Isolated Discipline and Prospect Evaluation

[SteelSD]

So Steel, gotta ask you two things here. One, what does your name stand for? And two, what are your feelings on Delmon Young? He is nearly by all acounts is the best prospect in baseball, but well, he also has a small margin between his AVG and his OBP of just 0.053 for his career.

1. I'm a Steelers fan (yes, I am a bad man) + the state in which I live.

2. I think Young has a shot. His IsoD was actually decent as an 18 year old but Tampa has rushed him tremendously because of his power potential. That hurts him. IMHO, the best shot Tampa has at capitalizing on his power potential is to send him back to AA pronto and allow him to work on his skill sets in an environment more conducive to his age. As-is, it appears they're counting on him to be a high-BA/high IsoP/low IsoD hitter. Rare creatures those are. Challenge thy hitters. But for gosh sakes don't hurt 'em and that's exactly what Tampa is most likely doing to Young.

[dougdirt]

1. I'm a Steelers fan (yes, I am a bad man)

Well at least you know you are in the wrong.

[Outshined_One]

Quickly, BP has been nice enough to post stats and translations for each team's system.

Link

[Outshined_One]

To quote Rocky...I didn't hear no bell. Next round!

You're right. Walks are not part of SLG. But your scenario didn't mention that both players would have identical Isolated Power numbers. If they did, then Player A would be the more productive of the two over the course of a season because his SLG would be higher.

Correct, but that was the point of my hypothetical. One of OBP's biggest flaws is that it rates all attempts at getting on base the exact same (1), be it a home run, HBP, or IBB. Those two guys had similar OBPs over the course of the season, but because Player A made more contact over the course of the season, he is more valuable because of that ability than Player B.

If both had similar IsoPs (let's say .070 for the sake of argument) Player A would have put up .300/.350/.370 and Player B would have put up .275/.350/.345, then there's still a gap between the two players in terms of their offensive output over the course of a season.

But, moving onward...

I applaud you for doing correlation numbers in regards to IsoP, IsoD, SecA, and RC/27. In all honesty, that kind of work is something for which I unfortunately do not have the time or patience. However, there is something that I am curious abuot in regards to the way you accumulated and spread out these correlations, as well as the application of said numbers. To put it bluntly, in these kinds of situations, I become very leery of spurious correlation and attempts at universality.

For example, in your primer, you correlated the numbers on a universal level. In my discussions to this point, I have specifically isolated power hitters because of the fact that they receive different instruction and different expectations as they move up the proverbial food chain. That is why I talked about that in my previous post; these guys are taught tools which they use in order to make contact with the ball and to properly harness their power. I would be curious to see how the numbers line up with those hitters, especially in the minor leagues as it relates to their development.

You hear a lot of people say that power is the last tool to develop, for example. That's not necessarily because these guys are rail thin when they are brought up and then have their helmet sizes increase tenfold as their careers go on (*cough*), although those guys do exist. More often, the guys with tremendous power tools are able to put on shows at batting practice, yet have a hard time translating it into the major league game until they are able to make the proper adjustments in their swing and approach at the plate. This is not a trait that is universal, however. A lot of guys in the major leagues simply do not have the ability to hit a significant number of HRs over the course of the season and therefore employ different approaches at the plate.

I'll happily agree that K/BB is not a very good stat to use for a number of hitters. Guys who lack that kind of power will not be fairly evaluated by that stat. I could care less about how often they make contact as long as they get on base at a reasonable clip. Those guys rarely come to the plate in situations where they will be looking to drive the ball. However, when it comes to these power guys, I am much more hesitant to overlook certain metrics because of that exact approach at the plate I have been talking about recently. I think this is something that can be covered statistically, but it requires a more in-depth and varied look. IsoD is a great statistic...but is it really the be-all, end-all when it comes to this kind of evaluation and projection? Or are there enough gaps in it that it could be detrimental to use it while ignoring other stats on a universal basis?

I find it amusing, though. We both clearly favor guys who have good plate discipline and understanding of the strike zone. However, I prefer a power hitter who is able to make regular contact with the ball in order to drive it. I want a guy with a strong understanding of the strike zone so that he can make contact with the ball on a regular enough basis to put himself in scoring position every time he comes to the plate. On the other hand, you seem to prefer a guy who, simply speaking, does not make outs. As I mentioned before, I place that second before my own preference with a power hitter. You place my preference second to your own on a universal level.

More in my next post...

[Outshined_One]

Something is gnawing at me in these discussions. It did not occur to me what it was until that discussion about pressure situations made me realize what precisely was bothering me. Essentially, it boils down to this:

Strikeouts are seen as just another out and typically are ignored or discarded with the vast majority of statistics you have gone over thus far. If a guy can get on base at a reasonable clip, what does it matter how he makes his outs?

So...if that is the case, why do we place so much emphasis on strikeouts for pitchers? Why do a number of people evaluating pitching prospects include strikeouts in those evaluations? What about the flip side of the coin?

I have done my fair share of prospect watching with pitchers (much moreso than hitters given that the Cubs have had a wonderful track record of position players) and the first numbers I find myself looking at are Ks and K/9 (along with BB/9, HR/9, HBPs, and WPs). Pitchers who have high strikeout numbers are often elevated above other pitching prospects. Let's face it, one of the big reasons people around here are giddy about guys like Bailey, Cueto, and Wood is because they have shown the ability to strike a lot of guys out.

It stands to reason that we should evaluate why the strikeout is important to a pitcher. The single biggest reason is because it is essentially a guaranteed out. There are rare cases where a player can reach base on a strikeout, but in the overwhelming majority of cases, the guy simply walks back to his dugout and takes a seat. There is typically not much of a defensive element involved in a strikeout save for the catcher's responsibility of receiving the ball and, if called upon, tagging the batter out or throwing to first.

However, with regards to other outs, more factors come into play. In simple terms, it is the only way to actually get a hit. We have seen many instances of batters hitting bad pitches, only to reach base. It requires the defense to take on responsibilities and make a play. Putting the ball into play also gives the runners an opportunity (or necessity, if that is the case) to advance. The term "productive out" comes to mind. Granted, it is usually a dumb idea to waste an out to advance or score a base runner unless in specific cases, but they do tend to produce results. Finally, and more rarely, most errors occur when a hitter makes contact with a ball and forces a defender to make a play. It is foolish to rely on the other team making errors, but at the same time, they do provide more opportunities.

Using only strikeouts in evaluating a pitcher is rarely a good idea, but it is something that helps give a stronger picture of a pitcher and his performance. Pitchers who strike out numerous batters in the minors on every level will likely be able to carry that ability over into the majors. It shows a good ability to command and control the strike zone. It can be indicative of the effectiveness of a pitcher's pitches. Strikeouts provide a better bounty of data than most other statistics when it comes to pitchers.

So, I hope you can understand why I am hesitant to throw out strikeouts with hitters. I find the double standard applied to them to be contradictory when used with hitters and pitchers.

[TRF]

This notion of strikeouts being bad is where I have disagreed with most of the "strikeouts are just another out" crowd. Funny thing is, I'm in that crowd.

I think they are bad, but only in the following situation: Extreme strikeout pitcher facing a team of extreme K batters. Here is why.

If said extreme K pitcher can pitch deep into the game, you are doomed. 7+ innings and at least a K per inning gives the opposing team 2 outs or less to do their damage. Smaller margin of error. Now wait. Harang had 7 K's last night. He also pitched only 5 innings and put 9 runners on. Not good. The extreme K pitcher that puts up 10+ K's in 7 innings is going to keep the hits and BB's low. teams prone to high K's are especially susceptible, because said pitcher ISN'T going to walk anyone. Therefore we have removed in essence one half (the most important half) of OPS. So we have to rely on SLG to do damage.

I know that these pitchers beat every team, but I believe some teams are easier prey for them than others.

All that said, it is an extreme case. These high K pitchers are not on every team, but Roy Oswalt had been killing the Reds for years because of this exact scenario.

[RedsManRick]

The difference in the way strikeouts are treated when looking at pitchers and hitters has a lot to do with what the ability to strikeout a guy (or not strikeout) says about the rest of the player's skill, particularly when looking at the predictive context.

History has shown that K/9 rates are more highly correlated with future ERA than is present ERA. Thus we can pretty generally say that a high K/9 is a good indicator of future success (and visa versa). Surely there are a lot of other variables involved, but the positive correlation is strong. This is largely because there is little downside to a high strikeout rate for a pitcher. There is no other skill set which is strongly correlated with the ability to strikeout guys that would balance it out (e.g. high K rate pitchers give up lots of homers).

However, with hitters, while most other types of outs are indeed BETTER than strikeouts, that relationship is dwarfed by the fact that there is a pretty strong correlation between high strikeout rates and power. Furthermore, the benefit of power grossly outweights a high strikeout tendency. Because of the nature of the strikeout, a player who strikes out a lot and doesn't hit for power, isn't likely to stay in the majors (presumably because he doesn't hit for average either). Therefore, the only real correlation between strikeouts and batting production is that the more strikeouts, the more production. Now of course we know this isn't a causal relationship. Sure, if Dunn could hit 45 homers and never strikeout, he'd be better. But that's not what we're looking at. We're trying to say, "what's the predictive value of strikeouts for hitters", and given the way it shakes out, the answer is not much. Because of selection bias based on skill set correlation, we cannot look at a hitters strikeout rate and make a general statement about his future productivity.

I guess in my head I see it this way. On the X axis graph strikeout rate of MLB pitchers (or hitters) and on the Y axis, productivity (measured in some general fashion like VORP):

Hitters: As you move left to right along the strikeout axis, productivity is all over the map, but generally increases. This would be shown as a moderate, positive R value (help me out here with a real number Steel). However, common sense tells us that strikeouts themselves are not a cause of greater production. So instead, we simply say that there is some other variable (the outcome of non-strikeout events) affecting both production and K rate which causes the observed correlation.

Pitchers: As you move from left to right on the K/9 axis, productive consistently and strongly increases. This shows up as a strong positive correlation (R-value... again, the real number Steel?). However, this relationship also passes the common sense test.

Part of the reason for this is that the strikeouts are distributed across the opposing hitters of varying skill sets. That is, assuming all pitchers face the same types of hitters, more strikeouts = less balls in play = less runs. There is no significant & consistent correlation between strikeout rates and the types of balls hit against said pitcher. With hitters, this relationship does exist and counfounds the relationship.

[SteelSD]

Correct, but that was the point of my hypothetical. One of OBP's biggest flaws is that it rates all attempts at getting on base the exact same (1), be it a home run, HBP, or IBB. Those two guys had similar OBPs over the course of the season, but because Player A made more contact over the course of the season, he is more valuable because of that ability than Player B.

If both had similar IsoPs (let's say .070 for the sake of argument) Player A would have put up .300/.350/.370 and Player B would have put up .275/.350/.345, then there's still a gap between the two players in terms of their offensive output over the course of a season.

If you're equalizing IsoP, then you are correct- Player A would end up with a .025 SLG advantage. That being said, when speaking to player projection, which of the two players is more likely to replicate that IsoP as he climbs the minor league ladder? There are variables, of course. Size, injury, instruction, etc. But assuming all of those things are equal, the correlations tell us that Player B has a better chance of replicating a .070 IsoP as he moves up because of his more advanced plate discipline skill set.

BTW, here's a list of the MLB players from 2000-2005 who produced a .300 BA with a .050 IsoD and a IsoP lower than .100:

Juan Pierre (2004)
Tony Womack (2004)
Luis Castillo (2003)
Juan Pierre (2002)

That's it. Castillo is an outlier in that group because 2003 was the lowest IsoD of his career. There are some guys (Ichiro Suzuki) who get near that profile, but even near that profile you'll find a ton of volatility as you do within that profile. Ditto for low IsoD/high IsoP power hitters.

But it's interesting to see Castillo's name pop up there along with Pierre's as Castillo was a .100 IsoD/low IsoP player at the minor league level while Pierre was a low IsoD/low IsoP player. Here are the minor league and MLB numbers for each:

Castillo Minors: .303 BA/.404 OBP/.358 SLG
Castillo Majors: .292 BA/.369 OBP/.356 SLG

Pierre Minors: .330 BA/.377 OBP/.392 SLG
Pierre Majors: .300 BA/.350 OBP/.370 SLG

Now the percentages:

Castillo MLB- % of Minor League Rate Stats:

BA: 96.4%
OBP: 91.3%
SLG: 99.4%

Pierre MLB- % of Minor League Rate Stats:

BA: 90.9%
OBP: 92.8%
SLG: 94.3%

Now Minor League K rates:

Castillo: 1.08 BB/K
Pierre: 1.21 BB/K

Castillo: 6.11 K/AB
Pierre: 14.50 K/AB

The key variable in accurately projecting future performance for those two players is IsoD. And look at how high Pierre's minor league BA had to be to even project anything resembling a decent OBP at the MLB level. While similar from an OPS standpoint over their careers, one thing is certain- Pierre is easily the more volatile of the two players. Even this season- not very good from Castillo's standpoint- we're looking at Pierre's second consecutive sub-.700 OPS effort; and during his age-prime seasons to boot.

If Pierre finishes this season anywhere near his current range, it'll be the third time in six full seasons he's failed to produce an acceptable OBP (.350+). If Castillo's 8th full season ended today, it would mark the second time he's failed to hit that OBP number. And his two misses (2001, 2006) would be .344 OBP and .346 OBP seasons. That's what you get when you have a higher IsoD player- hits that are more consistent and when "misses" happen, they're closer to the acceptable range.

And please note both the BB/K and especially the K/AB minor league differential. The BB/K rate looks fairly close. But Pierre's K/AB rate obliterates Castillo's. Yet Pierre is the guy who saw more degradation to his BA once he hit the MLB level. That also led to more degradation of his SLG. And it's happened because it's improbable that any player will reproduce such a high minor league batting average with so little power at the MLB level regardless of their contact rate. Even really fast guys.

From the minors to the Show, Pierre's IsoP has increased from .062 to .070. Castillo from .056 to .064. That's a wash even though Pierre has recently been acquiring a bunch of Triples to boost it.

I'll make an effort at the remainder of you excellent post(s) later.

[RedsManRick]

There was a point which Steel referenced here and I believe was made earlier, but that easily gets lost in all this.

If you are asking the question, "which season is better in retrospect?"
A: .300/.350/.450
B: .250/.350/.450
The answer is clearly A as those extra singles are better than walks.

However, when looking at the players who put up those lines and predicting future production, we would choose player B for all the reasons Steel has mentioned, particularly focusing on the consistency, repeatability, and reliability of each component. Player A will likely never become more valuable than his current line.

[TRF]

There was a point which Steel referenced here and I believe was made earlier, but that easily gets lost in all this.

If you are asking the question, "which season is better in retrospect?"
A: .300/.350/.450
B: .250/.350/.450
The answer is clearly A as those extra singles are better than walks.

However, when looking at the players who put up those lines and predicting future production, we would choose player B for all the reasons Steel has mentioned, particularly focusing on the consistency, repeatability, and reliability of each component. Player A will likely never become more valuable than his current line.

wait. If player A hits .300, and player B hits .250, then player A's singles are not more valuable than player B's walks, because player B has more extra base hits right? or do i have that wrong?

[RedsManRick]

wait. If player A hits .300, and player B hits .250, then player A's singles are not more valuable than player B's walks, because player B has more extra base hits right? or do i have that wrong?

I thought about that but unfortunately didn't have time to play with the numbers. So I played it out and here's what I came up with.

Player A (600 PA):
167 hits (109 singles, 42 doubles, 6 triples 10 homers)
43 walks

Player B (600 PA):
130 hits (73 singles, 25 doubles, 0 triples, 32 homers)
80 walks

You could look at it this way. Which would you rather have?
36 singles, 17 doubles, 6 triples OR 37 walks and 22 homers?

The answer to your question is that it can go either way. Homers have a drastic impact on SLG%, but reduce your count of extra base hits as well. Either player could have more extra base hits. The question becomes, do you think a homer is really twice as valuable as a double and how much more valuable is a single compared to a walk?

Player A concievably could have 130 singles and 37 homers. Or he could have 83 singles and 84 doubles. The slugging percentage is the same. Player B could concievably have 90 singles and 40 homers or 9 singles and 121 doubles.

So as you can see, the "problem" with player B is that he definitely had 37 times where there was no way for a runner to advance unless he was forced compared to Player A. Of the 130 hits which they "share", the composition of those hits is such that neither appears to have a distinctly more beneficial distribution.

Thus, given that neither player is significantly more productive (waiting to be corrected here...) at that .800 OPS production level, what is the likely development path for each player? I think history, and Steel, has proven that Player A will likely never be much more than what you see whereas Player B tends to develop in to the greater offensive force due to the way certain skill sets develop.

[Cyclone792]

From Rick's above example of Player A and Player B, and considering all other things (HBP, SB, CS, GDP) would be zero:

• Player A would have 92 runs created
• Player B would have 95 runs created

Not identical, but not different enough to be significant either. It's pretty much a coin flip for trying to determine which of those performances was better if they occurred in the past.

As evidenced by Steel's research in this thread, however, as far as future projections go, give me Player B every single time.

[RedsManRick]

Thanks Jason. What is really interesting is that you can see the value in using OPS (OBP + SLG) in a raw comparison of "which player is more productive?" While two players can have vastly different skill sets, if their OPS' are close, their run production is close as well.

Obviously this raises the question of SB/CS, GIDP, etc., but the impact of those things is tiny in the whole scheme of things. In a recent BP article, they mentioned that Jose Reyes' speed makes his .310 OBP more like a .340 OBP with no speed. Consider that SB can't drive in runs and just plop that extra .030 on to the OPS and you can see that even being the best base stealer in baseball doesn't change your OPS very much.

1 HR (instead of a single) = about +.006 OPS over 500 AB.
If we takes Reyes' approx 60 SB per year to equal .030 points of OPS, that's .0005 per SB.

In other words, 1 HR = 12 SB.... now if that doesn't tell you how skewed scoring is a 5x5 roto league relative to real contribution, I'm not sure what does...

BTW, according to this math, the "average" Tony Womack would need to steal about 468 more bases per year to be as good as Adam Dunn. That is, he needs to steal 2nd, 3rd, and home every single time he reaches base...

[SteelSD]

But, moving onward...

I applaud you for doing correlation numbers in regards to IsoP, IsoD, SecA, and RC/27. In all honesty, that kind of work is something for which I unfortunately do not have the time or patience. However, there is something that I am curious abuot in regards to the way you accumulated and spread out these correlations, as well as the application of said numbers. To put it bluntly, in these kinds of situations, I become very leery of spurious correlation and attempts at universality.

Leery you should be and "question everything" is as near a motto as I have.

For example, in your primer, you correlated the numbers on a universal level. In my discussions to this point, I have specifically isolated power hitters because of the fact that they receive different instruction and different expectations as they move up the proverbial food chain. That is why I talked about that in my previous post; these guys are taught tools which they use in order to make contact with the ball and to properly harness their power. I would be curious to see how the numbers line up with those hitters, especially in the minor leagues as it relates to their development.

Well, I think part of that is the fact that I don't necessarily believe in the concept of "power hitter"; only "productive hitter". Do high SLG and high IsoP hitters tend to produce more than those at the lower end of the spectrum. Certainly. That being said, I'm a firm believer that regardless of instruction at the minor league level a hitter's two primary functions trump anything else (i.e. make few Outs, acquire as many bases as possible when not making an Out).

So rather than just focusing on power, I'd rather focus on production. That's why you'll see me tend to not compartmentalize- because every hitter has the same basic primary functions. However, when looking at the top RC/27 hitters (and top OPS hitters if we want to go there) that list will include the top "power" hitters in the game and, more importantly, will also include the most productive hitters in the game. So here's the sort:

1) First, I selected the top 40 2005 MLB RC/27 hitters.
2) I sorted those 40 players in decending order using HR/AB ratio just to make sure we left nothing to chance.

For 2005, I only used the best of the best and that includes the best of the best "power" hitters. Here are the correlations with high HR/AB rate behavior:

K/AB: 0.361
IsoD: 0.162
BB/K: -0.254
BB/PA: 0.105

That's striking. K/AB rate is more highly correlated with top power production than was IsoD, BB/K, and BB/PA. In fact, high BB/K rates have a significant negative correlation with power production. We'd be better off simply looking at BB/PA rate. But we'd be even better off looking at IsoD.

And just to note how this all works, it's not like we need to be looking for players who strike out a ton. The K/AB correlation is a residual of power production- not a driver. Some portion of IsoD may be a residual, but that residual is an effect produced by the hitter (and things like IBB are produced by the hitter). It'll sound stupid to some folks, but every one of Barry Bonds' ridiculous 628 career IBB were caused by Barry Bonds and they were a result of his IsoP which was partially driven by his natural plate discipline and pitch recognition skill sets (IsoD).

And just in case you're wondering (because I did) about correlations within that 2005 40 most productive players...

IsoP to HR/AB: 0.933 correlation
IsoD to IsoP: 0.108 correlation
BB/K to IsoP: -0.246 correlation

If we're looking for a HR rate driver, that's IsoP. If we're looking for a secondary skill set driver for HR rate, that assuredly isn't high BB/K rates. Why? Because high K rates are a residual of high HR rate production. Now, more advanced mathematicians may want to question my methodology or my conclusions. That's no problem. I'm fine with that. But if we're looking at a driver/driver/driver scenario while confining our search to the most productive hitters and then sorting our sample by HR/AB rate while finding that IsoP has a 0.933 correlation with the behavior you're looking for and then finding that BB/K has a negative correlation with IsoP...well, that's a tall order.

You hear a lot of people say that power is the last tool to develop, for example. That's not necessarily because these guys are rail thin when they are brought up and then have their helmet sizes increase tenfold as their careers go on (*cough*), although those guys do exist. More often, the guys with tremendous power tools are able to put on shows at batting practice, yet have a hard time translating it into the major league game until they are able to make the proper adjustments in their swing and approach at the plate. This is not a trait that is universal, however. A lot of guys in the major leagues simply do not have the ability to hit a significant number of HRs over the course of the season and therefore employ different approaches at the plate.

You're right, but we've already established that hitters have two primary functions. The best guys within a "low power" subset tend to be the guys who are able to get on base any way they can. That, of course, includes free swinging "plate coverage" guys as well as high IsoD guys. The difference is that the high IsoD guys tend to work out better over the long haul because they've got a developed secondary skill set to compensate for a loss of bat speed and/or foot speed. And those high IsoD guys tend to be more consistent over time and hold their projections.

The free swinging low IsoD/high IsoP plate coverage guys can't be ignored and they ARE useful. That being said, they're useful for a shorter period of time, tend to manifest their abilities closer to their physical primes, and tend to flame out quicker than the high IsoD/high IsoP or high IsoD/low IsoP guys.

I'll happily agree that K/BB is not a very good stat to use for a number of hitters. Guys who lack that kind of power will not be fairly evaluated by that stat. I could care less about how often they make contact as long as they get on base at a reasonable clip. Those guys rarely come to the plate in situations where they will be looking to drive the ball. However, when it comes to these power guys, I am much more hesitant to overlook certain metrics because of that exact approach at the plate I have been talking about recently. I think this is something that can be covered statistically, but it requires a more in-depth and varied look. IsoD is a great statistic...but is it really the be-all, end-all when it comes to this kind of evaluation and projection? Or are there enough gaps in it that it could be detrimental to use it while ignoring other stats on a universal basis?

IsoD is NOT the "be-all, end-all" when it comes to player projection. Never would I attempt to claim that. We're not looking at absolutes; only degrees of "better" and I've found IsoD to be a hidden gem in the "degrees of better" department. And I've also found that it does apply universally. Hitters with better secondary on-base skill sets project to achieve numbers closer to their previous performance, they tend to last longer, and they tend to be less volatile than low IsoD players because Batting Average hitting is a mostly random event. In fact, high IsoD/high IsoP hitter can, over time, minimize that randomness because they project to hit more balls per AB that the defense can't deal with (i.e. Home Runs).

I find it amusing, though. We both clearly favor guys who have good plate discipline and understanding of the strike zone. However, I prefer a power hitter who is able to make regular contact with the ball in order to drive it. I want a guy with a strong understanding of the strike zone so that he can make contact with the ball on a regular enough basis to put himself in scoring position every time he comes to the plate. On the other hand, you seem to prefer a guy who, simply speaking, does not make outs. As I mentioned before, I place that second before my own preference with a power hitter. You place my preference second to your own on a universal level.

See, we're soooo close here. You want a guy with a strong understanding of the strike zone. I want a guy with a stong understanding of the strike zone. The only difference I see is that I don't see any evidence that contact rate has much of anything to do with understanding the strike zone.

[SteelSD]

Something is gnawing at me in these discussions. It did not occur to me what it was until that discussion about pressure situations made me realize what precisely was bothering me. Essentially, it boils down to this:

Strikeouts are seen as just another out and typically are ignored or discarded with the vast majority of statistics you have gone over thus far. If a guy can get on base at a reasonable clip, what does it matter how he makes his outs?

That's a good question and, for the most part, it really doesn't matter.

So...if that is the case, why do we place so much emphasis on strikeouts for pitchers? Why do a number of people evaluating pitching prospects include strikeouts in those evaluations? What about the flip side of the coin?

Well, there's the million dollar question. The answer is that we don't like high-K rate pitchers. We like high K rate/reasonable BB rate/low HR rate pitchers. That's what we call (well, what I call) the "trifecta". High K rate pitchers are fine but I can tell you right now that I've been at odds with folks who've positioned K rate without considering the rest of the mix.

Bascically, the pitchers we should like are the inverse of the hitters we should like. The only difference is that the pitchers we should like pitch against the average and their K rates work to minimize the damage on Balls in Play from the average hitter. But high IsoD/high IsoP players aren't the "average" hitter and they take defense out of the mix far more often than the average hitter.

And because hitters are hitting for themselves- even against high K-rate pitchers- they can maximize their offensive output by utilizing IsoD behavior to better select the few pitches they can drive (resulting in better IsoP output). Basically, the reason the "trifecta" pitcher can produce is that he has a better chance of neutralizing the average hitter. But a team stocked with high IsoD/high IsoP/high K hitters can strike quicky and strike hard. Ditto for a high IsoD/high IsoP/high K hitter. And against the "trifecta" pitcher, secondary base acquisition skills really matter.

I'll use Dunn as an example again here. His may, from a BA (.212) and OBP (.328) standpoint wasn't anything near optimal. But his SLG (.535) was solid and that presents a problem even to high K-rate pitchers. The interesting thing was that Dunn's K/AB rate was actually lower in May than during his 1.046 OPS April, but that's common with the kind of hitter we're talking about.

And yet, hitters like that will continue to produce even at their worst. Why? Because they have the ability to produce high IsoD/IsoP/SecA performance even when slumping because, when slumping, they still have the ability to take the defense out of play (BB, HR). That's where IsoD projection comes in because it allows us to find "slump-proof" players. And that, despite any protestation, is what high IsoD/high IsoP players (and teams) represent.

Teams need those guys because they can help pick apart high-K pitchers. They keep them honest. They may acquire fewer non-HR Hits versus that profile, but their ability to reach base will allow for more opportunity for the team.

I have done my fair share of prospect watching with pitchers (much moreso than hitters given that the Cubs have had a wonderful track record of position players) and the first numbers I find myself looking at are Ks and K/9 (along with BB/9, HR/9, HBPs, and WPs). Pitchers who have high strikeout numbers are often elevated above other pitching prospects. Let's face it, one of the big reasons people around here are giddy about guys like Bailey, Cueto, and Wood is because they have shown the ability to strike a lot of guys out.

And you are completely accurate. I'd say that you're looking at the right numbers.

It stands to reason that we should evaluate why the strikeout is important to a pitcher. The single biggest reason is because it is essentially a guaranteed out. There are rare cases where a player can reach base on a strikeout, but in the overwhelming majority of cases, the guy simply walks back to his dugout and takes a seat. There is typically not much of a defensive element involved in a strikeout save for the catcher's responsibility of receiving the ball and, if called upon, tagging the batter out or throwing to first.

However, with regards to other outs, more factors come into play. In simple terms, it is the only way to actually get a hit. We have seen many instances of batters hitting bad pitches, only to reach base. It requires the defense to take on responsibilities and make a play. Putting the ball into play also gives the runners an opportunity (or necessity, if that is the case) to advance. The term "productive out" comes to mind. Granted, it is usually a dumb idea to waste an out to advance or score a base runner unless in specific cases, but they do tend to produce results. Finally, and more rarely, most errors occur when a hitter makes contact with a ball and forces a defender to make a play. It is foolish to rely on the other team making errors, but at the same time, they do provide more opportunities.

Well, that's a reasonable take as long as you factor out "Productive Outs" (which are nothing but accidental events). Now, Errors can play into things, but how often? Does Error rate trump GIDP rate? No, but Error rate might be a mitigating factor. Problem is that for a hitter to hit more balls into play, we need to see a behavioral change and we know that it's unlikely that a high K player reaching a level at which Error rate factors in. But that has nothing to do with player projection because, as you noted, error rate for an individual hitter (especially a high-IsoP hitter) doesn't really factor in. Now, it might for the fastest of the fast and that shouldn't be ignored but that holds true for both high and low IsoD speed hitters.

Using only strikeouts in evaluating a pitcher is rarely a good idea, but it is something that helps give a stronger picture of a pitcher and his performance. Pitchers who strike out numerous batters in the minors on every level will likely be able to carry that ability over into the majors. It shows a good ability to command and control the strike zone. It can be indicative of the effectiveness of a pitcher's pitches. Strikeouts provide a better bounty of data than most other statistics when it comes to pitchers.

K rate does have a high correlation with pitcher success, but only when paired with reasonable BB rates and low HR rates (or low BB rates and reasonable HR rates). That's the "trifecta" and I'll encourage anyone to use it.

So, I hope you can understand why I am hesitant to throw out strikeouts with hitters. I find the double standard applied to them to be contradictory when used with hitters and pitchers.

I understand, but it's not really a double-standard and, thus, isn't at all contradictory. K rate for pitchers does correlate with long-term success, but as high K-rate hitting is a residual of power output for hitters, we have a 180-degree reversal. And, yes, it's important to differentiate that K rate for pitchers is a driver while K rate for hitters is a residual.