22 replies to this topic

[Sprowl]

Erik Bedard suffered some bad luck on groundballs and bloops, but his stuff looks pretty sharp -- only two of the Indians' hits were hit hard. His velocity was OK (average 92 on the 2-seamer and a little bit less on the 4-seamer), and his movement is outstanding, especially when contrasting the fastball and curve. He was working toward a 70-pitch limit, so he made every effort to be efficient. It has been a long time since a Red Sox starter threw this many strikes:



Bedard moved the ball around, but for the most part fastballs were above the belt, and sinkers below the belt or on the outside edge. Bedard has no qualms about throwing the curve for strikes, and its movement is sharp enough so that hitters have to be looking for the curve to get good wood on the ball. It's unusual to see a pitcher with such a sharp contrast between the movement on his three major pitches (the 4 changeups and 1 cutter look like show-me pitches at this stage), and it makes me think that Bedard can get by with just 91-92 velocity.



What did you see in the new guy's stuff on Thursday?

[5dice]

Agreed that he was victim of bad luck, especially in the second inning with ground balls finding holes and that bloop in RF. I felt like he was getting ahead a lot and I will take 0/5 bb/k any day of the week.

Edited by 5dice, 05 August 2011 - 07:42 AM.

[Tangled Up In Red]

Sprowl, do I read the second chart to suggest the two-seamer has similar movement to the curve? Or how does one interpret?

[JMDurron]

I was pretty pleased with what I saw as well. I have few worries about Bedard's capabilities on the mound, it's just a matter of how often he can be on it.

The only real "issue" in his outing was Bedard's inability to anticipate Gonzalez's "Kevin Millar on Mother's Day" decision to start playing 2B on a groundball with runners on 2nd and 3rd, and therefore he failed to cover 1B. That's not an issue for me as long as it is a rare incident, it struck me as a flukey thing.

[Smiling Joe Hesketh]

I was pretty pleased with what I saw as well. I have few worries about Bedard's capabilities on the mound, it's just a matter of how often he can be on it.

The only real "issue" in his outing was Bedard's inability to anticipate Gonzalez's "Kevin Millar on Mother's Day" decision to start playing 2B on a groundball with runners on 2nd and 3rd, and therefore he failed to cover 1B. That's not an issue for me as long as it is a rare incident, it struck me as a flukey thing.

Yeah, AGon was totally in the wrong there. That's an infinitely routine play for the second baseman; I'm having a hard time faulting Bedard for not covering when the whole world knew that was Pedroia's ball.

[teddywingman]

His curveball is a thing of beauty.

[bosockboy]

What he produced last night will win us a lot of games, and I expect him to get better. His control was fine which is the one thing I was concerned with.

[Ted Cox 4 president]

He finished up strong, retiring seven in a row, with three Ks.

[24JoshuaPoint]

Second start after some time off? I'm pretty pleased.

[luckysox]

I'm not sure why (if) anyone (at least ion this board) thought he'd have trouble pitching here from anything other than a health standpoint. The guy is a top-flight pitcher. He'd be a #1 on about half of the teams in the AL, and a solid #2 on all except maybe the Sox in a year like this year when both Lester and Beckett are awesome. If he stays healthy he's going to be a solid, solid piece of the rotation and make 3 out of every 5 games an almost guaranteed good start. His stuff was really good last night and I expect more of the same going forward, with hopefully fewer bloops and weird defensive plays, and more clean innings.

[dauber23]

I was very pleased with Bedard's stuff and performance. He is very calm and poised on the mound.

His knee seemed fine and his stuff looked very good, especially his curve. He pounds the strike zone.

For what was basically a rehab start, this was a solid A.

If Buch gets back for the Postseason, we may have 4 strong starters in the rotation.

[Eric Van]

His curveball is a thing of beauty.

Never mind that he threw 18 of 23 for strikes -- the rotation he gets on it is amazing.

When you throw a curveball, you are trying to do two different things, the first of which makes the second really hard. First of all, you are of course trying to get on top of the ball to get topspin when you rotate the wrist -- you are elevating the base of the palm at release relative to the fingertips. But you are also trying to square the hand to home plate, to point the fingertips towards third base (if LHP) or first base (if RHP) rather than home plate. That is, you are trying to make the spin axis of the ball perpendicular to the flight of the ball rather than more parallel.

Getting on top moves the break on the ball towards the 12-6 ideal, while failing to do so gives you more of a horizontal break and less of a vertical -- it makes the curve slurvier. But squaring the hand is even more important -- it makes the break bigger.

It is really hard to square the hand up if you get way on top (try it). Because the size of the break is more important than the direction, most MLB curveballs are thrown with the hand squared as much as possible, which limits how much on top you can get. So you'll see a curve like Beckett's with a spin axis of 45 degrees (halfway on top) and measured RPM of 1500 to 2000. That's the RPM in the direction of home plate, not the total actual RPM on the ball, which appears to be 2000-2500 RPM -- the more you square the hand, the more the measured (and effective) RPM approaches the actual spin on the ball. (In a pure slider, where the hand is pointing at home, the rotation of the wrist creates a spin axis which is parallel to the path of the ball and there is relatively little measured RPM at all -- typically about 500. In between, you get slurves.)

In contrast, almost every curveball I've seen that had a very low spin axis (0 to 30 degrees for a RHP, 330 to 360 for a LHP) has relatively low measured RPM. Manny Delcarmen had the classic low-spin-axis curveball, where he got way on top of the ball to get close to pure topspin -- but he hardly squared the hand up at all and would get very low measured RPM. And I would occasionally see visiting pitchers with similar curves. These guys get more 12-6 break, but just not a lot of it.

Bedard is the first guy I can remember seeing who combines an MDC-like low spin axis (getting way on top) with a Beckett-like RPM (squaring way up). The result is a curveball with classic big break but more 12-to-6 action than other curveballs with similar bite. (At its best, Buchholz's curveball has a similar freaky combination, but his spin axis is just a bit lower than usual while his RPM is off the charts -- so he's squaring the hand even more then Bedard while not getting quite as on top.)

Thing of beauty indeed.

(And the one cutter he threw looks like an absolutely nasty pitch. The three changeups look solid, too.)

[Sprowl]

Sprowl, do I read the second chart to suggest the two-seamer has similar movement to the curve? Or how does one interpret?

The chart shows two clusters in the upper right that correspond to the 4-seamer (average +10" vertical, +5" horizontal movement) and the 2-seamer (average +7" vertical, +10" horizontal). There are also 4 changeups with ~+8" horizontal movement mixed in to the clusters. The curveball is the cluster in the lower left averaging -9" vertical and -7" horizontal. I would interpret the chart to means that Bedard's 4-seamer and the 2-seamer are fairly close in movement (but still different enough to give batters problems when the 4-seamer is thrown up and the 2-seamer low and away), and that both of those fastballs are very different from the curve.

[IpswichSox]

Yeah, Adrian Gonzalez was totally in the wrong there. That's an infinitely routine play for the second baseman; I'm having a hard time faulting Bedard for not covering when the whole world knew that was Pedroia's ball.

What's nice about that, too, is that Bedard took it on himself for that play -- in the post-game, when asked about not covering the bag, he said something like he had a "brain fart, I guess." Given all the speculation about Bedard being a moody prick who hates the media, I was half-expecting him to say something passive-aggressive, like "Well, it's difficult for me to anticipate that the firstbaseman would range 30 feet to his right."

[joe dokes]

[snip]

Bedard is the first guy I can remember seeing who combines an MDC-like low spin axis (getting way on top) with a Beckett-like RPM (squaring way up). The result is a curveball with classic big break but more 12-to-6 action than other curveballs with similar bite. (At its best, Buchholz's curveball has a similar freaky combination, but his spin axis is just a bit lower than usual while his RPM is off the charts -- so he's squaring the hand even more then Bedard while not getting quite as on top.)

Thing of beauty indeed.

Maybe that's why he's wrecked both his shoulder and his elbow?

[Eric Van]

Maybe that's why he's wrecked both his shoulder and his elbow?

I don't know about shoulder, but it definitely seems to strain the elbow. I bet if we had pitch/fx data for Steve Stone it would be similar or even more extreme.

[Jnai]

Sprowl, do I read the second chart to suggest the two-seamer has similar movement to the curve? Or how does one interpret?

The movement numbers represent amount of movement due to the spin of the baseball. That chart looks at movement from the catcher's point of view.

From the catcher's point of view, a curveball (from a lefty pitcher) will move down and to the left. So, the cluster of dots in the bottom left of the graph is the curveball.

From the catcher's point of view, a two-seamer (from a lefty pitcher) will move up and to the right. So, one of the clusters of dots in the top right of the graph is the two-seamer. Two seamers have a reputation as being "sinking" pitches, but they really only "sink" in the sense that they fall slightly faster than four-seam fastballs and move slightly more. So, the other cluster, with a bit more vertical movement and a bit less horizontal movement is the four-seamer.

Someone could post the video from the Saber Seminar where I briefly go over pitch IDs and apparently look like a serial killer.

[czar]

From the catcher's point of view, a two-seamer (from a lefty pitcher) will move up and to the right. So, one of the clusters of dots in the top right of the graph is the two-seamer. Two seamers have a reputation as being "sinking" pitches, but they really only "sink" in the sense that they fall slightly faster than four-seam fastballs and move slightly more. So, the other cluster, with a bit more vertical movement and a bit less horizontal movement is the four-seamer.

One minor clarification (I think)-- harkening back to the age-old "do rising fastballs rise?" disco, the ball only looks like it's going up relatively to a ball thrown with no spin (it's downward sink (gravity) is partially offset by an upward pointing Magnus). In fact (and Jnai can correct me if I'm actually wrong), all pitch movement numbers are relative to pure projectile motion (the point 0,0) with no spin. Ex: a curve from a lefty moves down in the z-plane and negative in the x-plane (catcher's left) relative to a smooth ball thrown at the same velocity with zero spin (no Magnus).

Jnai didn't say anything incorrect, and this is probably way more nitpicky than 90% of the people who see those plots care about, just wanted to make the distinction clear that Pf/x isn't calibrated to like a catcher/hitter's eye plane or something (and that obviously, nothing is actually rising).

Edited by czar, 05 August 2011 - 02:57 PM.

[Jnai]

One minor clarification (I think)-- harkening back to the age-old "do rising fastballs rise?" disco, the ball only looks like it's going up relatively to a ball thrown with no spin (it's downward sink (gravity) is partially offset by an upward pointing Magnus). In fact (and Jnai can correct me if I'm actually wrong), all pitch movement numbers are relative to pure projectile motion (the point 0,0) with no spin. Ex: a curve from a lefty moves down in the z-plane and negative in the x-plane (catcher's left) relative to a smooth ball thrown at the same velocity with zero spin (no Magnus).

Jnai didn't say anything incorrect, and this is probably way more nitpicky than 90% of the people who see those plots care about, just wanted to make the distinction clear that Pf/x isn't calibrated to like a catcher/hitter's eye plane or something (and that obviously, nothing is actually rising).

Czar is correct.

To wit, "The movement numbers represent amount of movement due to the spin of the baseball."

[Al Zarilla]

One minor clarification (I think)-- harkening back to the age-old "do rising fastballs rise?" disco, the ball only looks like it's going up relatively to a ball thrown with no spin (it's downward sink (gravity) is partially offset by an upward pointing Magnus). In fact (and Jnai can correct me if I'm actually wrong), all pitch movement numbers are relative to pure projectile motion (the point 0,0) with no spin. Ex: a curve from a lefty moves down in the z-plane and negative in the x-plane (catcher's left) relative to a smooth ball thrown at the same velocity with zero spin (no Magnus).

Jnai didn't say anything incorrect, and this is probably way more nitpicky than 90% of the people who see those plots care about, just wanted to make the distinction clear that Pf/x isn't calibrated to like a catcher/hitter's eye plane or something (and that obviously, nothing is actually rising).

Question, what is that pure projectile motion then, like a baseball flung with a pitching machine at 60 mph or something? Also, I don't know the history of PitchFx, but why do they do it this way and show "positive" vertical movement numbers? I think I'd be a lot more "comfortable" with seeing the 2 and 4 seamers as negative vertical movement, but not as negative of course as curveballs and other "breaking" balls. Is it because they can show better resolution? Getting way off track here, I realize.

[czar]

Question, what is that pure projectile motion then, like a baseball flung with a pitching machine at 60 mph or something?

A smooth ball thrown at the same mph without any spin. It only takes into account gravity as a force-- a real baseball also has motion due to it's spin (aka the Magnus force-- has to do with boundary layer separation over the ball surface, yadda, yadda, physics).

Also, I don't know the history of PitchFx, but why do they do it this way and show "positive" vertical movement numbers? I think I'd be a lot more "comfortable" with seeing the 2 and 4 seamers as negative vertical movement, but not as negative of course as curveballs and other "breaking" balls. Is it because they can show better resolution? Getting way off track here, I realize.

It would actually be relatively trivial to plot it this way-- the cameras actually measure the raw (aka unadjusted) x,y,z locations of the ball, which I believe are included in the XML data (I think?).

All the spin numbers (RPM/angle) are actually "back-calculated" out from the "photographed" trajectory using a physics model after the "known" effect of gravity is filtered out. This is why it's easy to dump in +/- relative to pure projectile motion, because you're looking at the differences between a real pitch and basic physics theory. The specific details of these calculations (and how they end up in the data that you see at BrooksBaseball) are ones that someone like Jnai or alannathan can likely describe much better than myself.

Edited by czar, 05 August 2011 - 03:25 PM.

[Sprowl]

It would actually be relatively trivial to plot it this way-- the cameras actually measure the raw (aka unadjusted) x,y,z locations of the ball, which I believe are included in the XML data (I think?).

TexasLeaguers.com has an algorithm set up to give the actual motion by adjusting the pitchfx numbers for the effects of gravity. Here is Bedard's performance using those criteria:



Note that pitchfx is misclassifying several of Bedard's 4-seamers as sinkers.

[phragle]

TexasLeaguers.com has an algorithm set up to give the actual motion by adjusting the pitchfx numbers for the effects of gravity. Here is Bedard's performance using those criteria:



Note that pitchfx is misclassifying several of Bedard's 4-seamers as sinkers.

That's brilliant.