What Do YOU Expect from a Metric?

[leewileyfan]

It is obvious from the post content in these threads that there are a lot of savvy folks in here with a deep understanding of baseball, its history, its facets and execution of each ... which can also vary by degrees of the good, the bad and the indifferent.

I am in the middle of wrapping up a research work that has been several years in the making. Its focus is purely on defense. It covers every position except pitchers [for no special reason ... just the way I approached it]. It dawns on me that METRICS are presented as self-contained universes, in which mathematical models are devised, sorted, sifted, regressed and presented as EVALUATIVE of performance against whatever standard the metric maker has selected. Defense is subject to all sorts of nuances.

So, the question[s] are [include] these:

1. Do you want a metric that has a broad historic perspective? [My time frame is 1901-Present].
2. Do you want a metric that allows [to whatever degree] player comparisons from different generations of position play? [OR, do you prefer CURRENT evaluative presentations [snapshots]?
3. Do you prefer a metric that attempts to interpret basic data inputs that are uniformly available across the entire time frame being considered? [OR, do you prefer more complex, in-depth mathematical interpretations in which uniformity of available data may be restricted by available research efforts [back to 1950 ... 1970 ... 1988]?
4. How importance are matters of EQUIVALENCE to you? [Focusing on defense here, some/all of the following]:
a. Pitching staff K's recorded as compared to League, both Leagues, between-among teams
b. Park factors, like playable foul territory or HR dimensions;
c. Pitcher-handedness;
d. Pitcher fly ball, ground ball tendencies;

I raise the question of EQUIVALENCE especially with regard to pitching K's recorded. Yes, at the team level, a disparity of 500 K's by two competing staffs has a big impact on batted balls in play and fielder opportunity. By repeated application, I have found a K to be "worth" about .11 defense runs; so in the example cited, 500 * .11 = 55 defense runs [of fielding opportunity].

That K impact is distributed among positions; so that each position gets a piece.

I will try to word this precisely, to avoid any sense of bias:

1. IF a metric focuses on POSITION [individuality] and execution [as a rate of executing input bytes PO,A,DP,E,PB,etc.] do you want RATING to relate EXACTLY what happened on the field of play? IF the metric is constructed to adjust for evolving standards of expectation [equipment and design, playing surfaces, ball structure and dynamics] does that enhance the basic input product, or not?

OR

2. Do you want ALL data input calculations adjusted for pitching K's and OTHER factors that "level" the mathematically ideal model in which INTERPRETATION of EQUIVALENCY considerations are your sole bottom-line concern?

Of course, any additional ideas/challenges are invited as well.

[scorekeeper]

I’m guessing no one’s answering because its difficult to put one’s finger on what they want from all metrics when each one can and often does show something very different to different people who have different things they’re looking for.

I could give you an answer but won’t because my perspective is HS ball, so many of the intricate metric used for ML data won’t even work because there are things unavailable like BPF that don’t even exist.

[leewileyfan]

That's precisely why I posed the question exactly as I did. To narrow the scope, my focus is on defensive play. To cite possible examples that MIGHT encourage responses:

1. Can a defense metric possibly be acceptable if it is constructed [and defined and explained] to allow comparisons among players like Rogers Hornsby, Bobby Doerr, Bill Wambssganss, Dustin Pedroia and Orlando Hudson. What if it attempts to compare OF, like Tris Speaker, Taylor Douthit, Dom DiMaggio, Riche Ashburn, Willie Mays and Jim Edmonds?

2. Can a metric earn any credibility if it exhibits Gary Sheffield as a versatile player for playing 3B and OF, without presuming incredible shortcomings at either/both?

3. How deep a gap is there between fielding prowess/weakness as REACHABLE from basic box score inputs, as opposed to reliance on pitcher handedness, ground-ball fly-ball tendencies, OR play-by-play data that is available for limited periods of time?

The question is put as a direct and sincere POINT of discussion. Sometimes, we [myself included] can accept or reject concepts or presentations NOT because of inherent flaws in whatever is under discussion, but more because we are talking ACROSS each other's points of view because we haven't defined our terms or the context of our focus.

No trick questions here. Just what do you, individually, seek [expect, want] from a metric presented to represent a model for measuring defense, historically and position-specific?

[scorekeeper]

Again, I’m an exceptionally bad person to speak on any metrics designed or intended to compare ML players because my interests only lie in trying to discern what happens at levels lower than college. For defense especially, when there are rules allowing re-entry, it makes the scorer’s and by extension the statistician’s job much more difficult. Because of that, and because there aren’t 1,000 eyes paying attention to who’s playing where, the defensive metrics become very “unreliable” at the levels I’m interested in, to put it mildly.

So what ends up happening is, defensive metrics are seldom if ever used to make decisions as to player positioning, and that leaves most such decisions up to the “perception” of the guy filling out the lineup card. So whatever defensive metrics are used for at the ML level, they’re pretty much useless for levels below college. Sorry, I wish I could help, but I haven’t got the ammunition to take much of a shot. But you have my best wishes and I’ll be watching to see what kind of responses you get.

[leewileyfan]

scorekeeper:

I honestly get it. You are not interested in the topic of the thread because your personal interests lie at levels below MLB. However, I started the thread in hopes of getting some feedback from fans who ARE interested in defense metrics, their credibility, their applicability to fan interest in the game at large.

I date back to the late '30s. Like most young fans, my primary heroes were hitting and pitching stars. Eventually, I got keenly interested in the guys who made the great catches, turned DP's, exhibited great arms ... OR just made the game more interesting by making all that stuff look easy. Then too, there was that whole landscape for debating who was better ... or worse ... at this position or that one.

So, what do YOU want from a defense metric?

[scorekeeper]

Actually, I’m extremely interested in the thread’s topic. That’s what attracted me to it. I wish I could get others at the lower levels to be more “interested” in any metrics other than ERA, BA, SBPct, and FPct, but very few people trust the numbers, and for good reason.

If you would, please take a look at http://www.infosports.com/scorekeepe...special12a.pdf

Those are the defensive metrics I generate after ever game for our HS team. I don’t think you can get much more basic than that, but see how many teams you can find, even at the college level who put even that low level of defensive metrics out. IMHO, people at those levels are so caught up in offensive and pitching numbers, they have little time to study what I believe could help a team more than anything. But then again, I’m not a coach or manager, but just a numbers wannabe-wonk.

What I want in a defensive metric, is something more than FPct, but yet still simple enough for just about anyone to look at and get some valid idea as to how that player compares to others at his position. The trouble is, its really difficult to stay simple but get much validity.

[leewileyfan]

Then we are essentially on the same page: simple + credible.

For me, in devising a metric for defense, I had FIRST to determine what context of subject matter I wanted to make the basis of my focus. That came back in to parts:

1. historical
2. evolving

Then I wanted to consider data collection for equitable data inputs:

1. UNIFORMITY of availability for all seasons covered
2. SENSITIVITY to inherent [in-game] changes in rules, equipment, playing surfaces

I wanted a POSITION-CENTRIC approach, in which each position was appreciated as essential to the game, with positional "limitations" imposed only be the realities of game geometry and base advancement ... NOT by perceived images as some positions being essentially viewed as dumping grounds for aged or injured players.

Then I wanted a basis of comparison ... player defensive performance ... to WHAT. Strict numerical averages will vary from season to season and then too, between leagues. "Replacement" impressed me as being too low a comparison standard, conjuring [FOR ME]:

-images of the freshman class of New York Mets, a patchwork of aging vets and marginal young prospects predictably playing at around a .300 winning percentage

-images of "replacement" players recruited for the purpose of strike-breaking ... putting bodies on the field to encourage fans' bottoms occupying seats

I opted for a created Player X, always about 8-9 defense runs shy of whatever floating positional average prevailed in any given season. He is a good, solid investment that cannot be bought for a song, but rather at a relative bargain price in a pricey marketplace.

I recognize that, to a certain extent, pitching staff K's can restrict [or enhance] positional player opportunities to field batted balls in play; but I also recognized that the impact is diluted by being shared among positions.

Essentially, that's it. Pitcher handedness and ground ball/fly ball tendencies I consider "noise," for the purposes of my study. This may be a mortal sin to some; but MLB has a way of absorbing noise and tucking it away in an overwhelming reversion to distributive norms.

As a final note, anyone concerned with BIAS in any metric might just consider the game itself, notably at the MLB level of play. What is notable for fans going to the park is the disparity of stadium dimensions and conformations, despite more modern rules for design that strive to erase the more exaggerated cases. Baseball is loaded with rule, regulation, statistical data ... and glaring irregularities.

[scorekeeper]

Then we are essentially on the same page: simple + credible.

The reason it needs simplicity isn’t because stat geeks like Tango wouldn’t understand it, its that even though casual baseball fans, ex-players and players, and ex-coaches and coaches can and do recognize many of the metrics and buzzwords, most are woefully ignorant when you scratch very far beneath the surface.

As for the rest of what you said, while a lot of it shot well over this old head, I’m pretty sure I “get” the basis of what you’re looking for. In my looking at it, as soon as you go to things like “positional averages” or thinking about how he can be “acquired”, its completely lost any value as a metric for any but those in professional baseball. I’m not trying to say it isn’t elegant or doesn’t “work”, but rather that professional baseball is the only level of the game where players are commodities.

I’m gonna guess that what you’re doing has relevance and is useful for the level you find fascinating. I only wish it applied to all levels equally well.

[leewileyfan]

The reason it needs simplicity isn’t because stat geeks like Tango wouldn’t understand it, its that even though casual baseball fans, ex-players and players, and ex-coaches and coaches can and do recognize many of the metrics and buzzwords, most are woefully ignorant when you scratch very far beneath the surface.

Here you are taking the thread topic way off its intended search for information exchange. I have VERY DELIBERATELY abstained from mentioning any other metric. I have avoided names, personalities, approaches ... NOT out of any sense of intimidation ... but rather because many of them are widely published or widely familiar to many fans.

As far as being esoteric, not at all:This metric, if understood, is as simple and direct as it could possibly be. What makes it seem complex are the very core elements without which I would have trashed it entirely:

-built in sensitivity to game changes, such that as I did cold and basic data entries [letting the formulas do their thing] I got FEEDBACK, in the form of RATINGS clusters that were high ... higher than 1.000. It is NOT that > 1.000 was intolerable; but it told me that something[s] that held true for earlier ratings had changed.

-In every instance, this "shot across the bow" happened when a new glove design was introduced; or some element of play was altered; or some method [pitching] was outlawed; or something altered the playing surface [Astroturf, etc.].

:As for the rest of what you said, while a lot of it shot well over this old head, I’m pretty sure I “get” the basis of what you’re looking for. In my looking at it, as soon as you go to things like “positional averages” or thinking about how he can be “acquired”, its completely lost any value as a metric for any but those in professional baseball. I’m not trying to say it isn’t elegant or doesn’t “work”, but rather that professional baseball is the only level of the game where players are commodities.

This old head may well be several years older than your own. It is NOT I who is content with "positional advantages" or "disadvantages." The assigning of + credits to SS or 2B and - credits to 1B [for me] is an appalling concept that is a marked departure from appreciating the essence of the game. Also, speaking only for myself, ANY consideration of DH in discussions focusing on defensive value, is absurd. DH is non-defensive. It is a unique specialization, a contrivance, that has NO place in defensive evaluation.

When the DH was instituted, I favored it. My sole reason was that, for gifted hitters, who MIGHT still thrill fans with batting exploits, after perhaps, age, injury, joint fatigue or localized debilitation might have ended their playing days, perhaps prematurely. I'd be much happier, individually, if young pitchers were taught the importance of being full participants in games, where hitting was not considered some terrible burden not really a part of their participation.

:I’m gonna guess that what you’re doing has relevance and is useful for the level you find fascinating. I only wish it applied to all levels equally well.

You also mentioned "acquired" and "commodity" status of players, which I tried meticulously to explain in my aversion to REPLACEMENT as a player concept for comparison.

I explained the instances in which the very word evoked negative pictures in my mind; and I have more than once seen it [again, my opinion] as a tool best suited to the front office function of juggling and controlling payroll.

If I am mistaken, I apologize for mentioning this; but you do seem to have a way of introducing [imposing] your specific levels of interest here, while challenging my approach while, at the same time, being drawn to the topic. Much of what you have just attached to my approach is nothing more than the perceptions you have, which are out of context with what I am saying.

Bottom line: I am attempting to present a fan-friendly metric that allows defense to be seen as position-specific, within the broader scope of the entire game ... and to present the metric in a way that permits cross-generational player performance comparisons.

[brett]

That's precisely why I posed the question exactly as I did. To narrow the scope, my focus is on defensive play. To cite possible examples that MIGHT encourage responses:

1. Can a defense metric possibly be acceptable if it is constructed [and defined and explained] to allow comparisons among players like Rogers Hornsby, Bobby Doerr, Bill Wambssganss, Dustin Pedroia and Orlando Hudson. What if it attempts to compare OF, like Tris Speaker, Taylor Douthit, Dom DiMaggio, Riche Ashburn, Willie Mays and Jim Edmonds?

2. Can a metric earn any credibility if it exhibits Gary Sheffield as a versatile player for playing 3B and OF, without presuming incredible shortcomings at either/both?

3. How deep a gap is there between fielding prowess/weakness as REACHABLE from basic box score inputs, as opposed to reliance on pitcher handedness, ground-ball fly-ball tendencies, OR play-by-play data that is available for limited periods of time?

The question is put as a direct and sincere POINT of discussion. Sometimes, we [myself included] can accept or reject concepts or presentations NOT because of inherent flaws in whatever is under discussion, but more because we are talking ACROSS each other's points of view because we haven't defined our terms or the context of our focus.

No trick questions here. Just what do you, individually, seek [expect, want] from a metric presented to represent a model for measuring defense, historically and position-specific?

A real problem with historical defensive stats is that the original stats of putouts, assists, errors, double plays, do not ENTAIL what a player does defensively, so we are left extrapolating or correlating defensive performance to those stats. What if way back when, they had decided rather than recording errors, they would record "supurb plays" and give a player a hit when he was "robbed" of a hit? There will always be issues with subjectivity in scoring.

Aside from those 4 stats, an outfielder can a) cut of a potential double, b) reduce runner advances and I think these are two very big issues which may not correlate to the traditional stats. I have speculated for example that Joe Dimaggio comes out not quite top tier defensively in some metrics because he had such a huge outfield to cover and in large part lost putouts in exchange for cutting off doubles and triples in cavernous YS. We also have heard the old argument that the best outfielders lose some assists because they don't get challenged. I think that perhaps when an outfielder has a high ratio of putouts to opposition doubles, that might be a better stat because if you have say 400 putouts and the team allows 300 doubles some of the doubles may have been the result of aggressiveness, while another guy has 360 putouts but his team only allows 260 doubles his putouts might have been done alongside cutting off more doubles. I also think that if we look at opposition runs versus runs expected (via a sabermetric) we might have an idea of how outfielders hold down the running game.

[leewileyfan]

I agree with many of the points you have raised as to nuances of interpreting MLB defense. I am fortunate enough [old enough might be more exact] to have seen all three DiMaggio brother play the outfield. Joe and Dom at Yankee Stadium and Vince at the Polo Grounds.

What my metric attempts to do is to make reasonable assumptions about the in-play challenges at each position; then to model degrees of difficulty for the various plays to be expected at each position. After so many years of watching baseball, and having enjoyed many years playing it in tha amateur ranks, I did start studying various sabermetric essays, treatises, articles and books ... coming up with an unsatisfied feeling that the NUMBERS were at the heart of the studies and offerings ... but that the conclusions left something to be desired.

Then too, considering the WHAT IF of imagining transcribed broadcast play-by-play accounts of every inning of every game ever played readily at our disposal ... how much faith might we be able to put into that miraculous windfall? We know, from their own testimony, that several broadcasters, at the urgings of station/network mangers, "hyped" the dry and meticulous utterings of the teletype to "enhance" [put some life] into the games. Thus a circus catch bouncing off the wall by one of the Waner brothers might actually have bee a more-or-less routine play in actuality.

Even those metricians who are powerful or wealthy enough to have "spotters" at games across the country have complained that these reports lack consistency among interpretations of the kinds of batted balls witnessed.

I came to the conclusion that the box score is tangible. It gives us PO, A, DP, E, PB, CS, StB pretty much in dry, technical precision; and for the data most relevant to the majority of positions, the data exists for at least my time frame [1901-now]. I am convinced that if one devises a metric that has considered the nature of position play [at the time it is played], one can make a construct, without too much fear of abusing the facts.

So, a Joe DiMaggio may not get a number of PO due to the expanse of Yankee Stadium; but if one observes "others" making them [like Dom, or Wally Judnich], while other makes fewer plays, one can come away with the conclusion that Joe DiMaggio, over the course of his career, was not the number 1 defensive center fielder in the American League. Indeed, Dom was better; and a solid case can be made for others, in turn, like Mike Kreevich.

Am I saying either of these was better all-around? NO. However, if I am studing defense, then yes, others played center field better defensively than Joe D.

I have calculated 111 seasons of play; then reviewed them over again, satisfied that I have something very credible, even if it appears to come in a plain brown bag.

By my metric, an OF is not punished for assists not made. He just doesn't benefit from the 0 activity in that subset. More important, to me, is crediting those OF who do record assists. I am NOT inclined to credit a guy with 18 assists from RF with a lame arm, presuming runniers are going on him. To the contrary, I want to know how many he made pay a price for running on him.

Case in point: Mel Ott, under estimated on another thread because of the Polo Grounds dimensions. The expanse of the PG beyond the squinched corners is ignored, paying no heed to PO recorded by Ott in a "bandbox." Conversely, his 26 assists in an early season make it no surprise that a few years laters he was moved to 3B, where he performed solidly at about the MLB average for 3B.

I believe the routine plays and the gems all factor in with a metric custom-tailored for each position. Thanks for responding.

[scorekeeper]

Here you are taking the thread topic way off its intended search for information exchange. I have VERY DELIBERATELY abstained from mentioning any other metric. I have avoided names, personalities, approaches ... NOT out of any sense of intimidation ... but rather because many of them are widely published or widely familiar to many fans.

I’m not sure what you read into what I said, but obviously it wasn’t what I intended. If you think I purposely hijacked your thread, you’re wrong.

This old head may well be several years older than your own. It is NOT I who is content with "positional advantages" or "disadvantages." The assigning of + credits to SS or 2B and - credits to 1B [for me] is an appalling concept that is a marked departure from appreciating the essence of the game. Also, speaking only for myself, ANY consideration of DH in discussions focusing on defensive value, is absurd. DH is non-defensive. It is a unique specialization, a contrivance, that has NO place in defensive evaluation.

When the DH was instituted, I favored it. My sole reason was that, for gifted hitters, who MIGHT still thrill fans with batting exploits, after perhaps, age, injury, joint fatigue or localized debilitation might have ended their playing days, perhaps prematurely. I'd be much happier, individually, if young pitchers were taught the importance of being full participants in games, where hitting was not considered some terrible burden not really a part of their participation.

I don’t think we need to have an “I’m older than you conversation because its irrelevant”. I didn’t say “your” metric considered those things! I was speaking generically. And I don’t know where the DH stuff came from. I surely didn’t mention it.

…If I am mistaken, I apologize for mentioning this; but you do seem to have a way of introducing [imposing] your specific levels of interest here, while challenging my approach while, at the same time, being drawn to the topic. Much of what you have just attached to my approach is nothing more than the perceptions you have, which are out of context with what I am saying.

Bottom line: I am attempting to present a fan-friendly metric that allows defense to be seen as position-specific, within the broader scope of the entire game ... and to present the metric in a way that permits cross-generational player performance comparisons.

No need to accept an apology because there’s nothing to apologize for, mistaken or not. We have different interests and goals, but that doesn’t make us enemies. You’re looking for a better widget for one level, and I’m simply trying to find widgets that can be used at another. Most people at my level don’t even do defensive metrics, and if they do, there’s no doubt that all but a very very few are riddled with things that make them invalid.

As for me “challenging” your approach, nothing could be further from the truth. I’ve “challenged” nothing because I didn’t get into it enough to even worry about it. Tell ya what though, you tell me what I need to do to generate your metric at my level, and then I’ll consider challenging it. Until then, it will remain something that only interests fans of level you’ve generated the metric for.

If I’ve taken your thread off topic, I apologize.

[leewileyfan]

I don’t think we need to have an “I’m older than you conversation because its irrelevant”. I didn’t say “your” metric considered those things! I was speaking generically. And I don’t know where the DH stuff came from. I surely didn’t mention it.

The allusion to age [my being older than you] was intended in jest, since you referred to "this ole head." The reference to DH was in response to your citing "positional values," which I though referred to some metrics that assign + values and - values to positions [including DH], which I simply do not accept.

:No need to accept an apology because there’s nothing to apologize for, mistaken or not. We have different interests and goals, but that doesn’t make us enemies.

If you can write that with a straight face, you are taking this entire exchange far more seriously than I am.

:As for me “challenging” your approach, nothing could be further from the truth. I’ve “challenged” nothing because I didn’t get into it enough to even worry about it. Tell ya what though, you tell me what I need to do to generate your metric at my level, and then I’ll consider challenging it. Until then, it will remain something that only interests fans of level you’ve generated the metric for.

The metric is focused on MLB because, in the context of this forum in which we exchange messages, MLB is the centerpiece of most fan interest, and most metrics are applied to the topic of baseball where data are most uniformly and broadly documented.

The metric COULD easily be applied to any level of baseball where defensive records are kept. Degrees of challenge and the varied athletic executions and positional demands are fairly universal in baseball.

[scorekeeper]

… The reference to DH was in response to your citing "positional values," which I though referred to some metrics that assign + values and - values to positions [including DH], which I simply do not accept.

You’re sticking me in with the people you’re used to dealing with, which is ok. But that’s just one of the many reasons I stay away from ML numbers. There’s such a dearth of historical data at the HS and below level, there’s really no way to compute values like that.

The metric is focused on MLB because, in the context of this forum in which we exchange messages, MLB is the centerpiece of most fan interest, and most metrics are applied to the topic of baseball where data are most uniformly and broadly documented.

I understand that, but I’d surely like to see the fan base broadened, is all I’m trying to do. You probably can’t believe the interest in such things at lower levels, but the basic understanding just isn’t there, along with the lack of data. I’m ASSUMING you don’t know much about the wave washing over the youth baseball world(HS and below), of the new scoring apps.

People are suddenly finding out that there are not just a few other metrics than ERA, BA, SBPct, and FPct, but dozens and dozens of them. For people like me, many of those metrics have always been around, but not many people have taken the time to develop a database to work from. Now it doesn’t take much time at all, and suddenly there’s a whole new world of things to glean information from.

The metric COULD easily be applied to any level of baseball where defensive records are kept. Degrees of challenge and the varied athletic executions and positional demands are fairly universal in baseball.

Well if you can tell me how to do that step by step so I can program my scoring program to do it, I promise I’ll do it, and I’ll do what little I can to spread the gospel. If you looked at that link I provided, you pretty much know what data I have. It may be in the wrong format to do what you need to get done, but I’m pretty fair at making that happen. I’ll PM you my e-mail address, so if you want to continue, we won’t have to bore everyone to death with the machinations of creating something from nothing.

[leewileyfan]

In your fielding records link you did not include innings played for each player at each position. If a player, for example, had 220 innings at SS and 40 or so miscellaneous innings at numerous other positions, I would focus on the primary position, considering scattered "other" innings as not meaningful. The metric needs a denominator.

If "games" were the only denominator available in a limited records-keeping data base, that would do ... although games started and games finished at the position might tighten the calculations a bit. Innings is best.

Right now, I am very involved in editing/updating/polishing my manuscript while trying to get feedback on my metric outline on these boards. Beyond that, I hope to contact [or be put in contact with] some folks with graphic arts experience, so that the final product can be meaningfully and attractively illustrated.

If you provide innings for the data already linked [updated to incorporate innings in one document], I'll try to calculate each player at his primary position[s] and send results back to you. It may take some time [weeks]; but I could work it out and provide explanations.

Meanwhile, any feedback her or in my metric's primary thread [Evolving Defense] on this board, or player comp requests or straight-laced evaluation requests are welcome on either thread.