Archive for Daily Graphings

Rivera’s Cutters Working the Count

We’re pleased to welcome Albert Lyu as the newest writer for FanGraphs. He impressed us with his submissions to the Community Blog, and so we’ve found a spot for him here on the team. We think you’ll enjoy having his work published here regularly.

Dave Allen has talked at length about Mariano Rivera’s cutters and how well he locates them, but it’s always interesting to analyze what many consider to be the greatest pitch in the game. I don’t believe that there is any other pitch in the game right now that can be used so exclusively yet so dominantly the way that Rivera uses his cutter.

We know that Rivera has pinpoint control and likes to work the outer and inner edges of the strikezone against both right-handed batters and left-handed batters. We also know that Rivera is great at working the count, rarely getting to three balls. Combining both of these ideas, can we figure out how Rivera works the count based on the locations of his cutters?

I took all of Rivera’s cutters since 2007 and split them into different count situations not including the 3-2 count: first pitch, behind in the count, ahead in the count, and with two strikes. Borrowing Dave’s terminology, I wanted to see if Rivera’s cutters exhibit a bimodal distribution consistently and how the pitches are distributed differently around the inside and outside edges of the strike zone based on the count situation. First up, let’s compare the first pitch of an at-bat Rivera throws against right-handed hitters to left-handed hitters (all plots are from the catcher’s perspective):

Keep in mind that Rivera’s cutter moves toward LHH and away from RHH. It’s clear that Rivera likes to start off an at-bat by throwing his cutter in the same location, throwing outside to RHH consistently but occasionally going inside as well. Against LHH, Rivera almost exclusively throws inside on the very first pitch. Notice also that most of these pitches are within the strikezone, again, mostly outside to RHH and inside to LHH. Let’s take a look at how Rivera locates his cutter when he is behind in the count with more balls than strikes:

In this case, Rivera forms more of a horizontal bimodal distribution than on the first pitch, locating cutters on the inside and outside edges of the strikezone against both handed batters. He also consistently hits the strikezone, but isn’t afraid to go inside on LHH and out of the zone even when he’s behind in the count. Let’s compare this with his cutters when he’s ahead in the count with more strikes than balls:

Here, we see a very clear bimodal distribution, where Rivera works both the inside and outside edges to both RHH and LHH. Against RHH, he loves to go outside again, but goes inside a lot more than he does on the first pitch. Against LHH, he goes inside more, but also goes outside a decent amount. What amazes me about this specific graphic is that the middles of each of the four hot spots of Rivera’s cutter locations are bisected by the border of the strikezone, whether it’s the inside border or the outside border. This speaks to the uncanny control that Rivera has and how he loves to attack either edge of the strikezone when he’s ahead in the count, likely inducing both called strikes and swinging strikes. Finally, let’s look at Rivera’s cutter locations when he has two strikes:

These look similar to the plots when Rivera is ahead in the count except for two significant differences. Against RHH, Rivera goes higher in the zone, especially up and inside in addition to the outside edge. Against LHH, Rivera goes outside much more than he goes inside, very different from the other count situations against LHH where Rivera goes inside more. With two strikes, both RHH and LHH should expect the outside cutter most of the time.

Just looking at traditional statistics will appropriately show how dominant Rivera has been in his career (2.21 ERA, 8.2 K/9, and 1.00 WHIP in over 1145 IP). However, the plots above tell us how he has achieved such success: by living on the black against both right-handed and left-handed hitters and being able to consistently hit his various spots so that hitters are forced to swing at difficult pitches no matter the count.


Troy Tulowitzki: NL MVP?

On Friday, we recorded our latest version of the podcast, which focused on the postseason awards. During the discussion, I suggested that I’d lean toward voting for Troy Tulowitzki, and that was before he launched two more home runs on Saturday. His numbers in September are just crazy good(.357/.407/1.000, .577 wOBA), and he’s a big reason why the Rockies are right back in the NL West race. But I’m not throwing my support behind him just because he’s bashing the baseball of late; I think there’s a pretty decent argument to be made that he’s been the league’s best player, even after accounting for the time he spent on the disabled list.

I think we can probably all agree that he’s been the best player in the NL on a rate basis. His .420 wOBA ranks second to Votto in the NL, and he’s a shortstop. Yes, his numbers get a boost from Coors Field, but his park adjusted wRC+ is still a fantastic 158, and translates out to 42 runs above average per 600 plate appearances. Votto’s 174 wRC+ translates out to 53 runs above average per 600 PA, leaving a gap that is easily overcome by the difference in scarcity between SS and 1B.

Of course, Votto actually has 600 PA (611, actually), while Tulowitzki has racked up just 476 trips to the plate this year. While Colorado’s shortstop has been the league’s best player when he’s been on the field, Votto has a significant edge in playing time, which is why he’s right there with Ryan Zimmerman at the top of the WAR leaderboards, and why he’s likely to win the award.

Even with Votto holding a +1 win advantage by WAR, I still think a vote for Tulowitzki is justifiable. WAR is more blunt hammer than precise chisel, so while it does a great job at telling you whether a player is good or bad, it is not designed to be used to separate out small differences among players having similar seasons. Votto’s +7.0 WAR isn’t so much better than Tulowitzki’s +6.1 WAR that we should definitely say that Votto has been more valuable. The best interpretation of those two numbers is that both have been fantastic, and that there’s room for discussion about which one has been better.

While we’re obviously big proponents of the usefulness of Wins Above Replacement, we do not encourage the use of it as a definitive ender of discussion when the subjects are within the margin of error. It’s one thing to use WAR to declare that Votto has clearly been better than, say, Aubrey Huff, but its another to state that it is perfectly accurate down to the decimal point. Votto and Tulowitzki are both good candidates, as is the always overlooked Ryan Zimmerman. To me, it is not nearly as clear cut which one should take home the trophy as it is over in the American League. I think reasonable cases can be made for all of them, and given the level to which Tulowitzki has been performing, it’s likely that the gap between he and Votto will decrease even more before the season is over.

Quantity or quality? Tulowitzki has the latter, having shown himself to be the NL’s best player when healthy. Does the time spent on the DL hurt his value? Certainly. But I think he might be the NL MVP anyway.


Prospects Chat – 9/20/10


Four Factors: Kemp’s Down Year

Barring one of the most unlikely 60 plate-appearance-runs in baseball history, Matt Kemp’s 2010 season is bound to go down as his worst full season as a Major Leaguer. After putting up wRC+’s above 115 and wOBAs above .349 from 2007-2009, Kemp has a .314 wOBA and a 97 wRC+ in 618 plate appearances this season, which is about 25 runs worse on offense than last year. As Kemp has served as one of the Dodgers’ franchise players, the drop in performance has to be particularly worrisome for those in the LA front office.

Diving slightly beneath the surface, we can see that the culprits aren’t exactly his power stroke nor his ability to reach via the walk. Kemp’s 7.9% walk rate is nearly a percentage point better than his career average and his .181 ISO is exactly in line with his career mark. That leaves the other two factors of hitting – strikeout rate and BABIP – as culprits for this decline. Indeed, Kemp has seen has BABIP fall down to earth, a league average .299 mark after posting .340 BABIPs or better from 2007-2009, and his strikeout rate has skyrocketed from 22% to 28%.

Just over half of this decline can be attributed to Kemp’s relative BABIP issues. The nearly 40 point drop in BABIP has cost Kemp nearly 30 points in wOBA according to the Four Factors method. That leaves the other twenty points remaining between his .314 wOBA and his typical .360-.370 wOBA to his strikeout rate, which is confirmed by the method.

The difference doesn’t really get us much closer to understanding why Kemp’s numbers are down, we just know with which components of the game he’s struggling. Part of it is probably luck and random variation, but I believe a more significant cause can be seen in Kemp’s plate discipline numbers. Kemp thrives on making excellent contact, as that spurs not only his power production but also his excellent prior numbers on balls in play.

The best way to make solid contact is by swinging at pitches that are the easiest to hit – ostensibly, those that fall inside the strike zone. Kemp has had issues with that this season, as his Z-Swing% has fallen from 72.5% in 2008 and 68% in 2009 to 64.5% in 2010. Meanwhile, he’s still swinging at over 30% of pitches outside of the zone – it’s not simply a matter of Kemp swinging less. Combine this lack of swings at pitches in the zone with a lack of contact on pitches in the zone – a Z-Contact rate down by five percentage points this season – and Kemp is putting roughly the same amount of pitches in play, but fewer of them are of the easier kind to hit- the inside the strike zone variety.

It’s possible I’m missing something here – say, a hitch in Kemp’s swing – that I would have had to see more than a select few of Kemp’s plate appearances this season to adequately judge. But the numbers definitely suggest that Kemp’s struggles are due to striking out more and reaching on fewer balls in play – that’s hard to argue. Kemp’s plate discipline numbers suggest that he’s not hitting the more hittable pitches he’s seeing this year, which is making it difficult for him to maintain the brilliant BABIPs that have sustained his hitting lines to date. Regardless whether this is the case, Kemp and the Dodgers need to figure the reason why quickly, because Kemp is simply not a productive player if his bat continues to play at this level.


The Mariners’ Catchers Can’t Hit

In a season defined by Murphy’s Law, the Mariners’ inability to field a worthwhile hitting catcher ranks lower on the repair list than it would in most years. The Mariners’ collection of backstops has combined for a .246 wOBA. As faith has it, the 2005 Seattle club was the last team to have a catching staff so inept at the plate that they combined for a positional wOBA under .250. In other words, Seattle has seen this movie before.

The 2005 edition of Seattle’s finest highlighted their patent on Everlasting Gapstoppers. Miguel Olivo, Pat Borders, Yorvit Torrealba, Rene Rivera, Wiki Gonzalez, Dan Wilson, and even Miguel Ojeda contributed to the mess. This year’s group is more concise, with Adam Moore being the most responsible. Moore and Olivo actually have more in common from their awful seasons than one would suspect. Moore is older now than Olivo then, but their respective lines stack up well:

Olivo (’05): .151/.172/.276
Moore (’10): .169/.199/.260

Olivo and Moore each struck out in nearly a third of their at-bats and, while Olivo walked once for every 12.5 strikeouts, Moore is walking once every 11. Don’t mistake Moore as the only contributor to the awfulness. Rob Johnson and Eliezer Alonzo both own wOBA lines below .265, and Josh Bard’s .303 shines bright in comparison.

The free agent catching market is rarely strong but the Mariners would be hard-pressed to find a downgrade. After all, their wOBA is closer to Mario Mendoza’s career mark (.231) than Paul Bako’s (.275). The M’s won’t respond like they did in 2005: by drafting Jeff Clement during the season then adding Kenji Johjima in the winter. Jack Zduriencik has shown a savvy for trades, so maybe the answer arrives from that form. Whatever happens, history suggests the 2015 Mariners’ backstops are going to be atrocious.


No Free Passes Against Lopez

Of the 133 qualified relievers this season, only two have managed to compile walk rates better than one per nine innings pitched. Those two are Edward Mujica of the Padres and tonight’s featured pitcher: Wilton Lopez of the Astros. Lopez’s 0.73 BB/9 rate in 62 IP leads all relievers, and it makes him a rarity: a productive reliever who doesn’t rely on the strikeout. Lopez is the only reliever in the top 25 in reliever FIP to with a strikeout rate below 7.0 K.9.

Rate starts are nice, but when it comes to players like Lopez, it can be easier to understand just how well they are pitching with raw numbers. Lopez has walked only five batters all season of the 244 that have faced him, and that includes one intentional walk.

Lopez “pounds the zone” to a degree, to use a favorite cliche of announcers. According to Pitch F/X data, Lopez hits the strike zone with 51.6% of his pitches, a mark exactly five percentage points above the league average. That puts him in the upper echelon of qualified relievers, ranking 11th. However, simply hitting the zone often is no guarantor of limiting walks. The walk rates of those in the top 10 in zone% range from 1.25 (Rafael Betancourt of Colorado) to 4.09 (Robinson Tejada of Kansas City), and although we see more in the 1-2 range than 3-4, there’s more to avoiding walks than simply hitting the zone.

The key, at least for Lopez, appears to be the ability with which he can draw strikes with both his two-seam and four-seam fastballs. Lopez has thrown 448 four-seamers and 244 two-seamers this season, and both pitches have fallen for strikes over 68% of the time, whether it’s of the contact, called, or swing-and-miss variety. It should come as little surprise, then that when we take a look at his pitch selection by count, his go-to pitch has been one of his fastballs.

In the six (!) 3-0 counts that Lopez has seen this seasons, he’s thrown the four-seamer four times (three strikes), and the two-seamer twice (both strikes). In the eight 3-1 counts, we see seven fastballs (five strikes) and one changeup (for a strike). Lopez has thrown a more substantial 20 pitches in a full count this season. Again, the fastball has been prominent, featured 18 out of 20 times and thrown for a strike a remarkable 17 times. He’s also gone to his other two pitches, the slider and the changeup, one time each, with the result each time being a foul ball.

Most remarkably, of the 20 pitches Lopez has thrown in these full counts, 18 of them have drawn swings. Only two of them have been whiffs, but it’s still remarkable that only twice have hitters even left it up to the umpire to call the pitch a ball. When we take a look at the location chart from Texas Leaguers.com, we see that Lopez has been near the zone with a majority of his full count pitches.

These graphs aren’t perfect at predicting an umpire’s call based on location, but I wouldn’t bet that 90% of these pitches would be called strikes, and swinging 90% of the time or higher is likely far from the optimal strategy. That said, most of these pitchers are close enough that it wouldn’t be worth leaving the decision up to the human umpire. Lopez’s excellent control with the fastball – roughly 10 percentage points more strikes with his fastball than the league average – keeps hitters on their toes and expecting strikes.

Lopez has been extremely successful as a reliever this year, posting 1.2 WAR to date, and there’s no reason to think that he can’t continue this success. His arsenal – particularly, the presence of a changeup – suggests to me that he may be worth a shot in the rotation, but none of the current Astros starters are warranting the boot at this point. Regardless of what role the Astros use him in, Wilton Lopez should be able to provide the Astros with quality pitching for the next five seasons, chiefly due to his excellent control of the fastball.


Worthwhile Weekend Reading

For your enjoyment:

The good writing entry

Jonah Keri’s A Not-So-Brief History of Pitching Injuries

Easily one of the finest pieces of baseball writing and reporting produced this year. The plethora of quality here left me feeling jaded as I decided which paragraph to display in order to entice your click-through. In the end, I think this quote by Barry Zito exemplifies that he is more aware than expected:

“Baseball has always been the good old boys sport,” said Zito, whose father Joe was such a big believer in long-toss that he insisted on a clause in Barry’s first contract guaranteeing that the A’s wouldn’t interfere with his son’s regimen. “You’ve got a lot of old-school guys with old-school methods. It seems other sports will adjust and change with technology, whereas baseball has always been slow to adjust to the times, and to new technologies.”

The hardcore statistics entry

Kincaid’s ZiPS ROS Projections as Estimates of True Talent

This is FanGraphs-related. There is a lot of math there that supports what can be intuitive. The smaller the sample size, the more volatile the results can be. There is a tendency to ignore what exactly the projected numbers stand for. Maybe that is unavoidable, so an occasional reminder like this requires my appreciation.

The Joe Posnanski entry

Amazing Baseball Stuff

A collection of worthwhile baseball writing without Posnanski is like the best of soul music without any Marvin Gaye. Much like the aforementioned Keri piece, this much quality spread throughout this much quantity produces an insane overload that we should be more thankful for than we already are. What I’m saying is that Posnanski is a robot.


Torre Out, Mattingly In

With the Dodgers sitting at 72-75, 11 games out of the NL West leading and facing an elimination number of only 5, it’s safe to say that their season is over. Now along with that, Joe Torre’s managerial tenure in Chavez Ravine has also come to an end. Replacing Torre will be longtime bench companion and former Yankee Don Mattingly, who will get his long-awaited chance to manage.

Torre has seen his fair share of criticism of late, particularly in regard to his bullpen management (Scott Proctor immediately comes to mind) and his lineup construction. However, Torre’s teams have done fantastically well, as the last time a Torre led team failed to reach the postseason was in 1995, when the Cardinals finished 4th in their division and missed the playoffs in a strike-shortened season. Every single one of Torre’s Yankee squads reached the postseason, winning six AL championships and four world series titles, and Torre’s first two Dodger teams won the NL West.

Of course, the immediate counterpoint to the performance of Torre’s teams is the fact that they were, at least in New York, ridiculously talented. His Dodgers teams, too, at least relative to their division, were quite good. It brings up the question of how much we can credit a manager for winning with supreme talent. It’s easy to give credit to Buck Showalter when his relatively talent-less Orioles go on a winning streak; it’s not as simple when Joe Maddon has the deepest stock of position players on the continent or when Tony La Russa has the best player on the planet.

The question of whether or not Torre will take another managing job is purely based on his personal life. The question of whether or not another team should offer him a job is something different entirely. At 70, Torre’s age is a bit of a question mark and he certainly made some questionable decisions with the Dodgers: ruining Scott Proctor’s arm, batting Matt Kemp eight for a majority of 2009, and benching Manny Ramirez during the Dodgers’ last gasp this season are just a few. However, every manager is prone to these types of decisions, and although it’s easy for us on the outside to sit back and cry foul on them. Whether or not Torre is fit to manage is a decision that is much better left to those inside baseball than to us outsiders, even though on the surface it can seem apparent that Torre is hurting the team more than he is helping.

Now, Don Mattingly will take over a Dodgers team that is undoubtedly on the decline and is mired in an extremely ugly divorce case. This isn’t going to be an easy first situation for Mattingly, but at the very least it should serve as a decent test of his ability in the clubhouse.


Adam Dunn, Hall Of Famer?

In his post about Jason Heyward yesterday, Matthew Carruth referred to Adam Dunn as a “maybe” Hall-of-Famer. When I read that, I assumed Matthew had taken leave of his senses, or was making some kind of joke that I didn’t quite get. After all, Dunn has a career +27.7 WAR through age 30, and the established bubble for HOF players is currently about +60 WAR. It is highly unlikely that Dunn will be able to double his value in the second half of his career, especially given his skillset.

But, the more I thought about it, the more I realized Matthew was right – Adam Dunn very well may end up in Cooperstown if he can stay healthy.

Career HR totals are definitely one of the magic numbers that get voters attention. The guaranteed entry barrier used to be 500 before the offensive explosion of the 1990s. Only 25 guys in history have ever hit 500+ home runs, and almost all of them are or will be in the Hall of Fame. Now that non-superstars like Gary Sheffield and Rafael Palmeiro have joined the list, its exclusivity has been diminished, but it’s still the kind of number that will draw notice, especially for a guy who has never been linked to PEDs.

Well, Adam Dunn is only 149 home runs away from 500, and he’s averaged 35 HR per 600 PA in his career so far. Barring injury or a very early collapse of his skills, 500 homers seems like a very easy target for Dunn. That’s averaging just under 30 homers a year for the next five seasons, and the last time he hit fewer than 30 long balls in a year was 2003. His bat may begin to slow down, but he’s got enough power that he can lose some and still crank out that piece with relative ease.

Given just how bad defensively he was in the outfield, however, he’ll almost certainly be the worst player ever to reach that milestone. Even with the power, his offensive production has always been just good, not great, and he’s one of the worst defenders and baserunners of his generation. In terms of actual value, there’s almost no way Dunn will deserve to go into the Hall, but I’m not sure that will matter.

If he hits 500 homers, and is presumed to have done so while “clean”, there will be support for his candidacy, especially if he’s going up against other guys who have tested positive or been linked to steroids. If he gets anywhere near 600 and keeps away from scandal, he’s a mortal lock, and it’s not impossible to see Dunn launching 250 homers over the remainder of his career.

So, I have to conclude that Matthew is right – Adam Dunn is a “maybe” Hall of Famer. If he stays healthy for the next 5 to 10 years, he’s got a pretty decent chance at putting up the kinds of numbers that voters will take notice of. He won’t deserve to go in based on total value, but the sexy numbers just may be enough. He does the things that voters like, and the aspects where he fails, there is little emphasis. I can’t believe I’m saying this, but there’s a decent chance that my kids might walk past Dunn’s plaque in Cooperstown some day.


Jeremy Guthrie as Baltimore’s Bright Spot

2010 has been miserable for the Orioles. The vaunted young position players have been disappointing as a group, and no Orioles position player has reached three WAR yet this season. In particular, Adam Jones‘ incipient superstardom looks to be on hold yet again, as he’s not even at two WAR yet. The pitchers have been even worse. However, for those people who see ERA as the go-to pitching stat, Jeremy Guthrie (3.74 ERA over 190 innings so far) might seem to be a “bright spot” in this otherwise wasted season.

If you’re reading this, you probably know that ERA is not considered to be a first-call pitching stat around here. FanGraphs uses FIP when calculating Wins Above Replacement in an attempt to avoid the problem of giving pitchers credit or blame for the fielders behind them and batted ball luck. It’s not perfect, but it is more stable from season-to-season than ERA/RA. FanGraphs has Guthrie at 2.1 WAR currently, which is helpful, of course, but if that’s a team’s bright spot… His decent rating is mostly due to his endurance, given his number of innings pitched, as his 4.39 FIP is actually below average (FIP is always scaled so that it is equal to league ERA, which is currently 4.10). While Guthrie does “eat innings,” the differential between his ERA and FIP indicates he’s likely been the beneficiary of good fortune regarding the fielders behind him and some batted ball luck (.265 BABIP). But even that FIP is a bit deceptive, as Guthrie has had a career-best 7.8% HR/FB ratio this season. xFIP accounts for this by normalizing his HR/FB ratio to league average, which puts his xFIP at 4.90 — the fifth-worst among qualified starters in baseball this season. If we widen the net a little further, we find his xFIP to be about the same as less bright spots such as David Bush, Kyle Kendrick, Brian Bannister, and Tommy Hunter.

Guthrie’s main skill is that he doesn’t walk hitters, and his 2.23 BB/9 is the lowest of his career. Guthrie’s strikeout rate has never been impressive, but 2010 is his second season in a row with strikeout rate under five per nine innings. While Guthrie is getting more hitters to swing at pitches outside the zone than ever before (27.6%), this probably reflects the overall higher O-Swing percentage in the league more than an improvement on Guthrie’s part — he’s still below average. Moreover, when he does get hitters to swing at pitches outside the zone, they’re making contact with those pitches at a greater rate than before. Pitchers with Guthrie’s stuff usually survive by getting good groundball rates (which FIP doesn’t take into account), but Guthrie has never been exceptional at getting grounders, despite his this season’s improvement over 2009.

This sort of thing isn’t atypical for Guthrie. In his 2007 and 2008 seasons his ERA was much better than his FIP and xFIP. In 2009 his luck “evened out,” as it were, as his 5.04 ERA was relatively close to his 5.31 FIP and and 5.22 xFIP. One season isn’t a great indication of true talent on its own, but when looking at Guthrie’s 4.90 xFIP, it isn’t that surprising coming off of a 4.54 in 2008 and a 5.22 in 2009, and is probably a decent indication of his “true talent.” While Guthrie’s straight FIP of 4.39 is the best of his career, his 4.41 FIP and 4.53 FIPs in 2007 and 2008 came in seasons where the league average was 4.47 and 4.32, respectively. This season, the league average is 4.10, making his performance that less valuable than it might appear to be.

Guthrie was an excellent waiver pickup by the Orioles in 2007, and gave them a couple of decent years at the league minimum. Even after his poor showing in 2009, it was a low-risk move for Baltimore to offer him arbitration given his previous performance and the reasonable likely award (they settled without a hearing). Given his overall WAR of 2, it was a decent investment, and the Orioles have had success when he has started, at least in terms of preventing earned runs. However, once taking into account his poor strikeout rate, his good fortune on balls in play, HR/FB ratio, his age (Guthrie will be 31 in April), and the Orioles’ place in the “success cycle,” the notion that Guthrie will be part of the next contender in Baltimore seems far-fetched, however shiny his ERA may appear. Even if the Orioles keep Guthrie around in order to “stabilize the rotation” (whatever that means), a closer look at his performance this season shows him to be a very faint bright spot, indeed.