Mike Leake, Best of the Bunch?

[Falls City Beer]

I don't think anyone other than maybe the op is saying that Leake is the best of the bunch. I think most folks, however, are arguing that Leake is not a bottom of the rotation pitcher. And that context (defense, post-roid era) matters. Would Leake be less successful with crappy defense behind him? I'm sure. Good thing he's got great defense behind him. Leake will undoubtedly pitch up to if not over his current contract. Will he be a good candidate to sign long term? I don't know. It's not an issue right now.

[jojo]

And for the umpteenth time, that ignores the fact that the Reds staff changed in personal during those years, and those new pitcher were the reason why the production was better. They added Masset and Rhodes, who made up most of the improvement.

Two relief pitchers swung the BABIP of the entire staff lower by .049 pts? I understand it was tempting to pick out two guys with low BABIP to argue against the narrative that the Reds defensive quality was driving the staff's BABIP but that really wasnt the case as the table below demonstrates:

	Staff  Starters  Pen  UZR(rank)
2004	0.296	0.298	0.282	19
2005	0.309	0.307	0.313	27
2006	0.301	0.297	0.308	20
2007	0.31	0.308	0.315	24
2008	0.312	0.31	0.315	23
2009	0.283	0.285	0.278	2
2010	0.288	0.287	0.291	3
2011	0.282	0.286	0.273	3
2012	0.288	0.29	0.284	3
2013	0.273	0.278	0.263	7

The pen really didn't drive the decrease. As the Reds defense dramatically improved so too did the composite BABIP for the rotation and pen. In fact for the transition between the 2008 and 2009 seasons, BABIP actually decreased greater for the rotation than it did for the pen. So no, it wasn't two relievers driving most of the improvement in BABIP. BTW, for those that love statistics, the correlation between UZR rank and Staff BABIP was .89 (2013 actually tanked it down from .95).

None of the pitchers who pitched for the Reds in both 2008 and 2009 improved, except for Arroyo. Here are the numbers of the players who pitched for the Reds in both 2008 and 2009... for the umpteenth time.

2008 ERA - BABIP

Harang - .301
Cueto - .298
Cordero - .302
Burton - .301
Arroyo - .314

2009 BABIP

Harang - .331
Cueto - .291
Cordero - .301
Burton - .298
Arroyo - .265

You're not pointing to what you think you're pointing to...What you're demonstrating is that guys who give up more contact will have their BABIP affected more by the quality of their defense...aka Bronson Arroyo or Mike Leake which is one reason why Arroyo's BABIP and hence ERA was affected more by the dramatic defensive upgrade than the Reds staffs as a whole were affected.

Also, as to the notion that Arroyo has demonstrated an ability to post low BABIPs throughout his career, that's not actually accurate. As the table below shows, while in keeping with BABIP, Arroyo has had some large swings in BABIP throughout his career, the only consistent trend one sees in his BABIP is that his BABIP has been consistently lower post 2008 than it has been pre 2008:

Period		BABIP
Pirate		0.311
Redsox		0.279
06-'08 Red	0.296
09-'13 Red	0.267
		
00-'08		0.293
09-'13		0.267

[SteelSD]

Also, as to the notion that Arroyo has demonstrated an ability to post low BABIPs throughout his career, that's not actually accurate.

I'm not even sure why we're into discussing Arroyo at this point as he really doesn't have much of anything to do with Leake, excepting that he appears to now be the poster child for pitch-to-contact BABIP overachievers. But you're right.

	Arroyo	Team	Team  	BABIP vs
YEAR	BABIP	DER	BABIP	Team
2013	0.267	0.727	0.273	-0.006
2012	0.286	0.712	0.288	-0.002
2011	0.278	0.718	0.282	-0.004
2010	0.239	0.712	0.288	-0.049
2009	0.265	0.717	0.283	-0.018
2008	0.314	0.688	0.312	0.002
2007	0.309	0.690	0.310	-0.001
2006	0.270	0.699	0.301	-0.031

That shows Arroyo's BABIP alongside the Reds' team Defensive Efficiency Rating and team BABIP followed by Arroyo's variance versus the team. The correlation between Arroyo's seasonal BABIP and team DER with the Reds is 68.8%. That's a pretty high correlation if he's supposed to repeatedly break the system and the variances are consistently very very low for a guy who's alleged to limit the importance of defensive contribution.

But what's fascinating to me is that Arroyo only demonstrates two seasons out of eight that we might consider actual outliers- 2006 (-0.031 BABIP variance) and 2010 (-0.049 BABIP variance). I question the validity of the concept that something is repeatable skill when it happens only twice with the occurrences falling four years apart. Why, for example, would Arroyo demonstrate a skill and then just choose to mothball it for three seasons only to bring it out again for a single season and then retire it for good? Doesn't smell like a repeatable skill. But let's assume those two seasons were due to solely to Arroyo and just remove those from the sample to see how closely his performance correlated with team the rest of the time. The correlation is 96.22% for his seasonal BABIP to the Reds' team DER over the remaining six seasons. That's sort of a freakish correlation considering that we should expect to see limited correlation if Arroyo is just choosing to limit defensive impact on ball-in-play events. The real answer is that he's can't consistently choose to do such a thing over time and has demonstrated that.

Regardless, what does any of that have to do with Mike Leake? Nothing, of course. Leake hasn't produced and certainly hasn't reproduced any freakish BABIP variances versus his team or shown a propensity to outperform a low-DER defensive squad. He's not a top-flight ground-ball maven. His Infield Fly rate isn't out of whack. His HR rate isn't low enough to rate any real interest. He doesn't constantly and repeatedly outperform his advanced metrics. He's got exactly one season three years ago where he showed a small-ish out-performance of his team's BABIP and that's pretty much it.

To put it simply, Leake's numbers really aren't even as interesting as Bronson Arroyo's; whose aren't really that much more interesting.

[JayStubbs]

In order for any of the above correlation numbers between Arroyo's BABIP and the team's BABIP, to have any relevance, the team BABIP numbers must not include Arroyo's. There is a reason why his numbers correlate so well with the team's, he's providing around 15% of the team's numbers.

It's like saying that Jose Bautista's HR numbers correlates with the Blue Jay's HR numbers. Well... duh?

[jojo]

In order for any of the above correlation numbers between Arroyo's BABIP and the team's BABIP, to have any relevance, the team BABIP numbers must not include Arroyo's. There is a reason why his numbers correlate so well with the team's, he's providing around 15% of the team's numbers.

It's like saying that Jose Bautista's HR numbers correlates with the Blue Jay's HR numbers. Well... duh?

That really doesn't explain why Arroyo's BABIP is nearly identical to the team BABIP for 6 out of 8 seasons.

[SteelSD]

In order for any of the above correlation numbers between Arroyo's BABIP and the team's BABIP, to have any relevance, the team BABIP numbers must not include Arroyo's. There is a reason why his numbers correlate so well with the team's, he's providing around 15% of the team's numbers.

It's like saying that Jose Bautista's HR numbers correlates with the Blue Jay's HR numbers. Well... duh?

Yeah. That's not going to be happening. Removing ball in play outcomes from the team isn't anything at all like your Jose Batista example. You're actually asking me to skew the team's DER by removing team performance from team performance. That's not appropriate.

[JayStubbs]

Yeah. That's not going to be happening. Removing ball in play outcomes from the team isn't anything at all like your Jose Batista example. You're actually asking me to skew the team's DER by removing team performance from team performance. That's not appropriate.

Arroyo's IP consist of around 15% of his team's IP every year. Just like Bautista's HR totals consist of around 15% of his team's HR totals every year. It's exactly the same thing.

Bautista's HR total's are always going to correlate with his team's HR totals, because he makes up such significant percentage of them.
Arroyo's BABIP are always going to correlate with this team's BABIP because he makes up a such a significant percentage of them.

I don't have a dog in this fight over Arroyo and his BABIP. It could be the case that when you remove his stats from the team stats, you get similar results. All I'm saying is that it is bad math to include him in the team stats.

To put another way, if you want to compare how Arroyo does compared to the rest of his teammates in any stat, you can't include him as one of his teammates.

When comparing how the state of Ohio does in beer sales compared to the rest of the country, you don't compare Ohio beer sales to the nation's beer sales. You compare Ohio beer sales to the combined beer sales of every other state. Otherwise, you are dealing with skewed numbers.

[SteelSD]

To put another way, if you want to compare how Arroyo does compared to the rest of his teammates in any stat, you can't include him as one of his teammates.

Then I'd suggest that your next entry show the results of your study, set up however you'd like.

[jojo]

Here is the BABIP data for Arroyo versus the Reds staff minus Arroyo during the period spanning 2006-2013:

	  BA	Staff	  mlb
2006	0.270	0.302	0.298
2007	0.309	0.306	0.299
2008	0.314	0.306	0.296
2009	0.265	0.282	0.295
2010	0.239	0.294	0.293
2011	0.278	0.278	0.291
2012	0.286	0.281	0.293
2013	0.267	0.270	0.294
ave	0.279	0.290	0.295
SD	0.023	0.013	0.003
			
	  BA	Staff	  mlb
06-'08	0.298	0.305	0.298
09-'13	0.267	0.281	0.293

In his 8 years as a Red, he had two seasons where his BABIP was dramatically different than the rest of the guys pitching in his environment. He had 5 years where his BABIP was essentially identical and 1 year where it was 6% lower than his teammates. As can be seen by averages constructed by using the '08 offseason as a demarcation, Arroyo really didnt maintain a BABIP that was lower than exected until the Reds defense was dramatically altered and even then his BABIP dropped .021 pts while his teammates' BABIP dropped .024 from the pre-improved defense averages. So five years of identical performance and erring on his side, three years of lower than expected BABIP. Those arguing Arroyo has an ability to suppress BABIP have to concede that he has a track record inconsistent with their position because more often than not, he has failed to do so (and interestingly the narrative largely rests on his '09-'10 seasons because his BABIP is almost identical to his teamates for the previous two seasons and the three seasons following....this is a classic regression to the mean profile).

It's more compelling to conclude that like most pitchers, Arroyo has experienced a couple of seasons where his BABIP experienced an extreme fluctuation but if you look at him in totality, he really hasn't shown an ability to consistently suppress his BABIP. What looks like such an ability when simply looking at a career average really is mostly just the influence of the improved Reds defense, an influence that overall basically effected Arroyo and his teammates similarly. When looking at the results of defense-independent contact (HR/FB%) Arroyo has not demonstrated any ability to reduce quality of contact as his HR/FB rates as a Red has been consitently higher than expected.

Basically concerning whether Arroyo can control his BABIP or not, the best one can argue is that his penchant for giving up higher than normal contact rates magnifies the influence of the defense behind him. Arguing anything beyond that places one on thin ice concerning the data.

[wheels]

Let's hope the Leake we saw last year shows up for one more season (whoever brought up Kent Bottenfield earlier in the thread is brilliant), and let's hope Walt can find version 2.0 of Jim Edmonds in the trade market.

I dig Mike Leake, but there is absolutely no way I'd want the Reds to keep him over Latos. Flip him for a bat.

[M2]

Let's hope the Leake we saw last year shows up for one more season (whoever brought up Kent Bottenfield earlier in the thread is brilliant), and let's hope Walt can find version 2.0 of Jim Edmonds in the trade market.

I dig Mike Leake, but there is absolutely no way I'd want the Reds to keep him over Latos. Flip him for a bat.

If you can get the next Jim Edmonds for him then sure, but Bottenfield is a terrible comp (journeyman reliever who converted to starter and had a career year at age 30). Ideally you'd probably like Leake to settle into a routine where he's like this guy. Mildly bad case scenario I think you get this guy, because Leake should be better than that.

[SteelSD]

If you can get the next Jim Edmonds for him then sure, but Bottenfield is a terrible comp (journeyman reliever who converted to starter and had a career year at age 30). Ideally you'd probably like Leake to settle into a routine where he's like this guy. Mildly bad case scenario I think you get this guy, because Leake should be better than that.

Bottenfield was an extreme example; selected primary because of the Jocketty tie. You can pick whichever comp you want; Joe Blanton, Kris Benson, Elmer Dessens, Cory Lidle...etc.

[M2]

Bottenfield was an extreme example; selected primary because of the Jocketty tie. You can pick whichever comp you want; Joe Blanton, Kris Benson, Elmer Dessens, Cory Lidle...etc.

I knew you were tossing Bottenfield out there to make a point, that's why I didn't mention it at first. As for comps, I picked them above. He profiles as better than Suppan and he's on the cusp of Buehrle.

Interesting thing I noticed when looking at Blanton and Lidle's numbers (at least interesting to me) is that both of those guys consistently posted ERAs above their FIPs. Lidle absolutely relied upon unsustainably low BABIPs in his good seasons and Blanton's best seasonal ERA/ERA+/WHIP was a BABIP anomaly. It raises the question of what makes them ERA > FIP while a guy like Arroyo (and it looks like Leake might be headed down this road) can consistently ERA < FIP. I suspect it comes down to hittability, either in terms of the frequency or the quality (or both) of the hits the pitcher allows, or perhaps the types of balls he allows into play. It's not HR, at least it's not that flat HR/9. Yet there's a dividing line there and it would seem to based on balls in play.

[lollipopcurve]

Yet there's a dividing line there and it would seem to based on balls in play.

I agree. BABIP will go much the same way as UZR. The use of better cameras will allow better information re: balls in play (velocity, trajectory vs, fielders' positioning), and this will allow us to unpack the BABIP baggage.

[wheels]

If you can get the next Jim Edmonds for him then sure, but Bottenfield is a terrible comp (journeyman reliever who converted to starter and had a career year at age 30). Ideally you'd probably like Leake to settle into a routine where he's like this guy. Mildly bad case scenario I think you get this guy, because Leake should be better than that.

Sorry, should have been more clear.

I would be happy if Walt was able to parlay Leake into a Bottenfield - like return.