by Myron Logan

Bloomberg Sports is launching new fantasy and professional baseball products. Today, they unveiled their products at their NYC headquarters, which had my Twitter page abuzz. David Appelman has some really cool screenshots at FanGraphs, including this one:

I’m not sure how much of what they are doing hasn’t already been done on the ‘net, but the presentation sure looks nice. Plus, the ability to easily toggle through all MLB players is a huge bonus. This is the major league product, so I don’t think we’ll be seeing much more of it. Sure is fun to look at, though.

by Mike Rogers

Over at Project Prospect, there’s a nice and informative scouting report on the swing of 1st year pro, James Darnell. I had a three part series that looked at the 2008 college bats the Padres took in June of 2008 and I loved everything about James Darnell’s college statistics. I’m still a big time Darnell fan and think the former South Carolina Gamecock can be, at the very least, an average right fielder in the majors (not sure he can stick at third).

The scouting report makes a very interesting swing comparison: Justin Upton. Now, Steve Carter (who is using an alias but did play college baseball), cautioned he’s not expecting near the offensive production of Upton, but that they do have similar swings. An excerpt from the conclusion:

Darnell has the tools, ability, and swing to be a very solid offensive 3rd baseman. Eighteen home runs and a .998 OPS between Low-A Fort Wayne and High-A Lake Elsinore is certainly nothing to scoff at. But, the real test for Darnell will come at the Double-A and Triple-A levels. Then we will find out if he is just a polished college hitter pounding A-ball pitching, or a legitimate prospect. If he can improve his upper body pattern and make a few other refinements, he could go from a potential Casey Blake with patience type, and move toward his ceiling of a pre-2009 breakout Ryan Zimmerman with a higher on base percentage.

So, there’s some things to like. Casey Blake with patience definitely isn’t anything to scoff at and Ryan Zimmerman with better OBP is very good outcome as well. There’s a lot to like about Darnell.

Now, while you’re perusing the Project Prospect website, feel free to read my PITCHf/x look at Jake Arrieta, a Baltimore Orioles right-handed pitcher. Yeah, it’s self-promotion and I will not be ashamed of it! Why should you read about an Orioles pitching prospect? Well, because I’ll be writing these types of pieces for Project Prospect when prospects make their debuts. So, I want you to get used to the format for when I pen my piece on Mat Latos (which I already looked at his debut right here for FF) through the PITCHf/x lens.

by Mike Rogers

Clayton Richard made his San Diego Padres debut on Saturday against the Milwaukee Brewers after being acquired by the Pads for Jake Peavy (trade valuation coming from me sometime tomorrow on Beyond the Boxscore where I’ll be joining the BtB team of writers) on Friday. Before we jump into the PITCHf/x, lets look at the start through the numbers:

5.2 IP, 2 H, 1 ER, 3 BB, 5 K, 90 pitches (49 strikes), 8 ground outs, 4 fly outs.

Now, onto the graph-y goodness. First up is the release points (click images to enlarge):

Richard stands 6-foot-5 and weighs in at 240-pounds, but he was consistently “releasing” the baseball at just under 6-foot. So, he comes from a 3/4 arm slot. But, he repeats his delivery well and comes from a very consistent release point.

Now, the strikezone plot:

He only threw 49 strikes out of the 90 total pitches, so he was a bit wild. The view is from the catcher’s perspective, so he worked away from right-handed hitters for most of the night, but given the amount of right-handers he faced, that’s not surprising.

And speaking of how he attacked right-handed and left-handed batters, here’s a couple pie graphs:

Typical platoon split: drops the change-up versus left-handed hitters, and picks it back up against right-handers. Overall, though, he worked the fastball (either four-seam [FF] or two-seam [FT]) all night.

And now the pitch flight graph (again, courtesy of Harry Pavlidis):

My breakdown of Clayton Richard differs a bit from Myron Logan’s. I attempted to split up the four-seam and two-seam fastballs as best I could, but I still might be/am probably wrong on some of them. I wound up with 48 four-seamers and 21 two-seamers while averaging 91.7 MPH on the former and 90.4 on the latter. His four-seamer seemed to have a bit more arm-side run, but less sink (minimally). Just keep in mind that I was eye-balling the fastballs and made some judgement calls, so they aren’t anywhere near 100% accurate (and if they are, it’s absolute luck on my part).

As far as his breaking ball goes, MLB Gameday classified three of them as curveballs, but after graphing them out, there was virtually no difference in flight path from his slider. I don’t know if he throws a curveball or a slider, but on the flight paths I went with curveball as I believe that’s what the scouting report on him says. As you probably noticed, I classified them all as sliders on the other graphs — that’s just a little mistake on my part, but they’re still depicting the same data set. Either way, he’s got a bit of “sweep” on his slider and some decent downward movement.

Meanwhile, his change-up has some arm-side run but very little sinking action.

On the whole, Richard looks to have pretty average stuff for a big, tall lefty, but his results speak for themselves thus far: 4.32 FIP in 142.1 big league innings pitched. So, he’s definitely a back-end starter, but he’s got value as he’s probably above-average for a number 5 starter.

Overall, the Padres got a good haul in the Jake Peavy deal.

by Mike Rogers

UPDATE: Unfortunately, as I was doing this this evening, I mistakenly lumped Latos’ curveball’s in with his change-up’s. I have since fixed this, and re-written the paragraph on his change as well as penned one on his curveball. Whoops.

I’m just going to get this out of the way: I did not see one single pitch Mat Latos threw, unfortunately. I was busy and missed both his abbreviated outing in the Future’s Game, as well as his Major League Debut last night against the Colorado Rockies.

Here’s what I know about him: he can throw the baseball. Hard.

These graphs are made possible by Harry Pavlidis and his Pitch Flight tool.

First up, are the release points. Click on both images to enlarge.

That y-axis is in feet. That means, Mat Latos is “releasing” (data’s actually recorded 50 feet from home plate, I believe, so it’s not “true” release point) the baseball from nearly 7 feet in the air. He’s a big boy as it is, and his tall-and-fall mechanics lead to an even higher, over the top release point.

Now, on to the PITCHf/x flight path graphs. What this shows me is the movement of each of his pitches on its way to home plate. This sort of thing is hard to read for some people, so I am willing to answer any and all questions on it later tonight.

The Fastball (the green line) — 56 of his 70 pitches recorded by PITCHf/x were fastballs and all were coded as 4-seam fastballs. He averaged 94.9 MPH on them. From the graph, I see a very straight heater which is what I saw when I graphed out his Future’s Game performance on the message boards over at Project Prospect. I don’t see a lot of downward movement on his fastball from this outing against the Rockies, either, but coming from such a high release, I’m sure it’s difficult to hit. Ideally, though, you’d like to see a fastball that bends or sinks more than Latos’ does.

The Change-up (the red line) — Latos has a big time arm, but I think his secondary offerings are where his bread is buttered. He threw just one change-up and it clocked in at 82.2 MPH. It came from a lower release point (1st base view) than his other three pitches, but that’s likely because he threw just one of them. Good dive on it and a decent amount of tailing action.

The Slider (the yellow-ish line) — He threw only 8 of these at an average velocity of 82.5 MPH. He’s got a little sweeping action on it (bird’s eye view) and has a lot of downward movement, as well. Both his change-up and his slider compliment his fastball extremely well due to their similar release points, but both dive well below his fastball. I really like both of these offerings much more than his fastball.

The Curveball (the blue line) — He threw five of these averaging 81.8 MPH. It seems very 12-6 but with good drop on it. It comes from the same plane and arm slot as his fastball and follows it pretty well, but it’s kind of slurve-y which leads me to believe that his slider and his curveball are the same pitch, just classified differently by gameday. I’ll defer to Harry Pavlidis’ week New Arms Roundup on Beyond the Boxscore this weekend.

Conclusion

First off, I need to say that there really aren’t any “conclusions” from his outings that were recorded by PITCHf/x. He threw just 8 pitches in the Future’s game and then another 70 last night. That sample is too small to draw any hard conclusions. However, it’s enough to give a decent hypothesis on Latos and his stuff. He’s got the tall-and-fall mechanics with a clean arm action and they seem to be repeatable. Some have panned his mechanics but that’s likely due to not being a fan of the tall-and-fall motion. Alex Eisenberg over at Baseball-Intellect said he’s not a fan of tall-and-fall but that he doesn’t see a red flag from Latos, despite that motion.  I trust Alex’s opinion and some others in the Project Prospect forums have expressed opinions similar to Alex’s: not a fan of that style of pitching (would prefer drop-and-drive), but for what it is, it isn’t overly cringe-worthy.

His stuff features a big time fastball in velocity, but I’d like to see him throw more 2-seamers as his 4-seamer is very straight in both horizontal and vertical movement. His breaking ball and his change-up are great compliments to it — it’s easy to see why Latos’ stuff is heralded. Now, if he can only stay on the field.

Edit to my conclusion: I am not sure what exactly type breaking ball he throws, but they seem to be very similar.

by Daniel Gettinger

Tim Stauffer was very impressive on Saturday night.  Facing the Giants, he threw 7 innings, allowing 2 runs and 1 walk, while striking out 7 batters.  Besides the homerun he surrendered to Pablo Sandoval, Stauffer did allow many hard hit balls.

What surprised me the most was how hard Stauffer was throwing.  I had expected him to throw his fastball between 87 and 90 mph, but for his first 60 pitches of the evening, Stauffer was throwing his fastball between 90 and 92 mph.  His average fastball on the night is classified by pitch f/x at 90.5 mph.  Below is his velocity chart (not my favorite presentation, but in this case, it gets the job done) for the game:

Chart provided by Brooks Baseball

Although I was pleasantly surprised both by Stauffer’s results, and the speed of his fastball, I am taking a “cautiously optimistic” approach towards him.  In AAA this season, Stauffer had a 2.90 FIP.  Good, but perhaps a bit misleading.  In 23 AAA innings (not a huge sample size), he struck out 6.26 batters per nine innings, while walking only 1.29 per nine innings.  The low walk total is good, the strikeout numbers not nearly as overwhelming.  What Stauffer has done a great job at is keeping the ball in the park.  His 0.39 HR/9 is ridiculously low, and not sustainable at either the AAA or major league level.

I would love to see Stauffer succeed, and with a 90-92 mph fastball, good control, and some decent off-speed pitches, it certainly seems like he has a solid chance to be a reasonable back of the rotation starter. But I’m not quite ready to get too excited.

by Daniel Gettinger

Dave Allen finds that changeups are most effective when they are thrown between 5% and 12% slower than the pitcher’s fastball.  He reports a 12% differential is no more effective than a 5% differential.  However, once outside that range (on either side), “effectiveness rapidly drops off.”  

A year ago, using a different approach, Josh Kalk examined the same question.  He found the greater the differential between the fastball and changeup, the better.  His results did not give much credence to the theory that too slow of a changeup will help the hitter differentiate between the fastball and the changeup.

Like Allen said, “I am not sure how to reconcile these two different conclusions.”  But, in case you were wondering, Trevor Hoffman’s changeup is approximately 14% slower than his fastball.

by Daniel Gettinger

Following his poor April 26th start against the Pirates, there were rumblings that perhaps something was wrong with Jake Peavy.  While I did not find any conclusive evidence that he was pitching any different than he did last season, people who had actually watched the game thought he was “nibbling” too much.  Pirates hitters seemed to concur. 

On May 1st against the Dodgers, Peavy’s results were much better.  He threw eight innings of shutout ball, struck out eight, and only walked one batter.  Great results, but did he actually throw the ball that much better?  Again, I was not able to watch the game, but the pitch f/x data suggests the only actual differences were pitch selection and better luck.

The first question I had was did Peavy throw the ball harder than against the Pirates?  The answer: NO.  On April 26, Peavy’s average fastball was 93.1 MPH.  On May 1: 92.5 MPH.  Likewise, against the Dodgers, his slider was 2 MPH slower than against the Pirates.

How about nibbling?  Was Jake more aggressive against the Dodgers than the Pirates?   NO.  In fact, he might have nibbled more.  The data found at brooksbaseball.net has a nibbleness score.  Nibbleness is defined as “the arithmetic mean of the distance of each pitch, in inches, from the edge of a normalized strike zone. Lower indicates ‘more Nibbley’.”  The following chart lists Peavy’s nibbleness score for each of his pitches in both games:

	Apr. 26	May-01
	Nibbleness	Nibbleness
Fastball	5.33	4.96
Change	7.42	5.9
Slider	6.64	7.83
Curve	5.62	6.79
Cutter	3.33	5.63

The Pitch f/x data suggests Peavy nibbled less against the Dodgers with his slider, curve, and cutter, but more with his fastball (his most frequently thrown pitch), and his change-up.

Plots of Peavy’s pitch locations also seem to show that he did not nibble less against the Dodgers.  The only difference I see is a tendency to do his nibbling on the right side of the strike zone against the Dodgers, but the left side against the Pirates…

April 26 (Pirates)

May 1 (Dodgers)

So, if pitch speed was actually down, and Peavy was not actually any more aggressive, what did change?  The only thing I can find is a reduction in the percentage of sliders he threw.  Against the Pirates, 31% of his pitches were sliders.  He only threw the slider for a strike 63% of the time.  Against the Dodgers, only 15% of his pitches were sliders.  This time his strike percentage was even worse: 55%.  Instead of the slider, Peavy threw more four-seamers and cutters.  Honestly though, I am not sure what to make of this (or whether there is anything to make of it).

The other thing I found was Peavy was just luckier against the Dodgers.  On April 26, 36% of all balls in play went for a hit.  On May 1st, Peavy’s BABIP dropped to 12%.  Sometimes a little luck is all a player needs to “turn things around.”

by Daniel Gettinger

Jake Peavy has started five games.  He has pitched 31.1 innings, and has a 5.74 ERA.  Something must be wrong, right?  Well, not exactly.  While I will not try to argue Peavy has been exceptional in 2009, his underlying stats do not demonstrate any conclusive reasons to be concerned.  Pitch FX data does show a possible loss in fastball velocity, and while this is something to monitor going forward, I again am cautious about drawing any conclusions about Peavy’s future effectiveness.

While Peavy has a 5.74 ERA, his FIP, which attempts to weed out some of the luck inherent in ERA, stands at a much more acceptable 4.05.  This difference is mostly attributed to a ridiculously high 0.340 BABIP.  For comparison’s sake, in 2008, Peavy had a 3.60 FIP.  In fact, Peavy is actually striking out more batters per nine innings than he did in 2008 (9.19 v. 8.60 K/9).  What has hurt Peavy is an uptick in walks and home runs allowed.  In 2009 Peavy has allowed 3.73 BB/9 compared to a career rate of 2.92.  His HR/9 currently sits at 1.15 compared to 0.92 for his career (note: the career rate is adversely affected by a 2003 season in which he gave up 33 HR, a number he has never come close to approaching again).

Normally I would not be too concerned about these numbers.  Peavy has not pitched enough innings in 2009 to make any definitive statements regarding a loss of control or a decreased ability to prevent the long ball using the statistics alone.  What concerns me though is the possibility that the increased walks and home runs allowed are a result of decreased fastball velocity.

In 2007, Peavy’s best season, he threw his fastball 62.4% of the time at an average speed of 93.9 MPH.  In 2008, he threw the fastball 58% of the time at an average speed of 92.6 MPH.  In 2009, he has thrown the fastball 52.2% of the time at an average speed of 91.7 MPH.  The horizontal and vertical movement on the fastball have remained fairly constant.  Only the speed of the pitch has changed.

Instead of throwing his four-seamer, Peavy is relying much more on a curve and cutter than he has in the past.  There are two primary explanations.  The first, and the more disheartening of the two, is that Peavy no longer has the stuff to overpower batters with his fastball.  Instead he is substituting towards the curve and cutter, pitches he does not control as well.  Decreased pitch speed could explain the increase in home runs allowed, while throwing more curves might be contributing to the increase in walks.

I prefer to take a more optimistic view.  I do not believe there is ample evidence to conclude Peavy’s fastball is slower than in 2008.  It is very possible that as the weather warms, so will Peavy’s arm.  By June, Peavy may be back to throwing 92-93 MPH on average.  In addition, I believe Peavy is throwing the curve ball more often because he has improved the pitch, not because he believes his fastball is less effective.  Compared to 2008, Peavy’s curve ball in 2009 has had double the vertical movement.  Perhaps he is experimenting with the pitch because he knows eventually it will become a useful and effective part of his arsenal, protecting against a true loss in velocity down the road.

It may be true that Peavy no longer has the stuff to overpower batters with the fastball.  But, at this point, I feel such conclusions are premature.  In the past, Peavy has had 5-game stretches where he appears to have performed poorly.  And he will have such stretches in the future.  Most of the time, Peavy’s performance in these stretches has been a tad below-average, but also exacerbated by bad luck.  This five start stretch is no different.  Peavy has not pitched as well as we have seen in the past, but a 0.340 BABIP has more to do with his 5.74 ERA than his actual performance.

by Myron Logan

Peavy’s Saturday performance was not overly impressive, by the numbers, as he allowed two runs and walked three in three innings of work. It should of course be noted that we shouldn’t expect him to be in mid-season form, that he was facing a pretty good Canadian line up (that included guys like Justin Morneau, Jason Bay, Russell Martin, and Joey Votto — who crushed a homer off Jake), and that he also struck out four batters in those three innings and allowed only two hits.

Seriously, I don’t think I’ve talked that long about three innings of work on the history of this blog. It obviously has little meaning, from a quantitative perspective. But I figured that since the game was played in an MLB park, I’d take a quick look at Peavy’s PITCHf/x data, and see what comes out. First up is the standard velocity graph:

Peavy’s average fastball velocity was 92.6. He cracked 94 three times, and didn’t throw one below 90. Here’s the old movement graph:

The top left group are fastballs. Far left: changeups. And to the right are his sliders, with a pretty wide variation of movement, both vertical or horizontal. If I’m misinterpreting something here, it wouldn’t be the first time …

Here’s Peavy’s PITCHf/x card from last season, for comparison.

by Myron Logan

Edit: Check Mike Fast’s comment for additional info on PITCHf/x (it agrees with Bell).

In an article today in the UT, Heath Bell said he threw one changeup last year. Since you rarely get an exact number like that from a player*, I thought it be fun to look up what the two other data sources reported on his change.

Josh Kalk’s algorithm , using PITCHf/x data, says he threw 41. It recorded 1209 of Bell’s 1336 total pitches, so if we prorate it, we get 45 changes.

Fangraphs, using Baseball Info Solutions (which I believe uses video to classify pitches), says that Bell threw ~27 changeups.

First of all, out of 1300+ pitches, assuming Bell knows how many changes he really threw, the two data sources are not that far off. It’s not like they have him throwing 200-300 changeups. Anyway, there are (at least) two issues here.

1. Does Bell actually know that he only threw one change? Of course, he does. But is it possible that he threw an assortment of pitches, intended to be non-changeups, that ended up acting just like changes in spin, break, and movement? Should they count as changeups or should they count as whatever they were intended to be?

2. If Bell’s number is correct, why are the data sources off? Well, I think it just shows that it isn’t easy to classify pitches, either with numbers (PITCHf/x) or video (BIS).

*Can you imagine Jake Peavy saying, “Well, last year I threw 287 curves (10.5 percent of my total pitches) with an average mph of 79.1, an average x-break of 8.3 inches, and an average z-break of –2.9 inches. This year, I’m working on getting my z-break to crack –4.”