Cliff Lee: Counting Up From Zero

twitter.com/alexrolfe/status/217058748890488833

Above is a very interesting question Alex Rolfe asked me on Twitter during last night’s Cliff Lee start. Lee went seven innings, allowing five runs and once again failing to pick up his first win of the season. Lee has now allowed 14 runs in his last three starts spanning 20 innings, bumping his ERA all the way up to 3.72. Now all of Philadelphia is wondering: what’s wrong with Cliff Lee?

Amazingly, Cole Hamels — who started the other game in yesterday’s double-header — was in an eerily similar position as Lee back in 2009. Fans said many of the same things and asked many of the same questions of Hamels that are now surrounding Lee. Hamels had, in 2008, put himself on the map with a 3.09 regular season ERA and a World Series MVP award as the Phillies defeated the Tampa Bay Rays in five games for their first championship in 28 years. However, Hamels abruptly fell from grace, posting a 4.32 ERA and famously saying after a disastrous Game Three start against the New York Yankees in the World Series that he “can’t wait for it to end”. It, of course, referring to his terrible year. The Phillies lost in six games. Fans and media types spent time in the off-season dissecting Hamels’ season and wondering how they could ship him to Toronto for Roy Halladay.

At the same time, the stat-savvy among us pleaded for caution, claiming that Hamels’ ’09 was merely a fluke and his 2010 season would likely be much better. Myself, Matt Swartz, Paul Boye, and many more penned articles expressing this sentiment. These articles were met with skepticism (as any topic should) and disapproval, many claiming that the player we had written about was not the same player fans and media types had watched. To briefly recap the main points in favor of a Hamels rebound: the only difference between his 2008 and ’09 seasons was BABIP. His strikeout and walk rates were nearly identical, as were his batted ball splits, and he wasn’t allowing any more home runs. However, Hamels’ .259 BABIP in ’08 had risen to .317 in ’09, explaining much of the discrepancy.

As expected, Hamels rebounded in a big way in 2010. In fact, he not only regressed back to his mean, but he improved significantly. Hamels struck out more batters and induced more ground balls than ever before, and he added another pitch to his arsenal: a cut fastball. He finished the year with a 3.06 ERA, reestablishing himself as one of the best pitchers in baseball and giving the Phillies a daunting starting rotation that also included Halladay, Lee, and Roy Oswalt. Fans who had once stormed the castle with torches in hand in protest of Hamels’ continued presence had grown to love and accept him as one of their own.

On June 25, 2012, Philadelphia is on the verge of repeating  history. Lee is having a season not unlike Hamels’ 2009. Prior to his last two starts — both seven-inning, 5 ER affairs — a legitimate non-Sabermetric case could have been made placing Lee in a very preliminary Cy Young discussion. Despite having zero wins to his name, he had a 3.18 ERA, quality starts in seven of nine starts, and one memorable no decision against the San Francisco Giants. He had struck out 66 and walked merely 11 in 64.2 innings of work.

Then came the narrative train. No wins and three bad starts makes Phillies fans something something. Go crazy? Don’t mind if they do! Repeating the same cliches foisted upon Hamels back in ’09, the narrative has centered on Cliff Lee’s “struggles”. And yet, when you look at the stats — the same ones that controversially told us Hamels would rebound in 2010 — Lee is actually pitching quite well.

K% BB% GB% xFIP SIERA
2011 25.9% 4.6% 46.3% 2.68 2.72
2012 25.5% 4.6% 49.3% 2.88 2.75

There is one difference between Hamels and Lee, however: home runs. For Hamels, there was no jump in long balls surrendered, but for Lee, he has allowed nine home runs in 77.2 innings compared to 18 in 232.2 innings last year. On a per-fly ball basis, Lee jumped from nine percent last year and eight percent over his career to 13 percent so far in 2012, excluding yesterday’s start. If there are more home runs, then Lee has to be struggling with location, right? Doesn’t seem like that’s the case.

What about just fly balls? Again, not really a difference.

Let’s narrow it down to just home runs.

He is not being punished in any one particular area and, as the charts above show, he isn’t peppering a specific location where he shouldn’t be. Eight of his ten home runs allowed have been hit by right-handed hitters, compared to 79 percent over his career. And Lee’s pitches have been accessory to a home run in nearly equal portions: three fastballs, two cutters, two change-ups, and one curve. By the way, there has also been no meaningful change in his pitch usage.

We can narrow down the damage to Lee’s fastballs: sinkers and cutters. Opposing batters have mustered a meager .234 wOBA against his “soft” stuff this year and right-handed hitters are doing most of the damage (.341 wOBA compared to lefties’ .292 on “hard” stuff). Yet Lee is still doing his job against them as his 3.14 xFIP and 5-to-1 strikeout-to-walk ratio illustrate (career averages: 3.93, 3.2) and there is no pattern. Surprisingly, lefties have a .367 BABIP compared to the .301 of right-handers, including .356 on “hard” stuff. The distribution of hits, though, doesn’t scream “this guy is getting shelled!”

There is not one performance-based metric that is alarming regarding Cliff Lee. Yes, it is June 25 and he has zero wins and his ERA is just under 4.00, but such is life in small samples. Two weeks ago, there was nothing wrong with him and the narrative simply focused on his zero wins to date. As a fan, there is no fault in being frustrated with Lee’s last three starts, but when the irrationality of being a fan mixes with objective analysis, there is a problem, and it’s the same problem that arose in ’09 with Hamels. Performance and results are not always related. In some cases, like Roy Halladay in 2010, they match up perfectly, but other times, like Hamels in ’09, James Shields in ’10, and Lee this year, they are at odds.

Maybe, instead of hyper-analyzing Lee, we should be hyper-analyzing the traditional stats with which we evaluate him and others. Even if Lee’s 3.72 ERA was indicative of his performance thus far, his 0-4 record into late June should give us pause as to exactly why we continue to include this statistic in our conversations. Lee is the only pitcher this year to have thrown 50 innings and not earn a win. Is he the worst pitcher in baseball? Is he anywhere close to it? No. Then why use W-L ever, at all? Lower the threshold to 40 innings and Chris Volstad joins the party with his stunning 7.46 ERA. That tells you all you need to know about traditional statistics and their place in the conversation, especially when narratives are involved.

Cliff Lee has been really, really good this year and it’s important that Phillies fans realize this in what has otherwise been a very disappointing season.

Leave a Reply

*

31 comments

  1. BradInDC

    June 25, 2012 08:45 AM

    Some of the same people who say he stinks would also advocate things like trading him for Mike Trout. He is both worthless, and somehow worth a cost-controlled Über prospect/rookie in return. I haven’t heard anyone who matters say anything like that yet. But I imagine there are people thinking it. That’s my narrative for the day.

  2. Jim Z.

    June 25, 2012 10:21 AM

    I know that this blog is a temple to the fetishization of statistics, but let’s throw some of these numbers aside and reflect on the undeniable fact that Lee: a.) had the lead in all of his last three loses and b.) had the opportunity to register outs and failed to do so. So his SIERA / FIP are respectable – the fact remains, however, that he did not execute when he had the opportunity to do so in these last three starts.

    Here’s where you need to step outside of the stats box and say “He simply needs to get more of those crucial outs and stop giving up HRs at inopportune times and maybe he will, you know, actually win a game.”

  3. Bill Baer

    June 25, 2012 10:48 AM

    So, you’re basically saying that players have the ability to time out when they do or do not give up hits. Interesting. Lee should definitely stop intentionally forking over leads then. Sabermetrics is a sham.

  4. Bxe1234

    June 25, 2012 11:15 AM

    I can’t believe Cliff Lee decided not to pitch a perfect game yesterday.

  5. DS

    June 25, 2012 11:52 AM

    19% line drive rate last year, 16% this year. He makes Meg Griffin look lucky.

  6. topherstarr

    June 25, 2012 11:56 AM

    Bill, your appraisal isn’t all that different from Charlie Manuel’s (although maybe you could use a little more, ya know like, folksy charm):

    “It’s hard to figure out,” Manuel said. “He’s throwing good. He had nine strikeouts. He hit a streak there after the second inning where he looked super. It’s not like his stuff is not there, because it is. His stuff is there.”

  7. Noah

    June 25, 2012 12:20 PM

    Lee’s fine. Phillies not so sure.

  8. Jim Z.

    June 25, 2012 12:52 PM

    Bill, so what you’re basically saying is that players’ future performance is enslaved to their statistical past?

    There is no free will? There is no ability for individual players to sharpen their focus and intensify their effort in a specific situation? Are pitchers robots that perform uniformly according to their statistical profile and that cannot consciously elevate the quality of their play in a dire moment?

    If so, then sports as a whole is a sham that deserves nothing more than to be played out on an Excel spreadsheet. Let’s pull the plug on this elaborate sham and save alot of money by relegating every MLB season from here on out to the machinations of computer-based number crunching applications.

  9. Richard

    June 25, 2012 01:08 PM

    No, Jim. The point is that the statistical evidence, which is really all we have to go on, tells us that Lee is pitching about as well as he did last year. And you can sharpen your focus all you want, but that’s not going to result in perfection, won’t mean that all good pitches result in outs, won’t even mean all good pitches.

    Do you think he’s not focused with guys on base, that he doesn’t care enough to do so and is just flopping up bad pitches as a result? Does that sound like Cliff Lee?

    If he had recorded wins earlier in the year (say, the great 10-inning game against SF, or the two NDs he had, where the Phillies came back to win, or the first game off the DL, when he pitched well, but was on a pitch-count, left with a 2-run lead, only to see the bullpen blow it…) if he’d had say 4 or 5 wins already, prior to these last three games, do you think the prevailing narrative would be “what’s wrong with Cliff Lee?”

  10. JRFarmer

    June 25, 2012 02:33 PM

    >> There is no ability for individual players
    >> to sharpen their focus and intensify their
    >> effort in a specific situation?

    Methinks Jim Z. is a believer in “clutch” players.

  11. Greendale

    June 25, 2012 02:40 PM

    At least you didn’t say he was “cy young caliber all year” like you did last night on twitter.

  12. Jim Z.

    June 25, 2012 03:13 PM

    Statistics are not predictive, they are explanatory.

    A pitcher’s SIERA/FIP describe a body of work that encompasses any fluctuations in performance, focus, or effort. But that does not mean that they dictate the parameters of any particular set of individual future performances, only that they will give a guideline to describe an expanding body of work over time. A pitcher may, in fact, be more focused and effective in a particular game, or even an inning, than any other, and this may be, in fact, under his control, based on external and internal factors that affect performance. I am a tournament-caliber billiards player, and while you may devise some metrics that describe me as a player over the course of my career, I know that on certain days and in certain moments I played “better” than others, and much of that was due to circumstances that were within my control: i.e. I was mentally harried after having an argument with my girlfriend before one touranment and played poorly; I was distracted by some stray thoughts in another and played poorly; I was going up against a superior opponent and willed myself to a performance that was typically beyond my ability based on the sheer challenge of playing against such an opponent, which released adrenalin into my system and boosted my confidence.

    I would be shocked to discover that these situations never occur in professional athletes playing more popular sports such as baseball.

    In the case of baseball, is it too far-fetched to believe that Cliff Lee’s SIERA and FIP describe a body of work filled with both externally and internally affected (both negatively and positively) individual performances, and that these performances could be replicated in the future to fit along the guidelines of his current SIERA/FIP? Why can’t he have a “great” game in this past stretch of games? Why are they all “bad” or “average” games? Why can’t he “will” himself to a stellar, 1-hit performance knowing that he has a 2+ run lead? There is nothing about adherence to SIERA or FIP that precludes such a thing happening, as long as the ensuing body of work fits the guidelines set forth by his SIERA/FIP profile.

  13. Richard

    June 25, 2012 03:30 PM

    “Why can’t he “will” himself to a stellar, 1-hit performance knowing that he has a 2+ run lead?”

    Because there are other players involved, for one thing, his own fielders and the opposing team’s hitters among them.

    It’s certainly on Lee that he fell behind 3-0 to Conrad, for example (though I wasn’t watching, how close were the pitches?), so maybe he had to come in with a too hittable pitch, or the pitch he threw wasn’t exactly where he wanted it, whatever. But let’s say instead of Conrad hitting his double, he pops it up (hitters pop up meatballs all the time), or he takes it (because he doesn’t have the green light, or it wasn’t what he was looking for, or he sucks), so it’s only a called strike, or he hits it just as hard, but in the general direction of a fielder, who makes an easy out of it… these kinds of results happen all the time, but we don’t remember them, in part because we give the pitcher credit for getting out of a jam. Or we do remember them, when it’s the Phillies offense doing it, rather than giving the opposing pitcher (or defense) some credit. Etc.

  14. Bill Baer

    June 25, 2012 03:52 PM

    “Professional billiards player here,

    Math is stupid and I am an unbiased observer of my own experiences.

    Well, bye.”

  15. LTG

    June 25, 2012 05:34 PM

    “Statistics are not predictive, they are explanatory.”

    Funny that one finds the precise opposite criticism of statistical analysis in other places. That is, “Statistics are predictive but they don’t explain anything.” I would have thought this is closer to the truth too, although still not true. For, we can flip a coin lots and lots of times and become convinced that our data will be reproduced through future collections of coin flips, but have no idea why coin flips in large sets produce approximately equal amounts of heads and tails. We could predict the future flip-results on the basis of statistics, but not know that the 50/50 split is due to the construction of the coin (two-sides, balanced weighting, etc.)

    Certainly the coin-flip stats won’t predict any particular flip because the stats say 50/50 and each flip is 1/0 or 0/1. But then no one thinks stats are predictive at the level of individual tests (whose analogy to pitching could be pitches, BFs, innings, or even games, depending on what the statistic measures).

    It is also worth wondering whether predictive and explanatory are right opposed. Predictive is often opposed to descriptive, but even these strike me as related in important ways. But ‘explanation’ is an ambiguous concept. Often, we take ourselves to have an explanation if we can predict what will happen on the basis of the explanation. And it is hard to imagine that a good explanation would not allow us to predict what will happen when the explanatory conditions are met.

  16. Mike

    June 25, 2012 06:50 PM

    Is there a chance that both sides could be correct with this one – that Cliff Lee is having an “off” year and that the math can prove it. The problem, however, is that the current set of metrics does not support that theory. Therefore, how about this one – does a reasonable chance exist that a currently unknown metric, if formulated, would shine some light on this disagreement? My intuition tells me that we have not exhausted every single possible good explanatory metric, but I also have no idea what that metric would look like (and it’s pointless for me to provide some examples because they will be quickly shot down and my larger point will be missed).

  17. Richard

    June 25, 2012 07:23 PM

    “Is there a chance that both sides could be correct with this one – that Cliff Lee is having an “off” year and that the math can prove it.”

    I think you could argue he’s had an off stretch of, what, three games? (He has walked a few more than usual, for instance, in those games.) But the season is still early enough that three rough games, results-wise, can have a lot of impact on other numbers (like ERA). But his basic numbers, in particular his strikeout rate and overall BB rate, tell us his results should even out as the season goes on.

    However, prior to these last three starts, no argument whatsoever could be made that he was having anything like an “off” year.

  18. LTG

    June 25, 2012 08:00 PM

    Mike, your point is good at the meta-discourse level but not at the discourse level.

    You give us all (especially sabermetricians) a reason to be sensitive to reasonable disagreement when plausible alternatives are presented and to keep our credences in our stat-based conclusions from getting too high.

    But you don’t give us a reason to simply suspend all judgment. Compare the structure of what you say to the natural sciences. Given the history of revolutions in physics we have a reason to believe that there are phenomena yet undiscovered that would lead us to significantly revise the current amalgam of theories and, therefore, revise our beliefs about what the natural world is like. But we shouldn’t suspend all of our beliefs about the natural world for that reason. We should just be open to the need to revise when plausible alternatives offer solutions to intransigent problems.

    So, in the end, unless a plausible alternative account of Cliff Lee’s struggles is offered, we should not abandon the best one available.

  19. Frank Reynolds

    June 25, 2012 10:25 PM

    I read tend to agree with Bill. Good article. However, I would like to see him get a win soon so we don’t have to hear about it anymore. I don’t think wins are a good stat when evaluating pitchers. I would like our starting pitchers to get as many wins as possible because that means the entire team is winning the game.

  20. Mike

    June 26, 2012 09:17 AM

    First, let me make it clear where I stand on this issue:

    * I like Cliff Lee. He is an elite pitcher. I was ecstatic when he signed with the Phillies and very much want him to continue with the Phillies.
    * I would make exactly the same claims if Lee had a 3-4 record (with wins against the Astros on 5/15 and the Blue Jays on 6/16 when the bullpen blew late-inning leads and a win with his gem against the Giants on 4/18).

    Now, having said that, Lee has started 12 games. Here is the breakdown:

    * In 7 games, Lee pitched with a lead. In only 1 of those games did Lee leave with a lead (in the 6/16 game with Blue Jays, Lee left with a lead, but with the tying run on base and that runner eventually scored).
    * In 4 games, Lee never had a lead. He left the game on 4/13 down 4-1, left the game on 5/20 down 5-1, left the game on 5/30 down 3-1, and left the game on 6/24 down 5-3.
    * In 1 game, Lee always pitched in a tie situation (Giants, 4/18).

    Therefore, in only 1 out of his 12 starts this season, Lee left the game with a lead. So, is Lee pitching fine and is just unlucky or is he really not pitching as well as he pitched in 2011?

    Do his HR stats play a role in any of this? For example, Lee allowed 0.7 HR/9 in 2011 and 1.0 HR/9 in 2012 (not as good as 2011 but also not alarming). Maybe the issue is *when* he is allowing these HRs. Here is the breakdown of his 9 HRs this year just before the HR was hit (I realize this is a small sample size but this is really all we have to go on):

    * 3 times the Phillies were winning and the HR tied the game
    * 3 times the Phillies were losing
    * 2 times the Phillies were tied

    Therefore, the Phillies were leading before and after the home run for only 1 of his 9 HRs allowed. I would consider his other 8 HRs allowed to be “impact” HRs.

    Although Lee has K% and BB% numbers that are comparable between 2012 and 2011 maybe his “impact” HRs can help to differentiate those two seasons. If we had the choice, I think we can all agree that we would want the Phillies to have the best win% in the league, not the best K% and BB%.

  21. Richard

    June 26, 2012 09:36 AM

    Well, I know at least one other game Lee left with a lead: his first post-DL game, against the Mets, when the bullpen blew a 4-2 lead (as the bullpen blew every game in that series).

    Two other games he left after excellent starts with the game tied, but it only got tied because of some bad-to-weird defense (Astros, Cardinals).

    Anyway, no one is saying they’d rather have the best K% & BB% as opposed to team wins. The point–difficult to understand, I gather!–is that those two numbers not only ought to but generally do go with team wins.

  22. Anthony

    June 26, 2012 10:46 AM

    5/23

    Halladay is fine.

    Lee is fine, too. :-)

    You can’t be sabr guys and pretend HR/FB is a fluke. It’s one of the 3 main indicators of future performance! Why would that not indicate a loss in stuff, especially since placement appears similar?

  23. Richard

    June 26, 2012 11:17 AM

    “It’s one of the 3 main indicators of future performance!”

    um, no it’s not (cf. xFIP)

    (for the record, I didn’t think Halladay was “fine”, but I can’t say I knew what was up… when it turned out he was injured, it was actually kind of a relief)

  24. BobSmith75

    June 29, 2012 08:03 PM

    Bill – He was on the DL in April with a left oblique strain & came back earlier than expected. Those have a tendency to linger for a while without prolonged period of rest for 4-6 weeks.

  25. BobSmith75

    June 29, 2012 08:08 PM

    Curious to see how his numbers look since he came back off the DL. I imagine there is much difference to previous year’s either but just curious because he doesn’t seem like he has been quite the same pitcher since the SF outing.

Next ArticleRaul Valdes Now An Important Part of Phillies Bullpen