Johan Santana

[redsoxrules]

. Only players with a minimum of 100 pitches while PITCHf/x was on are listed. Be a little careful looking at the numbers below as some of them have very small statistics. Also, to try to give you a true representation of what a pitcher throws I am showing each pitch like it was thrown at sea level at standard temperature (59 degrees). If I didn't do this pitches thrown in very warm weather or at Coors would break up to 25% less and you would end up with even more confusing blobs to look at

 

 

thats the definition of your stats.....those stats are off and just admit it

[example1]

Why, all of a sudden, has the burden of proof shifted to me once the validity of the data you have used comes into question? It's disappointing you try to convince people of things and do not back them up with relevant data in this instance.

 

You say that 431 pitches, or 39.7% of the sample, does not skew the data at all. I guess there's nothing I can say to that which will convince you otherwise that the data is not reliable.

 

I'm sorry, I cannot classify anyone who sits at 91 a hard thrower, lefty or righty.

 

Secondly, again, show me ANY piece of data that can show me 5 LHP who consistently threw harder. If you can't, then I stand correct to assert that Jon Lester is a hard throwing lefty (at least relative to other left handed pitchers in professional baseball, which is, as far as I know, the data pool we're talking about here).

 

It's funny how you talk about "burden of proof" but don't actually prove anything, other than to state the obvious: that the sample is an incomplete (albeit large) one.

[example1]

. Only players with a minimum of 100 pitches while PITCHf/x was on are listed. Be a little careful looking at the numbers below as some of them have very small statistics. Also' date=' to try to give you a true representation of what a pitcher throws I am showing each pitch like it was thrown at sea level at standard temperature (59 degrees). If I didn't do this pitches thrown in very warm weather or at Coors would break up to 25% less and you would end up with even more confusing blobs to look at[/b']

 

 

thats the definition of your stats.....those stats are off and just admit it

 

That's consistent with every pitcher on my list, right?

 

You are wanting to claim that Lester is NOT a hard throwing lefty compared to other lefties. I have provided SOME proof, while stating very politely that I'm open to other data if I'm wrong. Nobody has provided any other data, so I won't "just admit it". I'll admit that they are ONE SET OF STATS.

[TheKilo]

Secondly, again, show me ANY piece of data that can show me 5 LHP who consistently threw harder. If you can't, then I stand correct to assert that Jon Lester is a hard throwing lefty (at least relative to other left handed pitchers in professional baseball, which is, as far as I know, the data pool we're talking about here).

 

It's funny how you talk about "burden of proof" but don't actually prove anything, other than to state the obvious: that the sample is an incomplete (albeit large) one.

 

I'll make this concession. He throws among the hardest of left-handers, however he's far from being considered a hard throwing lefty.

 

Happy?

[TheKilo]

That's consistent with every pitcher on my list, right?

 

You are wanting to claim that Lester is NOT a hard throwing lefty compared to other lefties. I have provided SOME proof, while stating very politely that I'm open to other data if I'm wrong. Nobody has provided any other data, so I won't "just admit it". I'll admit that they are ONE SET OF STATS.

 

Nice qualifier, "compared to other lefties".

 

If people threw underhand, softball style, but only maxed out at 60 MPH, would someone who threw 65 be a hard thrower?

[example1]

Did you watch the damn games last season?

 

This argument is foolish. It's about semantics. I don't consider 91, 92 to be a hard thrower just as much as you don't believe leaving 40% of the sample size out of data makes it any less relevant.

 

 

 

 

"Put it up" to 93, 94. Doesn't seem like that means he's "sitting" there.

 

I said before that he elevates it to 95.

 

Here's a completely different source: http://webusers.npl.uiuc.edu/~a-nathan/pob/Analysis.pdf

 

It's from ONE game (being honest about sample size). I see a number of pitches COMFORTABLY between 90-95. That, combined with the average velocity of the 600 pitches measured above, tells me he sits probably right between, which is somewhere close to 93.

 

Yeah, its semantics TheKilo. You know I respect you as a poster, but I can't help but wonder why you have to attack my work with one-liners when I go to such lengths to find actual data. Rather than just saying "did you watch the games last year?" you could say "thanks exmaple1 for finding all that data, here are the problems I have with it, and here is the site that I found in my own time to back up my claims". Instead I just get one liners while you sit around waiting for my next post.

 

I DID watch the games last year (as you well know, we watched a number of them 'together' on this very board). What I saw from lester is more consistent with the data I'm finding than what you guys are saying. I saw a lefty who threw HARD FOR A LEFTY, whose fastball I would guess was between 92 and 94 consistently and which occasionally went higher. It is surprising that he threw harder than Bedard or Santana--and I wouldn't be shocked if that was wrong--but for you guys to just let the "he's basically Ted Lilly" comment go without stepping is weak, because I KNOW you watched the games and I KNOW you know Lester threw harder than Lilly does.

[example1]

Nice qualifier, "compared to other lefties".

 

If people threw underhand, softball style, but only maxed out at 60 MPH, would someone who threw 65 be a hard thrower?

 

I said "hard throwing lefty". If I said "fast red car" or "giant moon of Jupiter" would you consider blue cars and moons of mars in your data set? Yes, apparently you would.

[example1]

I'm not qualifying anything. The reason that the term "hard throwing lefty" exists is because lefties tend not to throw hard.

 

I bet you don't believe there is a SINGLE hard throwing lefty in baseball, which makes the term--commonly used and, in my opinion, rightly so--actually have no meaning.

[example1]

Nice qualifier, "compared to other lefties".

 

If people threw underhand, softball style, but only maxed out at 60 MPH, would someone who threw 65 be a hard thrower?

 

And yes, if someone threw underhand in softball, and someone threw harder than most of his/her peers, I would say "hard throwing softball pitcher". The "hard throwing" modifies that pitcher, and is not some objective statement. In this case, however, I'd be very, VERY impressed if you were able to hit Lester's fastball. As you well-know, anyone who throws 90mph throws VERY hard, unless you are comparing them to other professional baseball players.

 

Of course, if you were comparing them to other baseball players, then you would also take "hard throwing lefty" and compare that to "baseball players who are left handed".

[example1]

I'll make this concession. He throws among the hardest of left-handers, however he's far from being considered a hard throwing lefty.

 

Happy?

 

God love you and your Family Guy avitars. I can't get mad at stewie! (it's a good thing you don't have a picture of that damn yellow-chicken who consistently attacks Peter!)

[redsoxrules]

I said before that he elevates it to 95.

 

Here's a completely different source: http://webusers.npl.uiuc.edu/~a-nathan/pob/Analysis.pdf

 

It's from ONE game (being honest about sample size). I see a number of pitches COMFORTABLY between 90-95. That, combined with the average velocity of the 600 pitches measured above, tells me he sits probably right between, which is somewhere close to 93.

 

Yeah, its semantics TheKilo. You know I respect you as a poster, but I can't help but wonder why you have to attack my work with one-liners when I go to such lengths to find actual data. Rather than just saying "did you watch the games last year?" you could say "thanks exmaple1 for finding all that data, here are the problems I have with it, and here is the site that I found in my own time to back up my claims". Instead I just get one liners while you sit around waiting for my next post.

 

I DID watch the games last year (as you well know, we watched a number of them 'together' on this very board). What I saw from lester is more consistent with the data I'm finding than what you guys are saying. I saw a lefty who threw HARD FOR A LEFTY, whose fastball I would guess was between 92 and 94 consistently and which occasionally went higher. It is surprising that he threw harder than Bedard or Santana--and I wouldn't be shocked if that was wrong--but for you guys to just let the "he's basically Ted Lilly" comment go without stepping is weak, because I KNOW you watched the games and I KNOW you know Lester threw harder than Lilly does.

 

 

BTW that ONE game was the hardest that he threw all year

[Jayhawk Bill]

Why would anybody dispute the validity of PITCHf/x velocity stats?

 

Josh Kalk is just using some utilities to consolidate the cumbersome PITCHf/x files. That's what example1 is using for a reference. Jon Lester's fastball's mean initial speed is 93.14mph. That's faster than the fastball of four out of five of the AL leaders in innings pitched; it's pretty fast for an average speed.

 

***

 

I'll make this concession. He throws among the hardest of left-handers, however he's far from being considered a hard throwing lefty.

 

Probably true. "Hard-throwing lefty" is a subjective term based upon the arbitrary beliefs of those who view MLB recreationally, and Lester isn't considered a hard-throwing lefty by most.

 

As example1 points out, Lester does throw hard: his fastball is above MLB average for all pitchers, and many lefties make their living with odd deliveries and slow fastballs, so he might be faster compared to lefties.

 

But that's merely objective truth. How Lester is considered can, and I perceive does, differ. :dunno:

 

Which may be why he's one of four players rumored to be in the Santana offer.

[TheKilo]

E1, you know I have no beef with you. One liners are how I roll.

[TheKilo]

Why would anybody dispute the validity of PITCHf/x velocity stats?

 

 

You misunderstand. I'm sure the data itself is relevant, however eliminating 40% of the sample makes me call the analysis into question.

[Jayhawk Bill]

You misunderstand. I'm sure the data itself is relevant' date=' however eliminating 40% of the sample makes me call the analysis into question.[/quote']

 

Do you similarly question every Gallup or Zogby poll? They query only about 0.001% of populations before stating results regarded as fact by the majority of Americans. Here we're charting roughly half of Lester's pitches, the sample determined quasi-randomly by when he happened to pitch at PITCHf/x ballparks. Why would you consider the sample unrepresentative?

 

Let's look at the entire sample set distribution:

 

http://baseball.bornbybits.com/plots/gifs/Jon_Lester3.gif

 

Do you still question that Lester has roughly a 93mph fastball?

[ksushi]

I'll make this concession. He throws among the hardest of left-handers, however he's far from being considered a hard throwing lefty.

 

Happy?

 

I believe Lesters average fastball is around 93. You might not describe him as a flame thrower, but I consider that throwing hard. You definitely can't seriously say he is "far from being considered a hard throwing lefty".

 

What do you consider throwing hard? And, would you consider Lester a soft-tossing lefting as opposed to a hard throwing lefty?

[example1]

E1' date=' you know I have no beef with you. One liners are how I roll.[/quote']

 

Absolutely.

 

Do you similarly question every Gallup or Zogby poll? They query only about 0.001% of populations before stating results regarded as fact by the majority of Americans. Here we're charting roughly half of Lester's pitches, the sample determined quasi-randomly by when he happened to pitch at PITCHf/x ballparks. Why would you consider the sample unrepresentative?

 

There is actually a point within a sample, as I'm sure you know, where it becomes almost futile to keep measuring. If you're using a random sample (which this likely isn't, but it certainly isn't intentionally biased either, so it's close) then you don't need to measure 50% of the population to know what general traits that population shows. If that population is something as objective as a mph number, the confidence can be pretty high that you've nailed it if you have 60%.

[Jayhawk Bill]

There is actually a point within a sample' date=' as I'm sure you know, where it becomes almost futile to keep measuring.[/quote']

 

You're right. I know.

 

The thing is, TheKilo is a very smart poster, and I'm pretty sure he knows, too. RSR seems smart enough, and RSR seems ready to throw out the whole PITCHf/x data set. I just don't get it. :dunno:

 

I frequently see opinions like these attributed to baseball insiders, but they're usually insiders with an agenda, either umpires or scouts. Obviously, scouts are becoming nervous that stats may take away their jobs: if we can discern how well a batter hits varying types of pitches in varying locations just by checking stats, we're getting to where scouts' skill sets are becoming obsolete. Umpires get testy about the whole issue of automated evaluation of balls and strikes because it reveals at least two things:

 

1) They're wrong a significant part of the time, and

2) The MLB strike zone isn't the one defined in the rule book.

 

It may also show two other things:

 

3) Umpires have individual strike zones, and

4) Umpires are biased against or for certain players and teams to a statistically significant extent.

 

Jonathan Hale has done some good work on the third issue I cited:

 

http://www.hardballtimes.com/main/article/a-zone-of-their-own/

 

but the fourth is a verboten topic. I picked up two games last year where the umpires were calling different strike zones for different pitchers, one where Matsuzaka and Okajima were being denied called strikes and one where Ian Kennedy was being given an absurdly expanded strike zone. One can determine the likelihood of such patterns of errors using binomial theorem (umpire's mistakes should favor both teams equally), and both of these games were biased at something over the 99% confidence level. The Red Sox game was biased, in fact, at the 99.99% confidence level. I contacted a few authors with whom I occasionally correspond, and literally nobody would even acknowledge the emails. (Before you say anything, they acknowledged previous and subsequent correspondence on other topics.) If one's living depends upon the integrity of the game being unchallenged, and if one needs professional contacts within a game, there are topics that cannot be touched. Those with integrity themselves simply avoid those topics.

 

Those with less integrity--sadly, often umpires--question the reliability of the system, as if triangulation from three precise digital images capturing pitches at widely divergent angles could somehow be less precise than a human relying on two imperfect eyes separated by only a few inches. Such criticisms do leak out. A few, rightly, can be attributed to either calibration or the view from one camera being obstructed, but the system is quickly turned off in such cases. Most criticisms of PITCHf/x and its less-capable predecessor, QuesTec, are unfounded, but they get released and they have their believers.

 

PITCHf/x data is excellent-quality data. It's literally better than scouting reports for such things as pitch velocity and location. Those who challenge it usually have agendas. The posters here have pretty good knowledge of the game, and some are very well-informed. I'm surprised that PITCHf/x data would be challenged.

 

...challenged here in the Johan Santana thread as an aspect of evaluating the talent of one Boston pitcher who might possibly be part of a Santana package.

[redsoxrules]

You're right. I know.

 

The thing is, TheKilo is a very smart poster, and I'm pretty sure he knows, too. RSR seems smart enough, and RSR seems ready to throw out the whole PITCHf/x data set. I just don't get it. :dunno:

 

 

I just dont think its a 100% acurate, the only live game I saw this year was when lester was pitching and he was no wheres near 95 , and he was actually pitching against kazmir ( who is at 92.9 on that data ) . and kazmir was a good 4,5 MPH harder .. unless i dont remember correctly ( it was the august 14th game ) :dunno:

[Jayhawk Bill]

I just dont think its a 100% acurate' date=' the only live game I saw this year was when lester was pitching and he was no wheres near 95 , and he was actually pitching against kazmir ( who is at 92.9 on that data ) . and kazmir was a good 4,5 MPH harder .. unless i dont remember correctly ( it was the august 14th game ) :dunno:[/quote']

 

Checking the MLB.com raw data files for that game, Kazmir was 2-3mph faster than Lester on the top end--that wasn't the case over the whole season. That game you saw was the only quality start for Lester in a five-game stretch where he posted an ERA over 6.00, and he was demoted to AAA after one more game amidst rumors of fatigue, IIRC. He posted a 3.34 ERA after he returned in September--he might've gotten those couple of mph back on his fastball.

[jacksonianmarch]

Do you similarly question every Gallup or Zogby poll? They query only about 0.001% of populations before stating results regarded as fact by the majority of Americans. Here we're charting roughly half of Lester's pitches, the sample determined quasi-randomly by when he happened to pitch at PITCHf/x ballparks. Why would you consider the sample unrepresentative?

 

Let's look at the entire sample set distribution:

 

http://baseball.bornbybits.com/plots/gifs/Jon_Lester3.gif

 

Do you still question that Lester has roughly a 93mph fastball?

 

that is extremely easy to call into question, though. Some guns and operators are different. I am going off the NESN reads. As we all know some guns are intentionally juiced (like on Fox for example) and others seem to lag. I think the NESN guns were rather accurate as most of them had Schilling in the high 80s-low 90s after his 1 hitter, which is where he said he himself felt like he was at. While FOX broadcast games still had him pouring it in at 93-95.

 

Regardless, I know what I saw as I was able to watch almost every one of Lester's starts. The NESN guns had him sitting around 90-91 with the occasional 93-94. Not once did I see a 95 from him last yr. And even the typically juiced guns in the WS did not have Lester throwing that hard.

[jacksonianmarch]

and going by that scale, he apparently threw a few 92mph changeups? Hilarious.

[MANNYHOF24]

that is extremely easy to call into question, though. Some guns and operators are different. I am going off the NESN reads. As we all know some guns are intentionally juiced (like on Fox for example) and others seem to lag. I think the NESN guns were rather accurate as most of them had Schilling in the high 80s-low 90s after his 1 hitter, which is where he said he himself felt like he was at. While FOX broadcast games still had him pouring it in at 93-95.

 

Regardless, I know what I saw as I was able to watch almost every one of Lester's starts. The NESN guns had him sitting around 90-91 with the occasional 93-94. Not once did I see a 95 from him last yr. And even the typically juiced guns in the WS did not have Lester throwing that hard.

 

I actually remember him touching 96 a few times on the ws gun. In that game he was around 94 all night.

[Jayhawk Bill]

that is extremely easy to call into question' date=' though. Some guns and operators are different. I am going off the NESN reads. As we all know some guns are intentionally juiced (like on Fox for example) and others seem to lag.[/quote']

 

Did you read Josh Kalk's reference to the combinatorial algorithms that accounted for calibration and differences in the atmospheres at different ballparks?

 

http://www.baseball.bornbybits.com/blog/2007/11/explanation-of-correction-code.html

 

The big corrections are for atmospheric differences, but he norms all factors, if I understand correctly.

 

I think the NESN guns were rather accurate as most of them had Schilling in the high 80s-low 90s after his 1 hitter, which is where he said he himself felt like he was at. While FOX broadcast games still had him pouring it in at 93-95.

 

Josh Kalk's PITCHf/x data has Schilling at 89.8mph for a mean fastball velocity.

 

I'm intrigued, though, that you trust Schilling's perception over any external metric.

 

Regardless, I know what I saw as I was able to watch almost every one of Lester's starts. The NESN guns had him sitting around 90-91 with the occasional 93-94. Not once did I see a 95 from him last yr. And even the typically juiced guns in the WS did not have Lester throwing that hard.

 

Lester was hitting 94.1 in the World Series per PITCHf/x raw data.

 

and going by that scale, he apparently threw a few 92mph changeups? Hilarious.

 

Kalk's classification algorithm groups pitches by horizontal and vertical movement. Those "92mph changeups" that look mixed into the fastballs by velocity and vertical break had a different horizontal break than the fastballs.

 

Hilarious? See, that's a subjective perception. One could criticize the clustering algorithm for clustering pitches over a whole season, not start-by-start--Eric Van has already raised that point--but Lester threw a few 90+mph pitches that moved as if they weren't fastballs. The chart reflects that. But I don't perceive hilarity in diligently tracking Lester's pitches by velocity and acceleration due to spin...YMMV. :dunno:

[example1]

that is extremely easy to call into question' date=' though. Some guns and operators are different. I am going off the NESN reads. As we all know some guns are intentionally juiced (like on Fox for example) and others seem to lag. I think the NESN guns were rather accurate as most of them had Schilling in the high 80s-low 90s after his 1 hitter, which is where he said he himself felt like he was at. While FOX broadcast games still had him pouring it in at 93-95. [/quote']

 

The point is that, as long as you're using the same measuring devices for ALL players, the variance in different guns shouldn't matter very much. It's not like I was stupidly comparing Lester with Pitch f/x but Kazmir using NESN's guns and Bedard was measured with a stopwatch.

 

Actually, measuring velocity is simply not that hard. There would be failure all over the place if, technologically, we were incapable of figuring out precise velocity. However, the military, NASA, etc., have long since moved beyond the struggle to get accurate measurements on something the size of a baseball hurtling through space--not millions of miles away, or even light years away--but right there, 50 yards away.

 

Here's a slate.com article about the system and its implications.

 

http://www.slate.com/id/2172223

 

 

Regardless, I know what I saw as I was able to watch almost every one of Lester's starts. The NESN guns had him sitting around 90-91 with the occasional 93-94. Not once did I see a 95 from him last yr. And even the typically juiced guns in the WS did not have Lester throwing that hard.

 

If the data is skewing Lester's fastball as too fast, how come it doesn't do the same for the rest of the Red Sox pitchers. Which of these seem wrong (remember, they are the speed at release point, not as it crosses the plate)?

 

[table] |Papelbon | 96.01 |

|Beckett | 95.74 |

|Delcarman | 95.38 |

|Matsuzaka | 93.95 |

|Gagne | 93.24 |

|Lester | 93.14 |

|Snyder | 92.73 |

|Timlin | 92.54 |

|Tavarez | 91.1 |

|Schilling | 89.81 |

|Okajima | 88.94 |

|Lopez | 87.09 |

|Wakefield | 75.62 |

 

[/table]

 

How about some Yankees? I know Jacksonianmarch would be familiar with their performance, likely having seen a few with his own eyes on tv or in person:

 

Yankees:

 

[table] | Chamberlain | 98.22 |

|Farnsworth | 96.37 |

| Bruney | 94.7 |

|Wang | 94.08 |

|Rivera | 94.03 |

|Vizcaino | 92.74 |

|Hughes | 92.26 |

|Clemens | 91.66 |

|Pettitte | 91.25 |

|Kennedy | 90.88 |

|Igawa | 89.67 |

|Mussina | 85.52 |

|Villone | 85.4 |

 

[/table]

 

There is a whole lot one could do with data like this. I would be very interested to know whether this year represents a low point in terms of LHP velocity, or whether it is a normal year in that respect. I would also be interested to see if there has been any effect of PED enforcement.

[ksushi]

Hilarious? See, that's a subjective perception. One could criticize the clustering algorithm for clustering pitches over a whole season, not start-by-start--Eric Van has already raised that point--but Lester threw a few 90+mph pitches that moved as if they weren't fastballs. The chart reflects that. But I don't perceive hilarity in diligently tracking Lester's pitches by velocity and acceleration due to spin...YMMV. :dunno:

 

Easily explained. Cut fastball. The chart doesn't have the cutter accounted for and its a 88-90 mph pitch that doesn't behave like a four seamer. Makes sense.

[Jayhawk Bill]

Concur, ksushi. Good call.

[das11209]

No Santana news in a week? That's a shame.

Hank is back after the holidays

Senior vice president Hank Steinbrenner said Wednesday that he believes the Yankees have made the best offer for Johan Santana, adding that the Twins "are not going to trade him before checking with us one last time."

 

The New York Daily News speculates that the Yankees' offer includes Phil Hughes, Melky Cabrera, Jeff Marquez and "another prospect" that isn't Ian Kennedy. "We're not desperate, so we're not going to chase anything," Steinbrenner said. "In the next two weeks, we're going to have to get everything done."

[rician blast]

Said Hank: "We're not desperate, so we're not going to chase anything," Steinbrenner said. "In the next two weeks, we're going to have to get everything done"

 

Uhhhh, yah. I think we all know the timetables and such that pour from Hank's mouth don't mean much.

[das11209]

Said Hank: "We're not desperate' date=' so we're not going to chase anything," Steinbrenner said. "[i']In the next two weeks, we're going to have to get everything done[/i]"

 

Uhhhh, yah. I think we all know the timetables and such that pour from Hank's mouth don't mean much.

 

He is getting it done in two weeks for the past 2 months already.