Cole, Don’t Change A Thing

(Well, there’s a couple things you could change…)

Another day, another poor start for Cole Hamels. Following what seemed to be a turn-around performance against the Florida Marlins last Sunday, Cole allowed six runs on four home runs in six innings, despite striking out seven batters and walking only one last night against the Arizona Diamondbacks.

The first three innings went very smoothly. In the first inning, Cole retired the side in order on three ground ball outs. He then struck out the side in the second, allowing only a single. Cole retired the D-Backs 1-2-3 again in the third with one ground out, fly out, and strikeout apiece. In case you’re counting, that’s four ground outs, four strikeouts, and one fly out — a great distribution of outs.

Cole unraveled in the fourth inning, struggling with location. With a runner on first with one out and Mark Reynolds (a.k.a. the right-handed Adam Dunn) at the plate, Hamels threw a four-seam fastball that was letter-high right down the middle. Reynolds fouled it off. Hamels came right back to the same exact location with a cutter, which Reynolds promptly deposited well beyond the fence in left field.

Adam LaRoche then worked the count to 3-2 after six pitches and three foul balls. All six pitches were at least waist-high, and so was the seventh, a waist-high change-up right down the middle. It was not surprising to see LaRoche — who routinely hits Phillies pitching — hit a home run down the right field line.

Hamels allowed a single to Chris Young before striking out Cole Gillespie, bringing up the D-Backs’ #8 hitter in catcher Chris Snyder. Snyder only needed one pitch — a knee-high cutter right down the middle — to notch the Snakes’ third home run of the inning, capping a five-run fourth.

The batted balls in the fourth went: fly out, soft line drive single, fly ball home run, fly ball home run, line drive single, strikeout, home run, ground out. Two line drives, three fly balls, one ground out, and one strikeout.

Kelly Johnson led off the bottom of the fifth with a solo home run off of Hamels that increased the D-Backs’ lead to four runs at 6-2. It wasn’t really a bad pitch by Hamels and more so a good piece of hitting by Johnson.

The batted balls in the fifth went: fly ball home run, fly out, line drive single, fly out, fly out. Cole added another two strikeouts and a fly out in his sixth and final inning.

All told, Hamels allowed ten fly balls (including three line drives) and six ground outs (44% fly balls,  19% line drives, 38% ground balls) along with the seven strikeouts, giving him a SIERA of 2.74 for the game as opposed to the 9.00 ERA. Of the ten fly balls, four were home runs, giving him a HR/FB% of 40%. We know that pitchers can’t control how many home runs they allow other than by controlling their rate of fly balls. While Hamels clearly wasn’t at his best last night, it also will not be par for the course — he will not have a 40% HR/FB rate every game.

That said, Hamels’ pitch selection was peculiar, as the following graph will illustrate:

His fastball and curve use has increased and his change-up use has decreased in each start. In other words, between his first and most recent start, Hamels has decreased the use of his best pitch by over 26% in favor of lesser quality pitches. While he has utilized his cutter in his last three starts, he is doing so at the expense of his change-up and that is not a winning strategy.

Still, Hamels has been unlucky. His 5.11 ERA is much higher than his retrodicted 3.13 SIERA. While he has been more BABIP lucky (.275), his HR/FB% (30.4%) is about three times higher than it should be. Meanwhile, his strikeout and walk rates are great at 9.5 and 2.2 respectively — a better than four-to-one ratio.

If Cole wants to get back on the winning track, he doesn’t need to change much — he just needs to ride out yet another wave of bad luck, be a little more precise with his location, and to stop using his other pitches at the expense of his change-up. That’s really it. Based on events proven to be within a pitcher’s control — strikeouts, walks, and GB/FB rates — he has pitched very well. With a few minor tweaks, he can put himself in a better position where he won’t be resting his fate on rolls of the dice.

Leave a Reply

*

59 comments

  1. Peter Hjort

    April 24, 2010 03:32 AM

    Kelly Johnson can be excluded, as the dude is super-human, and can’t be discussed in the same conversation as mere mortals :)

  2. gm-carson

    April 24, 2010 05:37 AM

    Really, we’re still going with the “unlucky” excuse?

    I hope you’re right Bill and Cole makes his “minor adjustments” and gets back to his 2008 form.

  3. Bill Baer

    April 24, 2010 05:42 AM

    Carson, it’s not an excuse. It has been statistically proven that pitchers control some factors and barely (if at all) control others. In other words, we can identify the factors that are symptomatic of good and bad pitching (strikeouts, walks, and ground ball/fly ball rates) and separate them from the non-factors (hits allowed per ball in play and home run per fly ball rate).

    There is no bias here — this is what the facts state.

    Others can choose to accept these facts or ignore them and continue to irrationally disapprove of Cole Hamels.

  4. gm-carson

    April 24, 2010 05:47 AM

    Guess I’m irrational. I look at results and the results right now show me a high ERA, WHIP, and homerun total. I truly am trying to become better statistically rounded in my view of the game, but in this case I just don’t buy the “unlucky” argument.

  5. Bill Baer

    April 24, 2010 05:58 AM

    I don’t mean to be insulting when I say “continue to irrationally disapprove of Cole Hamels.”

    But I do mean to say that this is not debatable, at least in the sense that most people are going to argue this. If you can disprove the findings that pitchers cannot control BABIP and HR/FB and that K/9, BB/9, and GB/FB rates are controllable by a pitcher, then the current anti-Hamels arguments may have merit.

    You can dislike him all you want and you can be extremely frustrated with his results so far, but by no means is the “ERA, WHIP, and homerun total” on any kind of equal footing with the evidence posted above, at least on an intellectual level.

    Simply put, you are looking at the wrong statistics. I know that statement will rub some non-stat people the wrong way but there simply is right and wrong analysis.

  6. TechniclySporty

    April 24, 2010 07:18 AM

    Ok, you had me right until the “intellectually honest” part. The stats you sight are a result of how Cole pitches. They don’t indicate how he threw but what the results were, so you are right, and you’re wrong. You can’t argue what the stats are, but you can argue how the stats were create. You have to agree, all fly balls are not the same. You can have a jam job that turns into a mile high pop out and a belt high fast ball deposited 20 rows deep. An off the end of the bat can-of-corn is not the same as even a warning track fly. Those results are the same, but they would be different levels of power and probably pitched.

    Three home runs in one inning is not unluky, it’s poor control. You said it yourself, A cutter to Reynolds in the same location, no change of eye level, a 3-2 change to LaRoche was the second homer. Let’s call it what it is a lack of control.

    Now, I don’t “irationally dissaprove” Cole. I like him, and I think he will get it back in the next few starts, but he will lose focus every now and then and lose control in the process. Let’s not get on non-stat geeks, like myself, and say we look at the wrong stats. I hate stats. Stats can be deceiving. My sole concern are results.

  7. Bill Baer

    April 24, 2010 07:26 AM

    You have to agree, all fly balls are not the same. You can have a jam job that turns into a mile high pop out and a belt high fast ball deposited 20 rows deep. An off the end of the bat can-of-corn is not the same as even a warning track fly.

    I do agree. However, to quote Matt Swartz:

    [...] pop-up rate was allowed to negatively affect SIERA because it is a symptom of the pitcher throwing the ball that generates an upward trajectory, which could lead to an increase in home runs. A pitcher’s skills are throwing strikes, making hitters miss, and throwing with angles and spins such that the trajectory of the ball is downward when it hits the bat. A popup almost always represents an out, but it also represents a potential problem for the pitcher in the future.

  8. loctastic

    April 24, 2010 07:30 AM

    I appreciate the analysis in this article, I suspected he wasn’t throwing the changeup enough. I feel like he’s still tinkering with the cutter. He should probably stop messing with it until next year at this point, because all that pitch does is get hit hard.

    I’m not sure why you think he’s still unlucky. 4 homers in a game is a result of bad location.. those weren’t gimme shots either, they looked gone right off the bat.

    I agree with your eventual conclusion that his location must improve and he needs to get back to throwing the change more. But that’s not anything to do with luck.. if you throw a fastball or a cutter over the middle of the plate, you’re going to be lucky if they DON’T hit it out.

  9. DrPete

    April 24, 2010 09:21 AM

    Following a comment by my wife, I noted that 4 of the Phillies 6 losses are with Schneider behind the plate — and all 4 starters in those games have been hit hard. (The other 2 were the Kendrick and Hamels gems.)
    How big of an effect does the catcher have on pitch selection and location?

  10. Bill Baer

    April 24, 2010 10:13 AM

    I don’t think there’s any way to know given the information at present, DrPete. Sabermetricians have studied catcher’s ERA (CERA) and haven’t found that it is an actual skill.

    At any rate, we need a much larger sample size than four games. And we would need a lot of qualitative information to support any findings.

  11. Jacob

    April 24, 2010 11:12 AM

    Bill,

    You really need to get off your knees when writing these posts about Cole Hamels. He’s obviously regressed and you might be the only person in the Tri-State Area who is too stubborn to admit it.

  12. Peter Hjort

    April 24, 2010 11:46 AM

    Carson,
    Results based analysis is a crappy place to start. Results fluctuate a great deal more than a pitcher’s fundamental skills.

    There are a million and a half ways for a pitcher’s ERA to balloon after a few starts, most have nothing to do with anything the pitcher can control.

  13. Matt

    April 24, 2010 11:55 AM

    Re: BABIP – does a weak grounder count the same as smoking a ball into the gap or destroying into the bleachers? If so, I don’t get the fascination with BABIP. Some pitchers induce a lot of grounders/weaker balls put into play. Others serve up batting practice. It seems to me that both of these fall into the “balls in play”, so looking at that stat as an indication that Cole has been unlucky because the ones that his opponents are hitting just so happen to find a home in between fielders.

    If Cole isn’t executing his pitches, or isn’t locating, doesn’t that increase the likelihood that a MLB hitter is going to be able to react with his bat in such a way that he can increase the likelihood of the ball landing in play?

  14. Chris

    April 24, 2010 11:59 AM

    As someone who has constantly defended Cole with similar arguments, albeit not as smart and in depth, I must admit it is becoming increasingly frustrating.

    Hopefully he does make these minor adjustments and in the end we can all say “Told ya so.”

  15. Bill Nodrog

    April 24, 2010 03:26 PM

    Another year of “Headcase Hamels”.
    This season for Hamels as a Phillie “…can’t be over soon enough.”

  16. Simon

    April 24, 2010 06:20 PM

    Just want to chime in and applaud you for continuing to use advanced stats and actual analysis, while others lose their heads over a couple of unlucky starts. Bravo, sir, and don’t let the haters get you down. :)

  17. Phylan

    April 24, 2010 07:03 PM

    I’m actually interested to see if he can sustain his K/9 at around 9.5, since it’s hovered around 7.8 the last two seasons. Not worried about Cole at all.

    Great analysis as usual.

  18. Peter Hjort

    April 25, 2010 12:11 AM

    Matt re: BABIP.

    Pitchers can influence their ground ball and fly ball rates, but limiting line drives is not a pitcher skill (well, in MLB at least).

    Fly balls fall for hits much less frequently than ground balls, but fly ball hits tend to be of the extra base variety while ground ball hits are usually singles, so ground balls are less likely to result in runs.

    Anyway, the statistic that embraces this theory the most is SIERA, which is currently the best measure of fundamental skill we have. Batted Ball FIP may be an improvement when that comes out, but right now it’s SIERA. Cole’s is 3.13, he’ll be fine.

  19. David

    April 25, 2010 02:41 AM

    Bill – You asked for proof that Hamels has regressed, arguing that the “wrong stats” give an improper perception of a pitcher’s skills and performance. The problem with the statistical analysis is this: the bottom line in MLB isn’t skills and performance, it’s results. In 2008 the Phillies won every postseason game that Hamels started; last year they won every postseason game that Lee started. Hamels may be pitching well, but he hasn’t dominated consecutive games like he did when he won his WS MVP, and the team hasn’t won with the same level of consistency. If you don’t include results in your measurements then you’re going to have trouble convincing anyone that’s actually paying baseball players and personnel – as well as the people that award the WS MVP.

    Besides, small sample size or not, you’re going to have trouble with reconciling your p-value with your skill and performance statistics when you look at streaks like the postseason one that pushed Hamels to prominence – and they happen all the time in baseball. Baseball is full of anomalies, suggesting there is much more than just the statistics that lead to overall results. Pitchers and players get hot and get wins and hits that they “shouldn’t” – it’s not just “luck”. A pitcher may have little visible control, for example, over defense and run support, but there are ways that a pitcher can affect these through leadership and inspiration (or lack thereof). These qualities aren’t going to come out in statistics.

    Also, of the statistics you’ve mentioned, how thoroughly have some of them been measured regarding hitting? For example, what kind of range is there among *batters* when it comes to HR/FB? Does someone like Ryan Howard have an unusually high ratio, or merely hit far more fly balls? Dismissing the number because it’s anomalous in the case of Hamels may be a mistake if it indicates that Hamels is throwing the wrong kinds of pitches to certain players.

  20. Bill Baer

    April 25, 2010 05:44 AM

    The problem with the statistical analysis is this: the bottom line in MLB isn’t skills and performance

    I’m not arguing that it is, and nor is that relevant to the discussion. We are discussing whether Hamels is actually as bad as he has shown over the 2009 season and the first fours starts in ’10. That’s why we’re using retrodiction.

    Hamels may be pitching well, but he hasn’t dominated consecutive games like he did when he won his WS MVP, and the team hasn’t won with the same level of consistency.

    Of course, no one is denying this. No one is burying their head in the sand and saying that Cole did not put up a 4.32 ERA last year. We are analyzing just how much of that 4.32 ERA is attributable to events within Hamels’ control (the answer: not much).

    If you don’t include results in your measurements then you’re going to have trouble convincing anyone

    You are missing the entire point of the discussion.

    Besides, small sample size or not, you’re going to have trouble with reconciling your p-value with your skill and performance statistics when you look at streaks like the postseason one that pushed Hamels to prominence – and they happen all the time in baseball.

    This supports my argument. Streaks happen and very rarely are they wholly attributable to one player’s level of skill.

    Pitchers and players get hot and get wins and hits that they “shouldn’t” – it’s not just “luck”.

    That is the very definition of luck.

    A pitcher may have little visible control, for example, over defense and run support, but there are ways that a pitcher can affect these through leadership and inspiration (or lack thereof).

    bahaha

    Also, of the statistics you’ve mentioned, how thoroughly have some of them been measured regarding hitting? For example, what kind of range is there among *batters* when it comes to HR/FB? Does someone like Ryan Howard have an unusually high ratio, or merely hit far more fly balls?

    Like with BABIP, hitters have much more control over HR/FB than pitchers. Ryan Howard has a career HR/FB of 31.3% which is a lot higher than Placido Polanco’s 6.7%.

    Dismissing the number because it’s anomalous in the case of Hamels may be a mistake if it indicates that Hamels is throwing the wrong kinds of pitches to certain players.

    Is Hamels facing lineups full of Ryan Howards?

  21. Gio

    April 25, 2010 06:31 AM

    All of this is BONK! Here’s hoping Hamels figures things out soon. He’s been more bad than good early this year, resembling the inconsistent ’09 version way more than the ’08 version. All of these made up stats can’t make up for the fact that he’s making the bad pitches, he’s giving up the homers, HIS ERA is over 5. Supremely talented, just doesn’t seem to be able to consistently master the mental part of being a dominant ace.

  22. InsaneFan

    April 25, 2010 07:03 PM

    I AM PANICKING

    SEND COLE TO THE MINORS

  23. Jeff

    April 25, 2010 11:21 PM

    When Cole leaves a pitch up in the zone, it’s going to turn into a fly ball. Whether that fly ball leaves the yard depends several factors, including:
    - Whether the hitter makes good contact
    - The wind/humidity/air density
    - Which part of the park the hitter hits the ball into

    Notice that Cole has ZERO control over these factors, which is why so much of HR/FB ratio is luck.

    BUT- the closer the pitch is to the middle of the plate, the better the chance that the hitter is going to make good contact. And the simple fact of the matter is that Cole is prone to making far too many batting practice quality pitches/mistakes, which I think leads to inflated home run numbers.

    I’d love to see (if it exists anywhere) some advanced analysis of Cole using pitch trackers to see how much of this “bad luck” is really due to a handful of god-awful pitches each start.

  24. Mratfink

    April 26, 2010 12:56 AM

    To follow up to the people that don’t understand what Bill is saying (and forgive me for putting words in your mouth Bill). Certain statistical factors (such as gb/fb, and k/bb) would suggest that if Cole were to continue these rates then he would show marked improvement. This will mainly be because his early season home run rate is abnormally high and should show improvement. Should that happen and were Cole to maintain his gb/fb and k/bb Cole will be putting up very good numbers this year.
    That may have been more confusing than intended but simply the purpose of this analysis is to show that Cole is actually do a lot right and being victimized by one thing he is doing very poorly (or as Bill would say is very unlucky )

  25. Bill Baer

    April 26, 2010 03:09 AM

    Yes, to sum up what Mratfink is saying in one phrase: regression to the mean. I think it’s a concept that most people intuitively understand but have trouble applying to baseball.

    Example: You draw a card from a deck and replace it ten times. You get a face card each time. That seems peculiar to you since there is a 12/52 (23%) chance of drawing a face card once, but only 4 in 1,000,000 ten consecutive times.

    The results of your drawing were odd but are unlikely to be repeated in future trials. The larger the sample size, the more the results mimic the event’s true probability (regression to the mean). For every ten cards you draw, you should be getting approximately two face cards.

  26. David

    April 26, 2010 04:25 AM

    Bill – The thing is, if you’re correct and Hamels should more or less pitch the same as he is now, then he should put up results, especially with a lineup like the Phillies have. Results *are* relevant to the discussion, because what you’re arguing boils down to is Hamels not changing being what will produce the results in the long run.

    What was the SIERA for Hamels last year? Based on run support and defensive play and the numbers that are supposed to predict performance, what should Hamels have managed in terms of record and ERA? I don’t have those numbers, but it seems to me that there’s a significant difference between where Hamels supposedly should have been and where he was and is. This is where the p-value comment comes into play – with 200+ pitchers that start over the course of a season, some of them are going to have results that don’t match up, yes, but if those results continually don’t match up then what’s statistically most likely is that numbers like HR/FB and GB/FB aren’t telling the whole story. One player may be streaky, but another may be Joe DiMaggio and put up probability defying hitting streaks.

    In other words, when regression to the mean hasn’t happened and isn’t happening – the mean is probably off.

    (Also, there’s a big difference between Howard hitting 3 fly ball outs and 3 home runs “luck” and inside the park home run funny bounce luck. Baseball players aren’t paid to roll dice. This is what I’m referring to with the Howard example – if hitters do have control over these numbers, and pitchers aren’t performing equally across the distribution of hitters, then again, performance and results numbers won’t match up: if you give Howard twice as many fly balls as Polanco because of how you’re pitching to them, you’re going to end up with a weird HR/FB number. Perhaps Hamels is making more mistakes against better hitters or in worse situations, making him seem “unlucky”.)

  27. Bill Baer

    April 26, 2010 04:42 AM

    What was the SIERA for Hamels last year?

    ERA: 4.32
    FIP: 3.72
    xFIP: 3.69
    SIERA: 3.55

    W-L record is useless, don’t even bother with it.

    but if those results continually don’t match up then what’s statistically most likely is that numbers like HR/FB and GB/FB aren’t telling the whole story.

    Hamels hasn’t consistently under-performed ERA retrodictors; he has only done so in 2006 and ’09, and thus far in ’10.

    One player may be streaky, but another may be Joe DiMaggio and put up probability defying hitting streaks.

    If Joe DiMaggio was immortal, it would be a long, long time before he hit in 56 consecutive games again.

    In other words, when regression to the mean hasn’t happened and isn’t happening – the mean is probably off.

    You’re talking like Hamels was supposed to regress to the mean prior to the start of 2009. In fact, Hamels did regress to the mean after ’08, and went further the other way. His skill level is somewhere in the 3.50-3.75 area.

    Give it time before you jump to conclusions. He has made four starts this year.

    Also, there’s a big difference between Howard hitting 3 fly ball outs and 3 home runs “luck” and inside the park home run funny bounce luck.

    Why is this germane to the discussion?

    Perhaps Hamels is making more mistakes against better hitters or in worse situations, making him seem “unlucky”.

    Home runs by batting order position:
    - Batting first: 1
    - Batting second: 1
    - Batting fourth: 1
    - Batting fifth: 3
    - Batting eighth: 1

    The 3 sticks out but we don’t know how many fly balls were hit (presumably the #4-5 hitters hit more fly balls than #1-2 and 8 hitters) and it is a very, very small sample of about 12 PA each.

    As for situations, Hamels has actually pitched worse with the bases empty.

    - Empty: .916 OPS, 5 HR
    - RISP: .532, 0 HR
    - Runner on first: .894 OPS, 2 HR

    Again, we need the # of fly balls in each split to truly make a judgment with those numbers.

  28. Steve

    April 26, 2010 08:29 AM

    Bill, perhaps I’ve missed this, but could you explain your SIERA stat real quick? I suppose I could also look it up.

    I’m not as worried about Cole as most, mostly because of said stats in this article. However, I do worry about his inability to stop the big inning. With Hamels, it seems to be either no runs or 5 runs. Once things start to go bad, he doesn’t seem to be able to get out of it.

  29. Bill Baer

    April 26, 2010 08:37 AM

  30. Ryan

    April 26, 2010 09:09 AM

    This is an argument that rages throughout message boards and comments sections, and to be quite honest, it is infuriating to me.

    It’s the division between a group that says “we have millions upon millions of innings of baseball that say, statistically, this is unusual, an outlier, and will regress to the mean” versus the a group that says “I watch the games, I know this is Cole’s fault, and the reason that his fb/hr rate is so high is because he makes crappy pitches”.

    Here’s what I have to say to the group who wants to trust their eyes above statistical evidence. Chart some games. I mean, seriously, find a pitch chart template online, and chart some baseball games. Chart Doc, chart Blanton, chart Santana, chart Tommy Hanson.

    You’re going to find some similarities… 1 – ALL PITCHERS THROW BALLS IN THE MIDDLE OF THE PLATE OVER THE COURSE OF A START. It’s impossible not to. Unless you are Greg Maddux circa 1995, through the course of 100 pitches, you are going to miss spots, and most of the time when you miss, its going to be back toward the plate.

    Now – where the statistical part of this comes into to play is, the numbers say the more fly balls you give up, the more hr’s you are going to surrender. It’s logic. Groundballs can’t leave the park, and neither can K’s. But, in order for an HR to be hit, several things have to happen – 1) the hitter has to be swinging (not always a given) 2) he’s gotta make close to perfect contact (very hard to do) 3) wind/air/fence depths still all play a role.

    Where does luck come in to this? Some times, you have starts where you miss down the middle, and for whatever reason, guys are looking offspeed, they are taking pitches early in the count, and the meatball you threw get’s taken. Some times the wind holds up ball that would have gotten 2 rows deep. Some times guys just flat out miss em and foul off a meatball or roll over a pitch they should drive. It’s the part of the game the pitcher can’t control.

    This is why its so important to look at the TON of data that we have on hand, instead of one game’s worth. In one game, you can have a day where the ball is flying out (come on, kelly johnson really going oppo gap?), and where hitters just don’t miss your mistakes. Some days your going to make mistakes, and they are going to find gloves due to whatever the circumstances.

    This is a representational psychology issue. What you WATCH and then process into conclusion – Cole made bad pitches, bad pitches were hit for home runs = every time cole gives up a home run it was a horrible pitch, every time cole makes a bad pitch it will be a home run.

    That just isn’t true. What is true, is that given his peripheral indicators, Cole is going to regress to the mean. He can’t continue to give up 4 of every 10 fly balls over the fence. It’s so statistically improbable it’s ridiculous. What he can effect, giving up fewer pitches that can be elevated, and if I’m not mistaken, he’s actually improved upon the fb rate since ’08. So he’s done that. Now its a matter of things evening out over time. He’ll have some games where he gets away with mistakes, he’ll have some games where he makes fewer mistakes, and as long as he and Dubee don’t panic and get crazy tinkering with things, he will be just fine.

    For all of you saying that Bill is blindly backing Cole with solid statistical information, you may want to understand the statistics behind the logic before you insult the actual research.

    Put simply, if Cole continues to pitch the way he has (k rate, bb rate, gb/fb rates), he would be one of the most significant statistical outliers over the course of 30 starts that I can ever think of. It’s just flat out statistically improbable.

  31. Chareth

    April 26, 2010 10:02 AM

    A few questions for Ryan and Bill:

    Has anyone look at the rate at which Hamels throws a meatball? Given that the issue at hand with Hamels seems to be a lapse of concentration or psychological breaking point, is it completely out of the realm of possibility that when Hamels hits one of these hypothetical points, that his meatball rate runs higher? And that this higher, temporary meatball rate turns into additional runs while not affecting his SO/BB and GB/FB rates?

    While I haven’t calculated SIERA myself to get a feel for how additional PAs really swing the numbers, I’ll go on a limb and guess that once you have enough samples, a few additional meatball PAs wouldn’t skew the number too much.

  32. Ryan

    April 26, 2010 10:36 AM

    Chareth – I can’t answer for Bill, but I have one general comment.

    If Cole were throwing more pitches in the middle of the plate, a higher rate of meatballs, I would think the first place it would show up is in his flyball and line drive rates.

    Basically saying, if he was missing on the middle of the plate more, there would be more pitches elevated/hit hard.

    Here are his ld rates and fb rates over the past 4 years –

    LD Rate –
    07 – 19.4%, 08 – 21.8%, 09 – 20.8%, 10 -16.4%

    FB Rate -
    07 – 38.7%, 08 – 38.7%, 38.7%, 10 – 37.0%

    And this is precisely what Bill was speaking to. The two areas you would expect to see increases if Cole where, in fact, throwing more meatballs, would be flyball and line drive rate. Essentially, pitches up and down the middle are easier to elevate, drive, and barrel up.

    Only, the data says that both his LD and FB rate have been down (or normal in ’09), suggesting, he’s not making any more mistakes than he was in ’07 and ’08 when he was really good (or his numbers were better than he pitched).

    To go along with that, his current gb rate is 6% higher in his 4 starts than his career numbers, meaning, logically, because he’s getting more groundballs, you would think his hr numbers would be falling, not increasing.

    So we come back to this – he’s getting more groundballs, he’s k’ing hitters at a better rate than he has throughout his career (missing more bats), and walking fewer than he has throughout his career. Essentially everything that he appears to be able to control, he’s doing as well, or better than he has when he was “good”.

    The only difference in the statistics? The hr numbers – he’s given up twice as many hr’s per 9, and twice as many hr’s per fb than he has in his career – despite the rest of the numbers suggesting that he’s actually throwing better.

    So he’s harder to elevate, harder to hit line drives off of, but for whatever reason, with fewer balls being hit in the air, more are going out.

    Essentially,it’s the definition of bad luck. And Jamie Moyer acknowledged it to start last year. Some times, every single one of those mistakes you throw end up over the fence, some times you throw 10 in a game, and not one is even hit for a single, much less an hr.

    Statistics say that Cole is due for a stretch at some point where he gives up 7 or 8 fly outs in a game without any leaving, and probably a stretch where he has more fb’s than normal over a 3 or 4 game period, with only 1 or so leaving.

    Once that happens, and he gets out of this poor stretch, ALL of his numbers will normalize – and as the hr numbers come back to earth, so to will the era and whip.

  33. Ryan

    April 26, 2010 12:03 PM

    This is one of those really hot button topics amongst Phils fans and I haven’t really seen much middle ground.

    The message boards are filled with the statistical supporters and those who have given up on him and want him to have a one way ticket to Lehigh/the bullpen.

    Those giving up point to last year, those who support point to the previous two years. In the end, it’s up to Cole to find that consistency and start to define a career that is, in many ways, at a crossroads.

    Many veteran pitchers go through rough seasons where they see babip or fb/hr outliers curtail a solid career – but the good ones find a way to keep plugging, keep positive, and put the work in to make sure that those fluky seasons only happen once or very rarely. I think that it is still up in the air in terms of Cole’s mentality if he can battle through this or if it will lead to struggles with confidence.

  34. Chareth

    April 26, 2010 12:08 PM

    Ryan, thanks for the reply. I completely get the stats and your conclusions.

    However, you’re looking at the data at a macro level and completely discounting plausible micro events. I think there are perfectly plausible situations where Hamels’ peripherals could look fine, but his other stats look bad because of a handful of particularly terrible pitches.

    What if those meatballs in his last game were particularly meaty at 87 mph with no movement? You don’t think a pitch like that is much more likely to leave the ballpark, than a much more average Hamels meatball?

    I think sabermetric analysis is fantastic but I think it shouldn’t be used to simply deride all human and emotional aspects of the sport. I’m sure there is a high probability that my hypothetical scenarios do not hold water, but that doesn’t mean they are completely implausible.

  35. Chareth

    April 26, 2010 12:19 PM

    One more thing to address your recent post. I think people are absolutely judging Hamels unfairly. Not only because we’re talking about a very small sample size; I can remember Cliff Lee having a particularly terrible stretch last year that was much worse than Hamels’ last 4 games, IIRC.

    On top of that, I think we as fans need to show more loyalty. That people boo him is disgusting. This is the kid that got us our championship, and we treat him like a batboy.

    Regarding his performance, I do think the truth is probably somewhere in the middle. Hamels is clearly not as confident as he used to be and I don’t see why this wouldn’t affect him negatively in certain situations.

  36. Ryan

    April 26, 2010 12:36 PM

    Chareth,

    I couldn’t agree with you more. I struggle, many times, with exactly what you wrote. The human, observation element makes it very difficult to just sit back and accept “hey it was just bad luck” in regard to anything.

    I coach hs baseball, and last week I watched my starting pitcher get bled to death. Bloop single here, infield hit, error, etc. And then, it resulted in back to back home runs that broke the game open.

    Now somewhere in there I have to find a way to evaluate his performance and his progress as a young man. To say that it was all bad luck eliminates some of the teaching points (perseverance in particular), but to say it was a horrible outing completely disregards the quality of pitch being made.

    In the end, we are talking much much more about how to predict future success, and a bit less about how to evaluate Cole’s last start. Was his last start poor? Definitely. Was it HE who made the bad pitches – yes. And my biggest criticism is, after the first home run, he has to have the swagger, confidence, and focus, NOT to let the next one happen – thats the micro, or event analysis of the game.

    My use for the stats is purely to say, look, if he does this over 30 starts, he will be fine. That is the macro end. If he doesn’t lose confidence, doesn’t change anything drastic, over the long run, as Bill said – he’s probably a 3.60 type era guy, or better if he keeps up the high gb rate and lowers the hr total.

    But in the short run, in terms of what I want to see out of Cole – its the reaction to the adversity that is my only beef with him. Cole seems to be dangerously into that over thinking, over analyzing territory, and not enough of focus on the next pitch and utilizing the wealth of talent he has.

    Part of what you wrote in your second post goes to that. The more pressure we put on him to be perfect, the more he puts on himself, and the more his focus is on the wrong thing during the game. That is just my opinion.

  37. Reverend Paul Revere

    April 26, 2010 01:59 PM

    I am, admittedly, a neanderthal when it comes to assessing advanced statistics, though I do respect them. My question is this: Given that last season and thus far this season seem to be statistical anomalies for Hamels, when can we reasonably expect his other numbers to regress back to the mean? Can that be predicted.

    I’m wondering because, forgive me if this sounds ignorant, it seems like these sort of things should even out over the course of a season. Cole had 32 starts last year and has four this year, and the numbers aren’t matching up, if I’m understanding this correctly. Is this something that statistically will take years to regress back to the mean? What if he doesn’t regress back to the mean? What does that mean? He’s just perennially unlucky?

    I’m not giving up hope on Hamels, but whether or not it’s the old-school or wrong way to look at it or not, I think we can all agree a 5.11 ERA has to do with a little more than luck or lack thereof, even if only slightly.

  38. David

    April 26, 2010 02:11 PM

    “However, you’re looking at the data at a macro level and completely discounting plausible micro events. I think there are perfectly plausible situations where Hamels’ peripherals could look fine, but his other stats look bad…”

    This is also what I’m getting at – though I think it could be more than just the occasional lousy pitch. I think the macro data may be deceiving. We’ve seen Hamels pitch to his expectations, and then we’ve seen him fall short of it, spectacularly. This kind of inconsistency can hurt (or help, if we consider the 2008 postseason results to be “lucky”) in ways that don’t match up with statistics like a nice bell curve distribution of performances would.

    A perfect example of this is Cliff Lee from last year: in his first five starts with the Phillies, he pitched 40 innings, giving up only 24 hits and 6 walks while striking out 39. Then he was terrible the next three games…and in the ninth outing he had a complete game shutout. Over 17 games with the Phillies he had 5 games that look just terrible and 10 that were fantastic – the exact opposite of a bell curve.

    For what it’s worth, I like Hamels and don’t think the Phillies should get rid of him, demote him, etc. – I think he’ll return to strong form…when he figures his inconsistency issues out. The game on Friday was like Lee’s game by game inconsistency applied to innings. If the bad pitches and poor decisions get concentrated into one lousy inning then it’s more likely to see the game get blown open, instead of having a run scored here, two runs there. There’s definite room for improvement beyond getting less “unlucky”.

    (Why is this germane to the discussion? Because there are no dice being rolled or coins being flipped when someone swings a bat and hits a home run. Hamels can’t control a lot of factors, but other players can; that’s not luck any more.)

  39. Mark

    April 26, 2010 02:19 PM

    This is absurd…

    First, you use statistics to back up your theory of “bad luck”, which is contradictory in itself.

    Second, pitchers can affect their hr/fb rate. Call it pitch location, skill, or whatever you want; good pitches will affect how well the batter can hit the ball. If these hits were weak ground ball singles or bloopers that’s one thing, but home runs indicate bad pitches. Cole had a bad game. It happens. It had nothing to do with luck; he just couldn’t make his pitches.

  40. Bill Baer

    April 26, 2010 02:31 PM

    Mark,

    Johan Santana’s career HR/FB is 9.4%. Jake Peavy’s is 9.6%. Roy Halladay’s is 10.2%. Josh Johnson’s is 10.5%. Dan Haren’s is 10.9%. Chris Carpenter’s is 11.0%. Vicente Padilla’s is 11.3%. Adam Eaton’s is 11.5%.

    The evidence is wholly in favor of pitchers having very little control on their HR/FB%.

  41. Mark

    April 26, 2010 02:54 PM

    Bill,

    Of course if you look over their whole careers, they will have a low hr/fb rate. There is a reason they are major league pitchers.

    You fail to realize that every at bat is an independent event. The past doesn’t have any control over what happens. If he can’t make a quality pitch, it will get hit far. If he makes a quality pitch, it will be a fly ball.

    He had a bad game. It happens.

    These stats are good for making predictions based on how a pitcher has thrown over his whole career, but when it comes down to it, the man has to make pitches.

    You act like Cole robotically throws the same game every time he takes the mound. He doesn’t. His past numbers mean absolutely nothing to the outcome. There is no reason why his numbers will ever return to his averages if he keeps throwing fat pitches. Pitching is based on skill, not probability… probably the main reason why I’m sitting behind a desk right now and not standing on a baseball diamond somewhere.

  42. Mark

    April 26, 2010 03:29 PM

    You even comment on how bad his location was on those home run balls. I don’t understand your argument that pitchers have little control on hr/fb.

    The only evidence you showed me is that all of those guys had the stuff to be mlb pitchers.

  43. Ryan

    April 26, 2010 05:38 PM

    Mark,

    You look at the career rate for Roy Halladay 10.2 and Adam Eaton 11.5 – and see that Eaton only gives up 1.3 more home runs per 100 fly balls, and all you take from that is “they are both mlb pitchers”.

    Those two couldn’t be further apart in terms of ability, and command of the baseball and yet they both give up hr’s at almost exactly the same rate. What Bill is trying to tell you is that despite the blatant disparity in “stuff” and ability, Halladay still isn’t markedly better at keeping fly balls in the park. In fact, he’s barely better at all.

  44. Mark

    April 26, 2010 10:06 PM

    Ryan, we’re talking about the home run to fly ball rates… and they were close, but i bet Adam Eaton gave up hits at a faster rate because of his poor command. Any proven pitcher in the major leagues will have many more fly balls compared to home runs.

    Bill, if you want evidence that location on a pitch matters, you should tune into the home run derby this year. I bet the hr/fb rate is well over Eaton’s 11.5%!

  45. Bill Baer

    April 26, 2010 10:11 PM

    but i bet Adam Eaton gave up hits at a faster rate because of his poor command.

    Patently false. Pitchers have very little control on their BABIP as well.

    Examples:

    Adam Eaton: .305 BABIP
    Dan Haren: .300
    Roy Halladay: .299

  46. Mark

    April 27, 2010 12:00 AM

    BABIP is not a rate. Adam Eaton gave up 11 hits per 9 in 2008 compared to Halladay’s 8/9 that same year, which really isn’t all that much, but if your team is on the losing end in each of those games (recording all 27 outs) the difference in average is .289 to .228.

    That’s a lot more than you make it out to be with BABIP

  47. Mark

    April 27, 2010 12:04 AM

    ha.. *BABIP is not the rate i was talking about* my bad

Next ArticleGraph of the Intermittent Time Period