young race: The Doc has it if he wants it.
UPDATE: Nobody, myself included, really likes that second chart. It’s a tough series to represent, what with overlapping values and three major data sets. Anyway, here’s the original version which may or may not be less eye-melting:
Two grammar/spelling errors in the -first paragraph-, “We can illustrate this distinct like so:” and “exagerate” in the second-to-last paragraph. I wasn’t even trying….sorry to now be one of those guys, but it seems to just get worse with every article.
Thanks, Sam. Sometimes they slip through; it’s unfortunate. :/
yea, you should of used a spell checker.
I hope this was sarcasm. Please tell me it was sarcasm.
Am I just stupid or is that a pretty awful graphic? What is going on with those colors…?
It certainly could have been done in a much better way.
Yeah, I’m with you guys. Please suggest a better way of representing it, and I’ll update it.
A cumulative line for each starter showing # of runs allowed. How many starts does Halladay have with <=2 runs allowed? How about Lee? Y-axis would be 0, 1 or fewer, 2 or fewer, etc, x-axis would be number of starts.
Ah! That’s where it’s at, Test. Gimme a moment.
I’m still trying to figure out why the graph has 7 different colors and the legend has only 3 different colors. Why are there four different shades of green?
Does solving this graph require an MFA in chromatics? What color results when you mix pink and light green? I have no idea.
I’m really quite fascinated.
Hmm… I can’t seem to get this concept to cooperate. My latest rendition makes it appear as though Halladay has allowed more runs than the other two, but each of the three pitchers are at 56 ER on the season.
I now suspect histograms for each pitcher would have been the best option, but who has the space to squeeze in three graphs into the second half of an article?
I thought pitchers did better for their teams, all else equal, when they concentrated a large fraction of their runs into a few “disaster starts” and were otherwise lights-out.
Agreed. And that is essentially what Halladay has done.
What the second graphic attempts to show (though it has a monumental task in that regard) is how Halladay’s starts have produced a smoother, lower distribution of runs allowed, whereas Lee has had more “disaster games” (relatively speaking) as well as more “lights out games.”
I think you need to read the comment you replied to again. Either Lee has more disaster starts (with essentially equivalent runs allowed, this is marginally better), or Halladay does. You seem to be saying both…
Oh, you’re right. My comment above should say:
“Disagreed. At least when it comes to these elite pitchers. A ‘disaster’ game for these guys is marginally much worse than a ‘normal’ game. One of their normal games by itself practically ensures a win (allowing only 2 runs at most), whereas as their disaster games give the other team a chance.
“The difference between 0 and 1 runs and 1 and 2 runs is much smaller than the difference between 3 and 4 runs and 5 and 6 runs, in other words.”
I think the avoidance of “disaster starts” is a different question than having a smooth distribution of runs allowed.
Pitcher A has one game giving up 0 runs in 9 innings, and one game giving up 4 runs in 5.7 innings.
Pitcher B has one game giving up 2 runs in 7.3 innings, and a second game also giving up 2 runs in 7.3 innings.
They’ve both averaged 7.3 innings/start and have a 2.45 ERA, which is almost exactly what Lee & Halladay have done this year. And neither have any “disaster starts”…. Assuming this is a year-long trend, which pitching style is more valuable?
I don’t know the SABR answer to this question. But Cy Young voters seem to prefer dominance over consistency.
@Yirmiyahu: I’d take Pitcher B any day. However, if you change the example up so that the first pitcher gives up 2 runs and 6 runs, and the second gives up 4 and 4, it becomes more interesting, and depends on the team he is pitching for, I think. In that situation, I’d take pitcher A if he was on the Giants, but pitcher B if he was on the Yankees. Obviously, these pitchers wouldn’t be in the Cy Young discussion, but I do think it makes a difference.
@Yirmiyahu
I did some calculations for the 2011 phillies using a tango distribution and the phillies’s bullpen’s siera by inning (based on the current bullpen). The opponent was an average team that would score 4.28 runs per game against an average pitcher. The phillies score 4.58 runs per game.
It turns out that the Phillies win 78.1% of the time with pitcher A and 76.5% of the time with pitcher B.
However, the Phillies have won 75% of games that Halladay has started and 68% of games that Lee has started this year.
As a sample size caveat, it’s just interesting to note that in just over 3.5% of Halladay’s starts, the pitcher of record is a utility infielder. It just points out how difficult it is to draw any meaningful conclusions from 28 starts, especially if you are trying to use winning (pitcher wins or team wins) as any type of measuring stick.
“I don’t know the SABR answer to this question. But Cy Young voters seem to prefer dominance over consistency.”
I rather see Lee pitch a complete-game shutout than Halladay pitch a 2 run complete-game. Shutouts are beautiful, and all-else being equal, I would vote for a pitcher with more shutouts such as Lee than a pitcher with a slightly lower FIP. I slightly prefer Lee over Halladay in the Cy Young voting.
Last pitcher with six complete-game shutouts in an MLB season is Randy Johnson 1998 with the Mariners and Astros
Is Doc a HOFer? How many more years of complete and utter dominance does he need to have for him to get in? Does he need a ring to get in?
At this rate, it seems like Halladay may pitch to or beyond age 40. If he makes it to 45, he could cross into 300 win territory. That would utterly endear him to the old school voters.
Of course, by then, most BBWAA voters will be former Fangraphs writers, so he’ll need a low FIP and a hardy WAR to pass muster. (Just kidding; I’m pretty sure Halladay is in by any nerd’s merit.)
Doc could get to 300 without pitching to 45.
And the Hall would be then be called “The FanGraphs Hall of fWAR – formerly The National Baseball Hall of Fame”. I’d imagine very player in the Hall would have their fWAR listed on their plaques, and the juicers would have their fWAR, and adjusted fWAR, assuming that they didn’t juice.
…one could only hope.
It’s universally agreed upon that he’s been the best pitcher in baseball the last decade. Assuming he doesn’t completely fall apart somehow over the next couple years, he’ll be a no doubt HOFer.
I’m not quite sure what “Hallady awesomeness” is.
It sounds like Halladay awesomeness, but it’s written sans an “a.”
Halladay awesomeness is an awesomeness akin to this mustache:
http://3putt.files.wordpress.com/2011/07/doc-holliday.jpg
Better awesomeness:
http://cdn.bleacherreport.net/images_root/slides/photos/001/252/912/the_worst_baseball_640_14_display_image.jpg?1314844176
You could probably post that picture anytime anything goes wrong with anything ever.
I think it would be more well represented if it was a simple line graph as opposed to this cumulative looking/overlapping thing you have going on.
That was my original thought, but such a chart gave me a nosebleed also. I’ll throw it up there for public consumption, but consider this the warning hazard.
I enjoyed the tone of the article – it made me chuckle. However, I don’t see how the FIP and ERA st. dev. are meaningful – much less statistically meaningful. Wouldn’t you want to use standard deviation of earned runs per start (or ER/IP per start or something similar)? Even then, “consistency” is a very intriguing term that has been discussed before, but never really described in this article.
What I’m trying to say is, how does the st. dev. information illuminate how Halladay has “it [the Cy Young award] if he wants it”?
In my mind, Lee’s ahead of the pack right now. The ESPN Cy Young predictor agrees:
http://espn.go.com/mlb/features/cyyoung
Oh, sorry if I didn’t clarify, but that is the st. dev. of ER per GS. I’m not sure if that clears anything up there.
The consistency issue is indeed a tricky one. For the sake of these elite guys, though, I think consistency is generally a better thing. Like I noted above, it’s very tough to lose games where the starters goes 7 and allows only 1 or 2 runs. The change in win probability is smaller if that’s 0 runs than if its 3 or 4 runs.
Yes, that does make more sense.
A pitcher who goes 9 innings every game and alternates between giving up 0 runs and 4 runs will win ~72% of games in the 2011 run environment. The pitcher who goes 9 innings every game and gives up 2 runs every game will win ~76% of the time.
However, Cy voters aren’t going to do calculations like that. Rather, they’ll just look at the W-L record (particularly just the # of wins) and call it a toss-up at this stage in that category.
Not sure I like the idea of evaluating a guy’s performance based on FIP and SIERA and similar measures. Yes, those are good ways to evaluate his true talent and predict future performance, but in the context of the Cy Young debate, I’m more comfortable looking at what the pitcher ACTUALLY did, as opposed to what he should have done based on his numbers. We can try say a person’s ERA should have been lower if various events beyond his control hadn’t happened, but we don’t KNOW that it would have mattered. Those stats are more helpful as tools for prediction, IMO.
So you’re a fan of a pitcher getting credit for his home park and defense when it comes to Cy Young awards?
But why would you want to evaluate a pitcher based on, as you say, events beyond his control. If they are beyond his control, we shouldn’t attribute them to him. FIP (and the like), are not just predictors, they incorporate, again as you say, what the pitcher actually did: Strikeouts, walks, and homeruns. A pitcher is not solely responsible for runs scoring or hits happening. FIP (et al) simply measures what the pitcher is solely responsible for so it’s perfectly fine to evaluate a pitcher using these tools, even for awards. That’s how I feel at least.
I am philosophically against evaluating a pitcher by his win-loss record because it is immensely influenced by his offense and to a lesser extent his bullpen, but I have little problem favoring a pitcher who might have a slightly higher FIP than another pitcher, but has a lower ERA. Essentially in this case, one would be rewarding that pitcher for luck (and park factors and defense).
I think most of the ERA variance among starting pitchers WHEN FIP or xFIP is controlled is caused by random variance. Defensive ability is certainly important when evaluating the value of individual players and does make a difference if the said player is extraordinarily good (or bad) defensively, but most variance in WAR among position players and teams is caused by variance in offensive production, not defense. For instance, according to the FanGraphs 2011 Leaderboard, the range among qualified position players for batting runs above average is 59.4 to -26.4 while fielding ranges from 22.6 to -22.3. (I’m too lazy to calculate the SD, but I’m sure that for individual players, fielding runs above average has a lower SD than batting runs).
This has been discussed a lot on this site. The issue is trying to distinguish what credit/fault goes to the pitcher and what goes to fielders.
Ultimately, what actually happened is reflected by runs/9 when that pitcher is on the mound. ERA is much worse because it includes arbitrary decisions made by a scorer and does that craptastic inning-reconstruction thing. Basically if you are basing value on ERA, then you are saying that the fielders only matter when they commit errors.
Tim L, FIP is what the pitcher ACTUALLY did.
Isn’t FIP what should’ve happened based on what the pitcher actually did (K/BB/HR)?
Not really. It’s just a number scaled to ERA. It’s not like, “ERA the pitcher should have had.” It’s just a number, and you compare it to the numbers of other pitchers.
It’s what a pitcher’s ERA should look like based off of his peripherals, or as it is put, ‘What he actually did’.
Well those 2 answers seem to contradict each other nicely.
I agree with you, Matt (you do have a great name, after all), but I think that is the problem with FIP. The fact that it is scaled to ERA is misleading because it does imply that FIP is what the pitcher’s ERA “should be” based on those three factors. Really, what FIP is, is simply a measure that combines Ks, BBs, and HRs with the right weights based on run values. Scaling it to ERA is just confusing.
From the Fangraphs Glossary: “(FIP) measures what a player’s ERA should have looked like over a give time period, assuming that performance on balls in play and timing were league average.” My name is also Matt, so I’ll just assume you were agreeing with me.
Touche, Santos. Well, I still think FIP should not be scaled to ERA if they’re going to use it in WAR. I also think that FIP ignores a lot of important aspects of pitching, but that’s a different issue (or is it?)
FIP is a weird compromise/hybrid between what actually happened and what should have happened. It incorporates some things that are mostly out of a pitchers control (HR/FB%, the BABIP effect on batters per inning), while excluding other similar things (LOB%, balls-in-play, stolen bases).
Runs per 9 innings is what actually happened.
Wouldn’t just runs allowed be what actually happened? And, is that really the best way to evaluate a pitcher. I mean, if you want to know what happened in the game, then sure, that’s fine to use, but if you want to know who is responsible for what happened in the game then I think there is too much noise in a stat like that. Personally, when I say “what actually happened” I’m not referring to the outcome necessarily, but rather, what actually happened that is attributable to a specific player. The general outcome (runs scored, wins/losses) is a combination of all players efforts. Stats like FIP assign responsibility to the individuals involved. It is still what actually happened, it’s just on a specific basis.
Is Hallady not ‘benefiting’ from a much lower than average HR/FB? And much lower than any he has previously had before. Lucky for him that this counts for FIP and thus WAR and thus prizes from the SABR community.
Maybe he took some lessons from Fangraphs favourite Matt Cain 🙂
Certainly any of the top SP would be worthy (Lee/Halladay/Kershaw/Hamels), but the fact remains that with the recent Cueto blowup, the NL Triple crown is in play (as it is in the AL) – and if someone nails the triple crown, then well I couldn’t care less if they have 1.5 WAR less than the leader.
Concerning the HR/FB issue: That’s why I didn’t hesitate to include xFIP, yet another stat in which he leads the league.
Frankly, all three of these pitchers have a bit of favorable (or favourable) luck in play, but that’s how Cy Young seasons typically go down.
Triple crown? You mean FIP, xFIP, and SIERA?
I haven’t done the number crunching, but I noticed Halladay’s got a great IFFB% this season, so his HR/OFFB% is not as giant of an improvement from previous years (though certainly still one).
Halladay/Lee/Kershaw all have very similar HR/FB% rates, but Halladay has clearly suffered the most in terms of BABIP. He allows the least amount of line drives, yet has the highest BABIP.
Very interesting.
Name | BABIP | LD% | GB% | FB% | IFFB% | OFFB% | HR/OFFB
Clayton Kershaw | 0.279 | 19.00% | 42.90% | 38.10% | 13.70% | 24.40% | 10.31%
Matt Cain | 0.262 | 19.70% | 40.70% | 39.60% | 10.80% | 28.80% | 4.95%
Roy Halladay | 0.310 | 18.40% | 50.70% | 30.90% | 16.00% | 14.90% | 10.58%
Ian Kennedy | 0.267 | 22.00% | 39.00% | 39.00% | 13.40% | 25.60% | 13.41%
Cliff Lee | 0.287 | 20.90% | 46.40% | 32.70% | 13.00% | 19.70% | 14.11%
For the record, I also lead in rWAR
That’s great Roy. I’m fourth this year in fly ball percentage.
That last graph is better, but I don’t see how it shows that Halladay is having a better season. Basically, Halladay had more starts allowing 3 ER or less, but Kershaw had more allowing 2 ER or less, and both Kershaw and Lee had more starts allowing 1 ER or less.
Also, could it be said that being a more “consistent” pitcher like Halladay is better on a team with a better offense, but having more dominant games (like Lee) is better on a team with a very bad offense? If a team is only averaging 3 runs a game, then I’d rather have a pitcher who gives up 1 or 0 most of the time but has bad games sometimes than a pitcher who gives up 3 runs every time. But if the offense scores 5 runs a game, then I’ll take the consistent pitcher every time. Obviously this is a simplistic example, but it seems pretty intuitive. Does it make sense to anyone else?
The second point makes excellent sense — and given Halladay’s context (the Phillies) it certainly makes an interesting explanation for his and Lee’s wins discrepancy.
One thing worth noting: Kershaw has one more start than the other two — and pitches in the cavernous Dodger’s stadium.
Really? No love for Jair Jurrjens for Cy Young? Do you not remember his first half numbers? He was unhittable!
(joking)
2010: Ubaldo Jiminez
2011: Jair Jurjjens
you mean the ubaldo jimenez who was only .3 WAR off halladay, a “lock” for the cy young?
I’m referring to his pre ASB ERA and BABIP from 2010.
Jiminez finished with a better second half (2010) than did Jurjjens (2011).
“striking out an uncharacteristically low number of batters (6)”
I wouldn’t call it uncharacteristically low. Lee struck out 20% of the batters he faced (6/30), which is a tick above his career rate (19.2%) and just a tick below his recent career rate (21.1% since 2008).
Yeah, it’s a little bit lower than his current season rate (25.4%), but not as drastically lower as K/9 (6.0 for this game vs. 9.0 for the season) makes it appear.
*Shameless Ian Kennedy name drop*
He won’t (and shouldn’t win), but he probably deserves an honorable mention for teh winz, and being pretty good under the radar.
That’s the part that bothers me about FIP … it ignores a component that spome pitchers have shown an ability to influence. We can form a long list of guys over the years that have BABIP against that are below league average or in the .290 to .300 range.
In 2011, the relevant example would be Ian kennedy. His BABIP against is .267 … which is exactly his career average. So, he supresses BABIP, but that doesn;t count for him. So his fWAR is 3.8 because it is assumed that he allows a .290-.300 BABIP … when he doesn’t.
Meanwhile, his brWAR is almost a full win better at 4.6.
Granted, all of this wouldn’t be such a big deal if we didn’t demand that we WAR for everything, and assume it’s precision (numbers to the right of the decimal), etc. We’ve all been in a discussion where we’ve stated that Player X is better or more valuable than Player Y because his WAR is 3.6 compared to 3.1 … even though those players are basically equals, considering the aspects of WAR/performance that fluctuate via single season data (or selection bias).
I think WAR should be THE major component of player vlauation and seasonal performance evaluation, etc … but not to the exclusion of rational thinking and consideration of other factors, or aspects that might cause WAR to represent an inaccurate scenario.
With “error bars” WAR does not tell us what we want to know. We don’t need WAR to tell us that Roy Halladay is much better than Kyle Lohse. We need WAR to tell us just how valuable those 2 guys are to their teams, and we need it to be more exact than just “He’s a 2-4 WAR player.” Given that range we’d almost be just as good with A, B, C, D, F letter grades for representing player performance.
A+ for Halladay!
Lohse gets a C-
“We’ve all been in a discussion where we’ve stated that Player X is better or more valuable than Player Y because his WAR is 3.6 compared to 3.1 … even though those players are basically equals,”
To be fair, doesn’t the same apply to ERA and batting average and OBP and SLG and HR and just about every number involved with baseball [or virtually any other subject involving numbers]?
Is a pitcher with a 3.50 ERA really that much better than a guy with a 3.80 ERA? Does a guy with 40 HRs necessarily have better power than a guy with 35? Etc…? Especially since we don’t even know about park and league affects with those?
That’s a very good point.
It’s probably more signant with WAR because it is viewed as a “Final Number”, rather than a component. By Final Number, I mean an accurate sum of all of the player’s contributions.
But, you do have a very good point in regards to other baseball statistics.
The interpretation of the second chart is totally wrong. The fact you see the red means exactly the opposite of supposed unappreciated awesomeness. It’s when the red curve is below the other two that it is showing Halladay is better. The points where the red curve sticks out above the other two is the completely appreciated conventional wisdom that Halladay has thrown fewer shutouts.
Also, the graph completely ignores innings in each start, which would seem to be kind of important.
Uh, this writer is forgetting Ian Kennedy as a Cy Young Candidate.
ian kennedy is a USA Today cy young candidate, not a fangraphs candidate
“Lee is now at 106 batters faced without allowing a run — 29 and 2/3 straight scoreless innings across 4 starts. Wow.”
Pfft….
Venter’s is (slightly) better than you, Craig. You just have higher status since you are the closer and get saves instead of holds.
WAR disagrees with you. Heavily.
I remember reading recently on FanGraphs that Venters has more shutdowns than Kimbrel. Shutdowns are supposedly the geek’s version of saves and holds.
And rely heavily on luck unless you K the side.
East Coast bias as usual (you might as well work for ESPN)!! Kershaw plays for a team with crappy offense and crappy defense (a blown double play by 3B Miles was cause for a three run 7th inning two nights ago after having pitches 6 scoreless with 9 Ks. Without looking at the circumstances, you think he gave up two earned runs in the 7th where any of the other top pitchers would have had the doulble play turned.
Kershaw is #1 or #2 in all of the stats that most people besides the “stat nerds” here on ESPN Fangraphs look at. You don’t even give that a mention. What a bunch of crap
And as for the “cavernous Dpdgers Stadium”. Also crap. The ball flys out just as much there for both teams and drops in as much in the alleys for both teams. There is no park advantage at Chavez Ravine.
http://www.andrewclem.com/Baseball/Stadiums_superimposed.html
Unless the rules of gravity are different at Chavez Ravine, then yes there is a park advantage for pitchers there.
Kershaw has significantly more strikeouts, a couple more innings pitched, a lower ERA, more Wins, a significantly lower batting average, OBP, and Slugging% against, and a lower WHIP (including significantly fewer hits allowed) than Roy Halladay. When you combine that with the fact that Kershaw plays for the Dodgers who not only are in a state of turmoil overall, but also are worse defensively than the Phillies at the important defensive positions, the case for Halladay having the Cy Young already isn’t as clear or pronounced as you make it out to be. Yes, Halladay has a lower FIP and xFIP than Halladay, which would seemingly give him the edge independent of the fact that he has better defensive support, but there is no way to actually account for the intangible factors that Halladay gains by playing for the better team all around, including the added benefit of pitching with the confidence that ground balls will be converted into outs and the added benefit and assurance of pitching behind a team that will most likely be able to at least score a few runs (which also allows him to not lose any confidence when he allows a run or two early in games, knowing that his team is capable of scoring and taking the lead at any time). Maybe these factors don’t really influence either pitcher’s performances, however, most pitchers do perform better with more run support and better defense. The Dodgers are playing defenders with negative UZRs at 2B, SS, and CF, with average defense at 3B, while the Phillies are relatively strong at those positions. Halladay has had a run support average of 6.00 compared to Kershaw’s 5.69, which may not seem like much, but considering how close their win loss records and ERA are, it definitely has some sort of impact. Moreover, while the difference in FIP between both pitchers is relatively large according to fangraphs calculations, Kershaw actually has a much lower Component ERA (CERA) than Halladay (a Bill James metric for calculating ERA independently of defense and luck using the number walks and hits allowed). Similarly, while the Defense-Independent ERA (DERA/DIPS) metric calculated using Voros McCracken’s formula shows Halladay having a lower dERA, the difference between the two pitchers is much smaller with Halladay leading the league at 2.36 and Kershaw a close second in the majors with a dERA of 2.62. Similarly, the Fair Run Average metric from BaseballProsepctus shows a narrow margin between the two. Kershaw also has a better average game score per outing (also a Bill James metric which awards points for good outcomes like strikeouts and completed innings and subtracts points for negative outcomes like hits and earned runs). Kershaw leads the majors with the highest Average Game Score, ahead of Lee (2nd) and Halladay (4th), (a metric by Bill James which calculated a score for each game started using 50 points as a baseline, while awarding points for positive outcomes like strikeouts and innings completed and subtracting points for negative outcomes like allowing earned runs). Lastly, while Halladay has a significant lead in WAR according to Fangraphs calculations, it is Kershaw who actually has a higher WARP (6.6) (wins above replacement player, calculated by BaseballProspectus) than (6.0), as well as a higher VORP (value of replacement player) at 61.4 compared to Halladay’s 57.6. Similarly, the rWAR calculations do show Halladay with the slim lead at 6.3 compated to Kershaw’s rWAR of 5.7, but also contend that the calculation includes defensive support, of which it calculated the Rdef (runs of support from defense) for Kershaw to be 0, while Halladay had Rdef of 1. Hence, when considering the other relevant information, some of which is unquantifiable, it isn’t as easy to name Halladay as the automatic Cy Young winner.
On another note, can someone explain to me how Halladay has a WAR of 7.4, compared to a WAR of 6.4 for Verlander and 6.2 for Kershaw? I understand fWAR is calculated using FIP, but is that the sole reason for the difference in WAR between Halladay and every other pitcher?
tl;dr
YDave:
“Lastly, while Halladay has a significant lead in WAR according to Fangraphs calculations, it is Kershaw who actually has a higher WARP (6.6) (wins above replacement player, calculated by BaseballProspectus) than (6.0)”
WARP includes offense and baserunning. If you look at PVORP (a more relevant stat in a Cy Young discussion), Halladay leads: http://www.baseballprospectus.com/sortable/index.php?cid=1012605
WARP also includes fielding, which I am okay at.
yea I stopped reading after “the added benefit of pitching with the confidence that ground balls will be converted into outs…”
For me, one of the things that separates Halladay from Kershaw is walks, and the various ratios that represent it.
-Strikeouts to Walk ratio (K/B)
Halladay: 7.50
Kershaw: 4.44
-Walks per 9 innings (BB/9)
Halladay: 1.15
Kershaw: 2.19
For me, that’s a big gap…especially considering how detrimental we generally view walks, in relation to performance. I don’t think, when attempting to directly compare two pitchers, that one can simply ignore the fact that Kershaw has walked nearly TWICE (!) as many hitters as Halladay.
Kershaw’s walk ratios are certainly very good, but there are 8-10 pitchers every year who post similar numbers. Halladay’s ratios are — without doubt — historic, on par with some of the best pitching seasons in the modern era.
In this current environment, I think we’ll find that batter walks are not nearly as productive as they once were, and likewise pitchers walking batters is not nearly as detrimental.
If you look at the leaderboards and select the lowest BB/9, you generally see good pitchers with lowish ERA/RA. When you select the higher BB/9, you still see some effective pitchers. They get decent K’s and/or don;t allow a buncha homers.
Strikeouts
Walks
Homers
Be effective in 2 of the 3 and you’re fine.
It seems that in this run environment a win is a little less than 10.5 runs. It seems closer to 9 runs per win (for example Jose Bautista currently has 74.3 RAR and a 7.9 WAR; the quotient is 9.4 runs per win). Do you suppose that there isn’t enough power in MLB to advanced walked batters, thus walks would not be as detrimental when batters had more power.
That’s exactly what I am suggesting.
Walks aren’t a huge deal provided you don’t give up a bunch of homers.
2-out walks likely have a very poor chance of scoring.
Yeah, ignoring Docs RIDICULOUS lack of walks really leaves out one of the biggest reasons why Doc is in the Cy Young discussion. You might as well leave out Kershaw’s SO total if you are going to leave out Doc’s walks.
Halladay’s stats would be awesome even with league average walk rates.
NL Ranks – Halladay
——————
K/9: 8th
HR/9: 3rd
IP: 3rd
NL Ranks – Kershaw
——————
K/9: 1st
HR/9: 6th
IP: 1st
It’s a pretty damn close race.
Halladay has 24 fewer walks allowed.
Kershaw has struck out 27 more batters.
Halladay has allowed 4 fewer home runs.
That’s a pretty amazing accomplishment for a 23yo pitcher.
“The Phillies team UZR/150 games is at -3.5 runs, whereas the Dodgers UZR/150 is 1.6 runs. The Defensive Runs Saved metric (DRS) gives the same data, as the Phillies are -30 as a team while the Dodgers are +2.”
Ugh, I guess that doesn’t shed a very nice light on me.
Great read though: http://www.thegoodphight.com/2011/9/9/2414142/the-nl-cy-young-race-an-extensive-statistical-perspective#storyjump