Adam Dunn is Better Than Tony Gwynn

[kflo]

Originally Posted by nanwynnfan

Hardly. It's clearly the guy who creates fewer non-productive outs. The RC/27 Outs, which I have used myself, can be seen to make a two-faced presentation in favoring the power guy, by stacking the deck with 9 of him, using stats that favor the power guy from the start. I've used it, and I plead guilty in that regard.

productive outs is far less than the bigger ticket stats we're talking about. all else equal, you'd go with the guy who gives you more non-dp balls in play. but i wouldn't simply discount rc/9 and ops or eqa advantages because of an advantage in productive outs. generating fewer outs trumps productive outs.

Yet Another Productive Outs Article -- The Hardball Times

The Truth About Productive Outs -- The Hardball Times


PRODUCTIVE OUTS - 3/17/06 - Amazinz.com™ - New York Mets Forum

Dan Agonistes: Productive Outs

Originally Posted by nanwynnfan

Okay. That's 42 GIDP in 600 AB for Gwynn, a record. In K's it's 185 K's for Dunn ... [I believe I mixed the two in my earlier post]. So, the corrected difference is 42 DP [not 28] for Gwynn; and the atrocious 185 K's stays the same for Dunn.

and 42 dp's is far worse than 185 k's for a team.

Originally Posted by nanwynnfan

I'm not about to convince you. I am well aware of that. However, I would urge readers of these posts to examine both Gwynn's and Dunn's career stats [I generally refer to Baseball-Reference] and just see the overall picture: a slugger who has batted as low as .215, whose TB are marginally = to his non-power opponent, his K's difference, the comparable distributions of extra base hits, and yes the almost damnable BA, the OB%, the steals and caught stealings, the sacrifice hits + sac flies per year ... tell me these guys are on the same offensive plain.

eqa factors in sb, sf, sh, sb, cs. i'm open to the notion that avg is more important, but haven't seen substantial evidence to support it.

Originally Posted by nanwynnfan

Major point. Dunn hits 40 Hr's in a season = 160 TB in that category vs. say 12 HR's for Gwynn = 48. That's a differential of 112 TB in one category with Dunn > Gwynn.

That basic element feeds into most RC statistical comps and gives a huge edge to the power guy. The power guy will also have pitchers playing at the edges to avoid the long ball, so their trade-off is the BB, which further props up the power guy in OB%; but here I would interject that the negative for a Dunn here is that the bat has been taken out of his hands.

So, at year end, if Gwynn bats .333 with 257 TB and Dunn hits .255 with 288 TB, the 112 TB expectation has evaporated down to 31; and Gwynn's 200 hits have the promise of greater advance for men on than the additional BB's issued to Dunn.

It's all nuance and layering; and the stats as presently emphasized have a bias for the long ball.

the stats as presently emphasized have the highest correlation to actual runs scored. that's why they're used.

Originally Posted by nanwynnfan

Wish list: Take down all OF walls for a season, allowing an OF strategy to play at whatever depth. Then shout, "Play ball!"

I wonder how many extra 3B and inside-the-park HR's a guy like Gwynn would collect and how many HRs would be long singles or doubles for a Dunn.

I'll say it just this one more time: the HR, once in every 16 PA or once in every 14 AB is along the lines of one every 4 games, is random, park dependent, and does not always come when you want it or need it most. Yet it pads the power guys stats [and paycheck - nature of the game].

which is why it's extremely important and relevant that dunn also gets on base at a high rate. it's not a love of the long ball. it's the importance of obp, and slg. get on base, create fewer outs, produce bases.

[nanwynnfan]

I'll Have to break this response down into several posts, starting with the first link you provided, from which I distill simply this portion:

Link 1: "The truth about Productive Outs is that they aren't something that are good in bulk, but only in limited context. A Productive Out in the ninth inning of a tie game is a good thing, but one in the first inning of any game is merely less bad than other outs. There is a heirarchy of plays in baseball:

Home Run
Triple
Double
Single
Base on Balls/Hit by Pitch
Reach on Strikeout
Reach on Error
Productive Out
Strikeout
Out on Ball in Play
Double Play
Triple Play"

Pretty consistent with my opening statement: Productive outs>K's a DP's. {and I didn't have to subject myself to the gyrations of whom to include and whom to exclude in an in depth study to get there}. Also note reach on error in the pecking order which happens far more often than reach on K, which author finds of greater value for some reason.

Also, please note that the core of my argument is CONTACT, of which the productive out is a byproduct, a benefit. I would never argue in favor of any kind of out "in bulk."

Link 2: As for the second link, after deciding special sabremetric conditions for what qualifies as productive outmaking, we do get this admission:

"A look at all teams shows that while productive outs tend to come at higher rates for teams that win, getting on base seems to be much more important. Winning teams over the past two postseasons had a .360 POP compared to the .301 POP of their opponents, a difference of .059. But winning teams also had a .364 OBP in those situations, compared to a .266 OBP for the losers, a difference of .099. Taking productive outs as a percentage of all opportunities, and not just outs made, you find that the difference between winning teams and losing team is only .008 -- .229 to .221."

Since winning is the main idea, any contribution is graciously accepted. Pleasw note that these writers in the links provided are so preoccupied as seein productive outs on a team level that they are a total departure from my POV, which is making comps between/among individual players.

Link 3: "The idea of productive outs has largely been a perplexing one. While I think we've established that there's no such thing as an "ability" to make productive outs at a team level, I was never quite sure what value to place on productive outs in comparison with other offensive outcomes. They do, of course, have benefits that strikeouts don't (advancing runners), and this action is better than not being able to. The question is how much these small gains are worth over time.

If you consider a productive out an out where a runner advances, then it's run value can be found using the Expected Runs Matrix. This data is from an old Baseball Prospectus guide, for a situation with a runner on first and no one out:

CODE

Base/Out Exp. Runs
1st/0 out .8998
2nd/1 out .6904
1st/1 out .5407

In runs above an unproductive out, the PO's value is .1497 runs."

Here a quantitative evaluation is presented, a step beyond my own effort. Thank you.

Link 4: from Dan Agonistes, whose work I admire:

"In order to calculate which outs are actually productive you simply look at the before and after state for the batter making the productive out. As in our previous example, before the productive out the run potential was .953 and after .725 so the run potential was decreased by .228 by the productive out. If the batter drove in the run via a ground out or sacrifice fly a 1 would be added to the run potential of the final state. The same can be done for the probability of scoring at least one run."

Essentially, Agonistes takes the run scoring expectancy from the 24 Base-Out Scoring expectaion grid, subtracting the odds after an out from the odds before the out was made.
I have no problem with that, so long as we don't classify and refine beyond that simplicity. My entire argument is that the before-after scoring odds HAVE A CHANCE of looking better via batted-ball-in-play outs > K or DP, which hold forth NO such promise.

Agonistes too, debates exclusion of position players and the sacrifice, which alluded to earlier: defining beforehand waht may or may not qualify as a productive out.

Please note that I did not present sac bunts or sac flies as productive outs, although their proper execution by competent CONTACT hitters is highly probable of maximized positive results.

[nanwynnfan]

Originally Posted by kflo

and 42 dp's is far worse than 185 k's for a team.

Here you attempt to transfer a courtesy hypothetical posed by you [and worked numerically to see the numerical result] as if it were reality. Then, you make the jump to equating the reality that never was into a value judgement without offering any support for that claim.

Originally Posted by kflo

eqa factors in sb, sf, sh, sb, cs. i'm open to the notion that avg is more important, but haven't seen substantial evidence to support it.

All I'll say on this point is that BA was, in isolation, more highly valued in my generation and my father's [1910-1960] than warranted; but BA nonetheless is THE honest measure of ratio of contact with positive offensive result to charged at-bats.

Originally Posted by kflo

the stats as presently emphasized have the highest correlation to actual runs scored. that's why they're used.

That's a pretty general assumption. There are, widely published stats, that are consistently being revisited and massaged, challenged and reworked, many by the "father" of sabermetrics himself, always in search of soething better.

Originally Posted by kflo

which is why it's extremely important and relevant that dunn also gets on base at a high rate. it's not a love of the long ball. it's the importance of obp, and slg. get on base, create fewer outs, produce bases.

And, getting back to the subject at hand, Dunn vs. Gwynn, Gwynn wins @ .388 to .381, if we're going to get meticulously linear in our arguments.

We can debate until the cows come home the relative sabermetric evaluations for more ancillary contributions made for sac hits, sac flies, stolen bases and caught stealing as well. Here, I would argue, that many sabermetric in-depth studies go through mental gyrations in developing formulae, chosing perhaps on one hand to eliminate position players or pitchers, or both, alternating to suit the postulate.

Here I would say is where CONTACT again raises its productive head. A sac situation, a squeeze, the need for a fly or grounder to advance a runner or score a run, are enhanced with contact skills experts holding the bat. Their execution will always exceed regressed projections, just as a gifted base stealer will defy [by craft more than raw speed] the pass-fail decimal values assigned to the steal as an offensive strategy.

Then factor in base running and defense & one can reveal a line drive spray hitter for average as a real value over an inconsistent slugger, who bats < .250 but clubs 40 HRs a year and whiffs 150 times.

[kflo]

my main 2 points have been 1) that dp's and strikeouts are different enough in value that they shouldn't really be added together to represent a single view of "negative outs". that was my point with the 42 and 185, that someone with a negative outs of 42 could be worse than someone with a negative outs of 185. it's an extreme situation, but illustrates the point.

and 2) that i haven't seen evidence to conclude that someone with a much higher ba, much lower k's, but higher dp's is automatically more valuable than someone with a slightly lower obp and materially higher slg. i don't doubt that there's benefit to making contact (to the extent it doesn't lead to an onerous amount of dp's). but that the overall value is fairly immaterial relative to the much bigger ticket items. which is what the links all show.

and results at the team level should translate to the individual level.

i have said that gwynn's fielding and speed (and peak seasons) make him an obvious choice overall. i just don't see where it's obviously one-sided from a pure hitting value standpoint.

as for stats being continually revisited - of course - and i'm open to adjustments for anything that can be shown to be materially relevant. i just haven't seen anything that supports negative outs, or positive outs, as being big difference makers.

[WilsonC]

Originally Posted by nanwynnfan

Here I would say is where CONTACT again raises its productive head. A sac situation, a squeeze, the need for a fly or grounder to advance a runner or score a run, are enhanced with contact skills experts holding the bat. Their execution will always exceed regressed projections, just as a gifted base stealer will defy [by craft more than raw speed] the pass-fail decimal values assigned to the steal as an offensive strategy.

In contrasting the two player prototypes, I do think that each type has his own advantages and disadvantages. For example, in situations with a runner in scoring position and less than two outs, there's an advantage to having a contact bat up - particularly in a situation where the leverage of one run is high. On the flip side, with no runners on, the player with more power has the clear advantage as he's more likely to drive himself in (assuming comparable OBP skills - if there's a significant difference there, it can skew things).

On a perception level, it's easy to give the advantage to the contact hitter, as his advantages tend to come in situations where success is the assumed result, so each failure feels like a greater failure. Conversely, the slugger's advantages come in situations where failure is the presumed outcome, making the successes feel more rare. However, there's also a difference in magnitude. Productive outs rarely score more than one run, and there will always be at least one more opportunity for the next batter to cash in that run in the event that the out fails to be productive. When they're of value, it's typically in situations where the player's team had at least a reasonable chance anyways, whereas power can create a game-changing swing even when the contest isn't as close, or can turn a tight game into a comfortable lead. It's less a frequent event but greater in magnitude. Both styles have their advantages, really.

Originally Posted by nanwynnfan

Then factor in base running and defense & one can reveal a line drive spray hitter for average as a real value over an inconsistent slugger, who bats < .250 but clubs 40 HRs a year and whiffs 150 times.

Inconsistent doesn't seem like the right word for someone like Dunn. I know that's just semantics, but it also helps add to create a negative perception that may or may not be warranted. In some regards, the slugger can actually be the more consistent of the two. For a hitter like Dunn, it takes the defense out of the equation more often than for a hitter like Gwynn. A power-patience hitter like Dunn succeeds or fails based directly on the duel between him and the pitcher, whereas a hitter like Gwynn is more likely to have stretches where his results are inconsistent with his performance during that stretch - you see it all the time when a hitter has a period where he just seems to be making solid contact right at fielders, or where everything he hits just seems to find the holes. The performance feels more consistent, but the results aren't always more consistent.

[nanwynnfan]

Originally Posted by WilsonC

Inconsistent doesn't seem like the right word for someone like Dunn. I know that's just semantics, but it also helps add to create a negative perception that may or may not be warranted. In some regards, the slugger can actually be the more consistent of the two. For a hitter like Dunn, it takes the defense out of the equation more often than for a hitter like Gwynn. A power-patience hitter like Dunn succeeds or fails based directly on the duel between him and the pitcher, whereas a hitter like Gwynn is more likely to have stretches where his results are inconsistent with his performance during that stretch - you see it all the time when a hitter has a period where he just seems to be making solid contact right at fielders, or where everything he hits just seems to find the holes. The performance feels more consistent, but the results aren't always more consistent.

Therein, IMO, lies the ultimate value of the BA as a statistic, in that it refelects a % of charged AB where successful contact resulted in safe hits. The more contact, the more balls in play, the more opportunity for safe hits, the more opportunity to advance runners who may be on base.

I have also mentioned several times in this thread the rather random nature of HRs as events. Of Dunn's 40 HR in 2007, 19 came with the bases empty. Furthermore, if Dunn is ahead in the count, he hits well; but at even counts, or even slightly in the hole, his average plummets to fair-hitiing pitcher levels.

[Triad]

Dunn is not only the only player in history to have three consecutive seasons of exactly 40 HR, but he's also the only player with any three seasons of exactly 40 HR at any time in his career.

46 HR three times:
Babe Ruth

44 HR four times:
Hank Aaron

40 HR three times:
Adam Dunn

39 HR three times:
Jeff Bagwell
Rafael Palmeiro
Jim Rice

38 HR three times:
Mike Schmidt

36 HR three times:
Jimmie Foxx
Dale Murphy

A pretty select group. I guess what typifies them is their consistency from year to year during their primes. Some of them might not have been flashy, but they were quite steady during their best years. I almost expected Steady Eddie Murray to show up on this list. Murray hit within 6 HR of his career high 10 times! That's got to be a record for players who have hit more than 30 in a season. If not for the '81 strike when he led the league with 22 HR, he would have had another, and it would have given him a four-year string of 32-33-32-33. (assuming 150 G that year) And if not for the '95 strike-shortened season, he would have had another in that range, which would have given him 12 seasons within 6 HR of his career high. That is the model of consistency. In fact, his 162 G avg. is at exactly 6 HR less than his career high. I'll bet that doesn't happen too often with home run hitters.

[bedir than average]

So somehow that makes Dunn better than Griffey I'm guessing?

[nanwynnfan]

I won't compare Dunn to Foxx, Griffey, Ruth, or even Guerrero or A-Rod. I'll compare those four consecutive years in which he hit 46, 40, 40, 40 HR and warranted such praise here, against the guy he's > than in the thread title, Tony Gwynn.

Here's a head-to-head comp of the two, ages 24 through 27 [the four years mentioned:

Player......Year.......Age.......HR.......BA/OB%/SLG%......TB......AB........K

Dunn.......2004........24........46....... .266/.388/.569.....323.....568.....195
Gwynn.....1984........24.........5....... .351/.410/.444.....269.....600......23

Dunn.......2005........25........40....... .247/.387/.540....293......543....168
Gwynn.....1985........25.........6....... .317/.364/.408....254.......622.....33

Dunn.......2006........26........40....... .234/.365/.490....275.......561...194
Gwynn.....1986........26........14....... .329/.381/.467....300.......642....35

Dunn........2007.......27........40....... .264/.386/.554....289.......522...165
Gwynn......1987.......27.........7....... .370/.447/.511....301.......589.....35

At first glace, those HR's look impressive; but distill those TB to the bottom line and we see base generation over the four seasons = 1,170 for Dunn to 1,124 for Gwynn, 46 total bases over 4 seasond of HR "consistency."

If I want offense via base generation, I don't want to see a HR total base edge over a four year period [664 HRTB, Dunn; 128 HRTB, Gwynn] squandered largely due to whiffs. Dunn fanned 722 times in 2,194 AB, a 32.91% clip, almost one K for every 3 charged ABs. For that kind of performance, I'm expected to see a fair trade in being rewarded with 166 HR in 2,194 AB, or .0757, a chance of 1 HR every 14 AB, or 1 HR every 3+ games for a guy who walks a lot.

Add to that the fact that half of those will come with nobody on, and I'm seeing a very poor return on my investment.

Let's not even consider comparing Dunn to Foxx or Ruth.

[WilsonC]

I don't think Triad was intending to compare Dunn to Foxx or Ruth in quality, but rather that despite the high K rates, he's been a very consistent player, which is true. The past four years he's been one of the more consistent players in baseball, on a year-to-year basis.

While it's true that half of a player's HR will generally be solo shots, that also means that half will have runners on. Also, what's wrong with a solo HR? In the very same close-game situations where contact-oriented hitting is most likely to generate meaningful productive outs, a single solo HR can be a critical difference maker between winning and losing. And, while HR do happen in a random variety of game situations - some which have an impact on winning, some which don't, the same can be said about the opportunity for meaningful productive outs.

If we look at Dunn's numbers last year, he had 43 K's with RISP. Assuming a random distribution based on the number of outs, that puts him at about 28 K in situations where he could have reasonably gotten a productive out. 12 of those had a runner at 3rd. That's the difference on the extreme end - Adam Dunn compared to Joe Sewell - there'll be typically less than 30 situations a year where the K has the potential to hurt Dunn's value. And, keeping in mind that there'll always be someone batting after him, it's reasonable to expect that roughly 25-30% of his failed attempts the next batter will get a hit anyways, netting the same result as if Dunn had advanced him. Many of those situations will also happen in games where the score's not close enough for it to matter anyways.

Adam Dunn's not as good a hitter as Tony Gwynn was, and you'll not hear me say otherwise. Dunn plays in a better hitters park and a better hitters era, which inflates his numbers relative to Gwynn. It's perfectly reasonable, however, for a superior Dunn-style hitter - say, Mike Schmidt - to be a better hitter than Gwynn, however - K and all.

[nanwynnfan]

Originally Posted by WilsonC

And, while HR do happen in a random variety of game situations - some which have an impact on winning, some which don't, the same can be said about the opportunity for meaningful productive outs.

I prefer to look at it this way: In 600 AB, a guy who whiffs 32.4% of the time
will fan @ 194 times, where, regardless of game situation, nothing positive will happen. He will put the ball in play 406 times & he may connect for a HR 10% of those contacts.

A hitter who fans 7.5% of the time will put the ball in play 555 times, fanning 45 times. While he may not give me the long ball, he will give me 149 extra contact opportunities; and over 162 games, that's 92%, almost one per game. I'll take those odds.

Now, given an open market to complement my contact hitter with a new acquisition for power, given a Dunn or a .285/.365/.510 guy with @ 80 walks and @ 30 HRs and fewer than 90 K's, I'll take the latter, every time.