Over the weekend the Marlins swept the Phillies to move within four games of them in the NL East and three games out of the Wild Card lead. The Marlins are a real surprise this year, as some projected them to be worse than the Nationals. Last night Josh Johnson won the game for the Marlins, pitching six innings of four-hit, one-run ball while striking out 6, walking only one and getting seven ground balls on 14 balls in play. Johnson has been great this year, doing everything a pitcher should do, getting lots of strikeouts and ground balls while limiting walks. He has a FIP just over 3, good for ninth in the league.

Johnson has come back from his 2007 Tommy John surgery in a big way. His fastball averages over 95 mph, third highest among starters and his slider averages 86.4 mph, is one of the ten fastest. He also throws a changeup.

As you can see he throws a four-seam fastball with a good 10 inches a rise. This is really interesting as he is a ground ball pitcher; he gets over 50% of his balls in play on the ground. Most ground ball pitchers do that with a sinking two-seam fastball, like Derek Lowe, or get their groundballs on a slider or curve, like Brett Anderson. Johnson’s four-seam fastball induces 50% ground balls, even though it has 10 inches of rise. He is able to do this by locating it low in the zone.

The difference appears subtle, but over the course of the a season has a huge effect. This placement makes Johnson’s four-seam fastball a ground ball pitch, unlike most four-seam fastballs. He is able to locate his ‘rising’ fastball low in the zone. Here is how he uses his three pitches to lefties and righties.

+----------+-------+-------+
|          |  vRHB |  vLHB |
+----------+-------+-------+
| Fastball |  0.66 |  0.67 | 
| Slider   |  0.30 |  0.19 |
| Changeup |  0.04 |  0.14 |
+----------+-------+-------+

The slider is a strikeout pitch getting lots of whiffs and out of zone swings, but also lots of fly balls. His changeup is not a strike out pitch, like Tim Lincecum‘s or Rich Harden’s. It gets very few whiffs and out of zone swings, but is an extreme ground ball pitch (over 67% ground balls per ball in play). This leads to a very interesting platoon split.

+--------+--------+--------+
|        |   vRHB |   vLHB |
+--------+--------+--------+
| OPS    |  0.646 |  0.713 | 
| K/PA   |  0.228 |  0.179 |
| BB/PA  |  0.073 |  0.099 |
| HR/BIP |  0.033 |  0.019 |
+--------+--------+--------+

His slider is much better than his change and he fastball has a normal platoon split, so has a whole he has a pretty big split. He can locate his slider well in the zone and gets lots of whiffs with it, but it induces FBs. As a result against RHBs he gets strikeouts and walks them rarely, but they hit more HRs against him. Against LHBs he goes with his changeup which gets few whiffs or out of zone swings, but lots of ground balls. So he actually gives up fewer HRs against lefties.

Josh Johnson is one of the ten best pitchers in the game and a valuable asset to the Marlins in their playoff push.

A couple weeks ago I looked at Chris Davis‘s contact struggles. In that post I displayed a histogram with all contact rates from 2003 to 2008 and Davis’s half season of 2009. Here is the same figure without Davis, but with Mark Reynolds’s 2008 contact rate indicated.

Not as much of an outlier as Davis, but last year Reynolds’s rate of 62.3% was the lowest of any regular since 2003. This year he is similarly lowest in the league with 63.4%. Reynolds, unlike Davis, makes it work. Even with this very very low contact rate Reynolds has a wOBA over .400, 12th best in the league.

One thing he has over Davis is much better plate discipline, swinging at only a quarter of pitches out of the zone compared to Davis’s 35.4%. As a result, and because pitchers don’t throw in the zone to him that much, he has a healthy walk rate of 11.8%.

In addition the pitches that Reynolds does make contact with are very likely to be HRs. Fly balls make up 46% of his balls in play and 28% of those are HRs. That means 12.9% of his balls in play are HRs (tops in the league with Adam Dunn second at 11.6%). As a result Reynolds is second in the league in HRs with 32.

Reynolds represents what Davis needs to be if he is going to succeed with a huge whiff rate. He needs to stop swinging at pitches out of the zone, and make sure the pitches he hits go a long way. Not rocket science, but it is helpful to see someone how makes it work with a big whiff rate.

The A’s came into the season looking to contend with some newly acquired position players and a group of young pitchers. The acquisitions didn’t hit and the young pitcher’s have been shaky, so the A’s traded off Matt Holliday and Orlando Cabrera to continue to stock their growing farm system and push contention off for another year. In a frustrating year there have been some bright spots, one is Brett Anderson. He has a FIP right at 4 and had the third best FIP in July.

Throughout his minor league career Anderson has been known for his very good command and great ground ball numbers. Interestingly it looks like he throws a four-seam fastball, which only gets 38% GBs. Most ground ball pitchers throw a ‘sinking’ two-seam fastball, which gets most of their grounders. Anderson, though, doesn’t get many grounders from his fastball. Instead his main ground ball pitch is his slider.

According to the BIS classification Anderson is just one of four pitchers to have thrown over 30% sliders so far this year. The pitch is very interesting. It has the speed of a slider,

and the movement of a curve.

Unlike most sliders which show a pronounced platoon split, Anderson’s shows none, more like a curve. As a result he throws it almost equally to both LHBs and RHBs. The top sliders in the game get huge whiff rates, in the 40% range, by comparison Anderson’s is rather meager at just 25%. But Anderson can get his in the zone over 52% of the time and it induces ground balls over 68% of the time. Both of these rates are very high for sliders. All together it is worth almost 13 runs and is the third best slider in the game.

Anderson is a really interesting pitcher. Most ground ball/control guys do it by throwing two-seam fastballs around 70% of the time. Anderson does it with a slurvy-slider which moves like a curve, has little platoon split, he can locate in the zone and hitters pound into the ground.

The Twins picked up shortstop Orlando Cabrera from the As, looking to solidify their middle infield as they chase the Tigers and fend off the White Sox in hopes of a AL Central title.

Cabrera replaces Nick Punto at short for the Twins, a major offensive upgrade. Cabrera has a rest of the season ZIPS projected wOBA of .316, while Punto a meager .287. Over the maybe 200 at-bats he will get this is a 5 run upgrade. Calculating the difference in defense is a little tougher. Cabrera’s UZR has been very bad this year, but UZR is prone to rather large fluctuations. It is better to just use his career numbers, +5 UZR/150, as our best guess for him going forward. Over Punto’s career he is a +15 UZR/150 shortstop, but that is over only 1.5 seasons, not really enough to be confident. Plus he is only a +3 UZR/150 second baseman. Combining those and regressing our best guess is that he is around a +7.5 UZR/150 shortstop. So the difference is fielding is probably less than a run over only 50 games.

EDIT: It looks like I was wrong here. The Twins are going to shift Punto to second and Brendan Harris to the bench. So Cabrera’s offense replaces Harris’ and Punto defense at second replaces Harris’. Harris has a projected wOBA of .296, so Cabrera’s .316 over 200 PAs should be worth 3.5 runs. I think Harris is probably about -5 run defender at second while Punto something like a +5 run defender. Over 50 games this is another 3 runs. So over all probably still about half a win. Thanks to the commenters for pointing this out.

All told this is about a half a win gain for Twins, not an insignificant number in the thick of a playoff race. Cabrera does not have any value beyond that. His contract says he cannot be offered arbitration, so he will not be any worth picks if the Twins do not resign him. Last year Cabrera was a Type A free agent and he felt, rightly, that status depressed his value, so he included this clause in his current contract.

The Twins gave up their 2008 second round pick, 21 year old shortstop Tyler Ladendorf. Ladendorf had a great 74 plate apperances in rookie ball with an OPS over 1.2, but has had a tough 65 at-bats since being promoted to A ball, with an OPS of 0.559. He fell just outside of John Sickels top 20 Twins prospects, but Sickels likes him thinking he has the tools to stay at short.

I like the deal. The Twins are exactly the type of team, in tight contention for a playoff spot, who should be dealing. Thus even half a win is quite valuable to them. Cabrera’s contract limited his value, because he cannot be offered arbitration, so the As did a nice job getting something of value for him.

The big name we will hear a lot today either because he is or is not traded is Roy Halladay. Dave Cameron broke down his value and if he is traded we will have analysis of the trade here. Until then I thought it would be cool to look at his stuff.

For the past two years he has had five strikeouts for every walk and over half of his balls in play have been groundballs, amazing. Here are the pitches he does it with.

Interestingly his pitch proportions against RHBs and LHBs are quite similar. He throws a lot more changeups against lefites, but the proportions for his other pitches are very close.

 
+-------------------+------+------+
|                   | vRHB | vLHB |
+-------------------+------+------+
| Two-Seam Fastball | 0.34 | 0.31 |
| Cutter            | 0.39 | 0.43 |
| Curveball         | 0.26 | 0.20 |
| Changeup          | 0.01 | 0.06 |
+-------------------+------+------+

His cutter and curveball are both amazing, worth 1.5 and 2 runs above average per 100 over his career, respectively. The cutter’s worth comes from his ability to get it in the zone over 60% of the time while still getting a healthy number of whiffs (14%) and a good amount of GBs (48%). His curveball is a whiff-machine, it has the third highest whiff-rate of any curveball in the game this year 39%, while still being in the zone 44% of the time, fairly high for a curveball. His two-seam fastball, is also very good worth about half a run over 100 pitches. It gets most of its value because it induces 65% groundballs per ball in play. Tons of strikes, tons of groundballs, and a fair number of whiffs, what more could one ask for?

Combine those two amazing pitches with one very good one and you have one of the greatest pitchers in the game.

The Phillies acquired Cliff Lee yesterday with the eye on the playoffs, they are six games up in NL East and by at least these estimates have over an 81% chance of reaching the post season. So I thought I would look to a potential Philly playoff berth. With Lee, their potential playoff rotation looks like:

Cole Hamels, LHP
Cliff Lee, LHP
Joe Blanton, RHP
J.A. Happ or Jamie Moyer, both LHPs

This is three-lefties-one-righty rotation is all the more interesting in light of the potential righty-heavy lineups they could face. Dave noted that the Cardinals have a very righty-heavy lineup, and the Giants only have one lefty in their regular lineup now that Ryan Garko replaced Travis Ishikawa. On the face of it this could present Philly with a real problem, but let’s look at these lefties’ career platoon splits.

 Career OPS against
+--------+----------+----------+
|        | OPS vRHB | OPS vLHB |
+--------+----------+----------+
| Hamels |    0.698 |    0.716 |
| Lee    |    0.737 |    0.710 |
| Happ   |    0.711 |    0.651 |
| Moyer  |    0.741 |    0.764 |
+--------+----------+----------+

Hamels has a devastating changeup, a pitch that generally does not show a platoon split, as a result Hamels has a reverse platoon split.

Cliff Lee, with a good changeup and curveball another pitch with no platoon split, has shown only a modest platoon split over his career.

Happ, on the other hand, has a quite large platoon split, which he had in the minors too (0.678 vRHB and 0.622 vLHB). It is backed up by his pitch usage which is very fastball heavy, with almost no curves and few changeups.

Moyer is an interesting case, over his career he has had a reverse platoon split on the strength of his good changeup and curveball. This year, though, he has a very drastic split in the other direction (0.885 OPS against RHB and 0.700 against LHB). Normally I would write this off as noise, but as I noted in this post Moyer is throwing many fewer changeups and curves this year and many more two-seam fastballs, which have a large platoon split. If he keeps this pitch breakdown going forward I would assume his split will stay like it has been this year.

All told I don’t think it should be much of a concern. Lee and Hamels are both great pitchers who do better or just marginally worse against RHBs, so I don’t think the Phillies should worry about facing righty-heavy lineups in the playoffs.

Over three games Mark Buehrle was perfect through 15 innings. Starting on July 18, when Nick Markakis flied out to center in the 8th, through Thursday’s perfect game and ending in the sixth inning of last night’s game not one batter reached base. As a result Buehrle set the major league record with 45 straight batters retired. Coincidentally the previous record holder Bobby Jenks, tied with Jim Barr for 41 batters, was sitting in the bullpen watching as it happened.

Over those 45 batters Buehrle got nine strikeouts, 22 ground outs, eight fly outs, two line outs and four pop outs. A truly amazing performance. I wanted to partially analyze partially commemorate the streak by looking at the location of Buehrle’s pitches. The image is a little messy, but there are some striking trends.

To begin with look how much he is in the zone. If you are going to go 15 innings without walking anyone, this is what it looks like. Beyond that though, against RHBs, he stayed out of the heart of the plate. One lonely curve down the middle, and everything else in this donut shaped area along the edges of the strike zone. Tons of pitches in the zone, but nothing in the heart, that is a recipe for success. Against righties you can see Buehrle’s plan, cutters and fastballs up and in and changeups low-and-away. That is just a beautiful pattern of pitch types and locations.

As I talked about in my Washburn one-hitter post you can’t get a no-hitter or one-hitter without lots of help from the defense. So here are all the outs at US Cellular during the streak, the one Markakis fly out from the 18th and all his balls in play during the perfect game.

You can see that amazing DeWayne Wise catch. Here are all his outs at the Metrodome during the streak.

Only one ball in play out of the infield and no line drives.

As a whole, notwithstanding the Wise catch, these were some very fieldable balls in play. Only 2 LDs over 36 balls in play, probalby a result of Buehrle staying out of the heart of the plate. In addition, out of 12 fly balls four were infield fly balls, the easiest balls in play to field. Three of these came on those up-and-in fastballs and cutters to RHBs, so Buehrle gets credit for those very fieldable balls in play. Still the defense needs to be given credit, 22 ground balls and all outs, nice work.

An amazing 15 innings for Buehrle, and wonderful for us that we could watch this piece of baseball history.

Over the weekend Tom Tango linked an interview Brian Bannister did talking about how he used PITCH/x information to change his approach this year. It is a really interesting interview that I urge you to listen to. This morning Harry Pavlidis posted a great breakdown of the things Bannister talked about. I wanted to expand on a couple of things Harry looked at.

As some of you might know Bannister is a student of sabremetrics. Back in 2007 when he had a great ERA, build largely on a unsustainably low BABIP, he understood what was happening and that his current approach would not work going forward. In the interview he talked about going into 2008 looking to increase his strikeout rate to decrease his FIP. To do so he tried to work on his four-seam fastball. As we know it didn’t work. His strikeouts didn’t increase, his BABIP regressed and he had a very poor 2009. So going into 2009 he realized the other way to increase his FIP would be to increase his GB rate and thus decrease his HRs. So going into 2009 he wanted to stop using his four-seam fastball as much and replace it with a cutter that he had throw in the minors but given up. The cutter has less ‘rise’ and gets more ground balls. He also talked about changing the grip on his changeup to throw a power change, which results in a changeup with less ‘rise’ and thus get more ground balls.

I wanted to look at this change in approach of replacing his four-seam fastball with a cutter as his main pitch and the change in his changeup. First I went through and reclassified his pitches for this year and last year since both BIS and PITCHf/x are mislabeling his cutter.

+--------------------+-------+-------+
|                    |  2008 |  2009 |
+--------------------+-------+-------+
| Four-Seam Fastball |  0.58 |  0.23 |
| Cutter/Slider      |  0.20 |  0.51 |
| Changeup           |  0.13 |  0.16 |
| Curveball          |  0.09 |  0.10 |
+--------------------+-------+-------+

I had a hard time telling the difference between his cutter and slider. To the best of my understanding I think in 2008 most of those are sliders and in 2009 mostly cutters. So the big difference is in exchanging his main pitch from a four-seam fastball to a cutter, just as Bannister said in the interview. Here is a comparison of his 2008 four-seam fastball to his 2009 cutter to see the results of this shift.

+--------------------+-------+-------+-------+-------+
|                    |  zone | whiff | oswing| GB/BIP|
+--------------------+-------+-------+-------+-------+
| Four-Seam Fastball |  0.59 |  0.11 |  0.20 |  0.30 | 
| Cutter             |  0.58 |  0.10 |  0.24 |  0.47 |
+--------------------+-------+-------+-------+-------+

It looks like he can hit the zone and get whiffs and out-of-zone-swings at roughly the same rate with his cutter that he did with his four-seam fastball, but it gets about half again as many ground balls. That is an exchange anyone would be happy to make.

Bannister also mentioned his new power change that is a little faster has less rise, or vertical movement.

It is working, the GB rate on his change has risen from just under 50% last year to 69% this year.

Overall Bannister’s transition has been great. He is striking out and walking about the same number of hitters, but instead of being a 40% GB pitcher he is a 48% GB pitcher. That is a huge difference, which should result in many fewer HRs and probably make him a league average pitcher.

I think this is really cool. Bannister looked at his stats, understood something had to change, went about changing it and so far it looks like it has worked. I am very happy for him.

Over the weekend Matthew noted that the only starter who comes close to getting as many swinging strikes as called strikes is Rich Harden. He does it with just two pitches, a four-seam fastball and a changeup. Sliders and curves, generally, get the most swinging strikes so that makes Harden’s feat that much more amazing. Harden used to throw a splitter and a slider as well, but gave them up in hope of decreasing his injuries.

Harden four-seam fastball averages about 92 mph with over 10 inches of ‘rise’. The fastball has a 18.7% whiff rate (misses per swings). Among starters only Ted Lilly, Jonathan Sanchez, Justin Verlander and Max Scherzer have fastballs with a higher whiff rate.

His changeup is an even bigger reason for his huge number of swinging strikes. Hitters miss 48% of the time they swing at his change. As Harry Pavlidis highlighted that is the third most of any pitch in the game, and tops amongst changeups. The changeup has about 8mph separation from his fastball, with less tail and ‘rise’. Because he lacks any other secondary pitches he throws the changeup almost equally to lefties and righties, about 37% of the time. Harden starts about a third of of his at-bats with the changeup and has had better results when he pitches it before his fastball (1.3 runs per 100 above average) then when he throws it after (0.6 runs per 100 above average). This is the opposite of what we saw with Tim Lincecum. Hitters must be expecting the fastball on the first pitch of the at-bat and the changeup trips them up.

Harden’s results so far this year have been poor, but that has largely been driven by his high BABIP and HR/FB. His K and BB rates are inline with his career numbers, so going forward we should expect him to be very good.

According to this DL tool it looks like Harden has spent 26 days on the DL this year and 38 in 2008 way down over from 100 days in both of 2007 and 2006, so it looks like getting rid of the splitter and slider may have helped him stay healthy. That left him with two pitches, both of which rack up tons and tons of swinging strikes. It as a testament to how those pitches are that he can succeed as a starter with just the two pitches.

Before the season Dave Cameron looked at the trend of some teams putting a greater emphasis on outfield defense. He specifically noted that the Mariners, Rays and Orioles were planning on effectively playing three center fielders. The outfield defense plan has worked out well for the Mariners and Rays, but not for the Orioles.

Ichiro Suzuki, Franklin Gutierrez and Endy Chavez were just as good as advertised and the Mariners limited defensive butchers Wladimir Balentien‘s and Ken Griffey’s time in the field. Chavez is out for the season with an injury but his playing time has largely been replaced Ryan Langerhans, also a very good defender. By UZR’s measure the Mariner’s outfield defense has saved 30 runs over average, playing a big part in keeping them in AL West contention.

The Rays outfield has been worth 25 runs over average, second only to the Mariners. Their outfield innings have gone to equally to Carl Crawford, B.J. Upton and the Gabe platoon, with a little sprinkling of innings from Ben Zobrist. All of these guys have played well this year, right in line with their career UZRs.

For the Orioles things have not gone as well, giving up an additional 20 runs in outfield defense. Nick Markakis and Adam Jones are two of the game’s up-and-coming star outfielders and their futures are bright, but their defense this year has not been good. Together they have given up over 15 runs, but it is important to remember this does not mean we should expect them to be poor fielders going forward. UZR and other defenses metrics, by all accounts, take huge numbers of innings before they become predictive and are subject to big fluctuations even of the course of an entire season.

Markakis and Jones came into this season with great fielding numbers in 2008. Markakis at +11 UZR/150 in over 1300 innings in RF and Jones at +11 UZR/150 over 1000 innings in CF. This year Markakis is -16 UZR/150 over 760 innings in RF and Jones -12 over 600 innings. This new data put together with the old tempers our expectations; maybe Markakis is only a +5 to +0 corner outfield and maybe Jones is just a +5 to +0 center fielder. With so few innings it is also important to weigh our projections going forward with scouting reports, at least one set of which calls the two great defensive outfielders.

Luke Scott and Nolan Reimold picked up most of Felix Pie’s, the third member of the Orioles planned three CF outfield who couldn’t hit enough to justify his playing time, innings and have played poor defense. Scott has a history as a +5ish corner outfielder and Reimold was projected as an average corner outfielder, so their performance should get better. Going forward we shouldn’t expect the Os outfield to play as bad as it has, but we can’t expect them to play like three center fielders either.