Lee's Caught-in-the-Webbing Line

[leewileyfan]

Hi. Although I have been a poster here for several months, in starting a new thread that invites questions and comments, I thought it would be a good idea to open the thread with some personal introduction and an outline of the thread's purpose.

My user name, Leewileyfan, comes not from some obscure ballplayer but from a husky-voiced blues, standards and jazz singer popular in the '30s and '40s. When musicians were asked about working with her, they generally agreed it was like having a special kind of instrumentalist in the band with them.

Having thus dated myself, I've been a Red Sox fan since 1939; so I am right at home with a degree of frustration .... but equally comfortable at being patient.

Although I am am at home with math and statistics, I do not refer to myself as a sabermetrician. That is a simple statement of fact, not a knock on sabermetrics, which I've studied for several years now.

My major interests lie in the historic aspects of the game, up through the present; and I've developed a deep and keen interest in defensive play, 1901-present. My metric permits equitable comparisons for position players across that 110 year span; and I hope some readers and posters here share that interest.

I've considered the "what ifs" of Gold Glove Awards existing across that span; and have come to evaluations that challenge some ridiculous awards made and some glaring biases in repeat awards.

I am now working on team defenses and have completed 1901 through 1985, along with study into the three principal game components and their relative contributions to team success [or failure].

So, the doors are open and we hope some folks here have some fungoes for us to field. If I don't have the answer, I'll try to get it or refer you to a site that might. If it will take a bit of work, I'll also let you know that but will get back to you as fast as I can.

So, welcome.

[jaxxr]

I would like to ask a general question about defensive evaluation/comparison.

How does one best account for the time era differences ?

Particularly for the lack of detailed box scores, equipment shortcomings and field conditions or turf, especially for infielders.

Honus Wagner has limited box score data available, Marty Marion never played on Astro turf, Louie Aparicio has limited TV film to analyze, while Jimmy Rollins has, probably every game on video.

I realize the negro leaguers are not in your specific player base, and perhaps that is reasonable, as Pop Lloyd is evaluated almost 100% via anecdote or observation.

Best of luck with the thread.

[leewileyfan]

Thanks for the question, jaxxr.

I first decided what I wanted to study and how I wanted to go about it. So I took the following steps:

1. Determine an historic framework during which the MLB game is very much as we know it today: rules, physical dimensions, deliveries, mound, etc. That gave me a start date of 1901, new century, new game - although 1902 would be technically more correct.

2. Limit data inputs to those as readily available for 1901 as they are for 2009. Thus, my starting headings were Assists, PO, DP, E, PB for every position.

3. Incorporating local ball fields into my walking-jogging routine and pacing off areas covered by each position. During this phase I jotted notes on playing conditions back to my grandfather's [catcher] recollections of game play; my father's [3b; P] @ 1911-1921; and my own [P], 1944-1954.

4. Treating each position as a separate universe, assigning degrees of difficulty to types of plays executed and the relative cost values of errors for each.

5. Recalling a column by an old scribe @ 1911 to follow the assists, particularly in the IF, creating individually weighted models [+/- value] adding up to a position model. The results were, naturally, varied as the positions.

6. Creating a conversion formula, so that player performance against model would be uniform and familiar. The formula converts ratings into [usually] a .900 - 1.000 range, similar in appearance to fielding %.

7. By trial and error, at first, creating a template set @ 1.000 rating at each position. Calculating player ratings against position templates allows ratings to be converted into runs. This also allows flexibility:

a. say a SS rates .937 and relative perfection = 1.000. However, most fans and evaluators do not compare individuals to such a model, preferring to compare to average, replacement]. I subjectively created a Player X, created by me out of observations for his generation. I envisioned him as a journeyman player, competent, tried, past his peak and around 8-10 DR below predominant performance numbers at his position. For these threads, I am using comps to average.

Our SS above @ .937 would be +0.78 DR > average, with overall average @ SS = .935 by my metric.

8. At the key to the answer to your question is the feedback built into my metric.

As playing conditions evolve [groundskeeping. glove design, ball liveliness, playing surfaces, domes], player ratings set for earlier periods of play start to rise toward 1.000 or above; and I have tried to respond to these signals by expanding template expectations. Thus, the metric adjusts to the time for each position [they do not all change instantly or in exactly like manner]. The flexibility of the templates, I believe, makes my metric uniquely suited to player comps among generations in the 110 year period studied.

As for Negro League players who never made it to the big leagues, I am inclined to accept the oral histories. I was lucky enough to have seen the battery of Satchel Paige and Josh Gibson at Dexter Park in Maspeth [Queens/Brooklyn border] around 1942. My Dad pointed out to me the [B]Krrraaack sound made when a Paige pitch hit the pocket. I do recall a basically side-arm whip motion, lots of strikeouts, and not many hard hit balls.

The only numerical evaluations I do for black ballplayers is for MLB games played.

Thanks again for the question. I hope you find this responsive.

[jaxxr]

I am really not sure exactly how the specific "templates" are created, but really see no reason for an elaborate or fully detailed explanation, as the purpose is fairly clear and reasonable.

I am still curious as to how a unique factor such as astro turf for a SS, for example, is accounted for. Somewhat of a ballpark factor, I suppose.

Is there a good way to evaluate the defense of a SS who may have played 60% of his games one season on a turf infield, in comparison to one who may have played 20% of his games on turf that same season ?

[leewileyfan]

I am really not sure exactly how the specific "templates" are created, but really see no reason for an elaborate or fully detailed explanation, as the purpose is fairly clear and reasonable.

I am still curious as to how a unique factor such as astro turf for a SS, for example, is accounted for. Somewhat of a ballpark factor, I suppose.

Excellent question and on a topic that could make for a book of its own. The way I treat playing surfaces is NOT to get sidetracked by particular ball park in which a player half his games at home.

The individual player and position models are basically a weighted series of + inputs and a - input for misplays, based on positional damage done by errors. The model is, therefore, a sum. The template is generally a considerably higher sum, established by progressively incremental bests at position, set to convert to 1.000.


Is there a good way to evaluate the defense of a SS who may have played 60% of his games one season on a turf infield, in comparison to one who may have played 20% of his games on turf that same season ?

In the next day or two, I'll try to cite such an example as the one you posed. Meanwhile, laying out some notes I'd already jotted down, here are some observations that are more general, but directly related to the question:

1. Using Houston as the Astroturf master model for artificial surfaces, between 1966 and the present, Houston players noted for defense have essentially been: C [like Bateman, Edwards and Ausmus]; 1b [Watson, May, Bagwell]; 3b [Doug Rader]; and OF [Wynn, Staub and Cedeno]. This suggests that up-the-middle, as 2b and SS, the artificial surface inhibits player range due to limited reaction time for ball in play.

2. When I list players in #1 above, I am not listing either formalized award winners or my own top-rated position players in a given season. I am listing guys in my top 3-5 at their position in any season. Tommy Helms [2b] and Roger Metzger [ss] were two player who excelled on the surface.

3. Comps for players playing % on turf and off turf relative to each other is not as simple a task as it may seem. This is because between 1969 and into the '80s, there was a craze of sorts among MLB franchises who converted to artificial surfaces, then converte back to grass. The craze appears to be a thing of the past.

4. The most avid proponents of artificial surface [Whitey Herzog] may have put up a defensive + for Ozzie Smith @ SS, since he played a big portion of his career on AT. They took us back to slap and run small ball, discounting power for speed [both on offense and defense]; with triples nudging HRs and high bouncing balls in the gap flying high above OF heads. This, plus bad hops on seams and hard surfaces, contributed to ending the craze.

5. The common denominator of horticultural natural surfaces is the human element of the head groundskeeper. Whether weather zone defined, humidity, pest control, maintenance costs, the human element can determine height-of-cut, coarseness, grain - becoming an invisible infielder at home.
The slowness or fastness of the natural surface can be tweaked a bit to comport with home team strengths and weaknesses on defense and offense.

Bottom line: C, the corners and the OF suffer less from AT in defensive execution; but 2b and SS are hindered by reduced reaction times.

I let the numbers for inputs determine the findings. As I see it, grass>artificial surface for best defensive play.

[jaxxr]

" 2b and SS are hindered by reduced reaction times. "

Artificial turf probably does that, however, the "true-ness" of the trajectory, the lessening or near elimination of bad hops, Etc., might offset some part, or all of the reaction time lost, from the presumed increased velocity of every grounder.
In addition, the artificial surfaces are generally more consistent throughout, in relation to natural grass ones.

Perhaps my perception that a SS, or 2B, who had a majority of his games on AT, should be evaluated differently, is overly nit-picking ?
Ozzie Smith, for example may have had little or no difference in his stats, on AT or natural surfaces ?

[leewileyfan]

" 2b and SS are hindered by reduced reaction times. "

Examination of defenses of three teams transitioning from grass to AT or vice versa for three years before and after, may leave me corrected on the quote above.

In the first year of change @ 1b, 2b, 3b and SS, there does appear to be a drop, across the board, in assists; but this may be more a matter of adjustment than truly measurable & fixed decline in defensive opportunities.

Thereafter, the drop is more persistent at the corners, with 2b and SS faring better in ensuing years. I guess the bunt having beeen largely reduced as a winning strategy for the offense, 3b and 1b would experience that effect. Then too, the corners would probably feel the reaction time affect as well: hard-hit balls to a reduced distance from the plate.

A defensive player I've always rated well rises like cream in this regard: Doug Rader at 3b.


Artificial turf probably does that, however, the "true-ness" of the trajectory, the lessening or near elimination of bad hops, Etc., might offset some part, or all of the reaction time lost, from the presumed increased velocity of every grounder.

That "trueness" is part of the decline in popularity of AT, largely because it does not last as long as once was believed; and when an investment like that wears, so does "trueness" and reliability.


In addition, the artificial surfaces are generally more consistent throughout, in relation to natural grass ones.

Yes. And no doubt enticing for savvy head groundsmen to exploit for home team advantage, no matter how small.


Perhaps my perception that a SS, or 2B, who had a majority of his games on AT, should be evaluated differently, is overly nit-picking ?
Ozzie Smith, for example may have had little or no difference in his stats, on AT or natural surfaces ?

Not at all. Exploration of any changes, like playing surfaces or indoor-outdoor park design are critical. However, my metric does have formulaic effects on player rating outcomes which prompt me to review templates when spikes occur.

I am thinking it's an expensive wash [AT] that is falling out of favor. Triples hitters have enjoyed the fruits of the surfaces, as some slap hitters have as well. However, no essential or contention benefit of lasting value has ever really been a payoff for AT surfaces.

Ozzie Smith would have excelled anywhere. I don't believe grass would have affected his defensive scores much, either as + or -.

[RuthMayBond]

Bottom line: C, the corners and the OF suffer less from AT in defensive execution; but 2b and SS are hindered by reduced reaction times.

And if the ball gets to (not past) them, they will have MORE time to throw out any runners.
And bunts which might have died for hits on natural grass may get the batter thrown out.

[leewileyfan]

And if the ball gets to (not past) them, they will have MORE time to throw out any runners.

I am supposing here that "them" refers to players on artificial surfaces. Reviewing my notes on ball coefficients, bat speeds and pitch speeds at contact, and diamond batted ball trajectories slightly into the hole between 3b and SS, I get .25 seconds added time for the SS [in this example] > reaction time in the live ball era than the slower batted ball movement in the dead ball era.

On artificial turf, I estimate a reaction time for the same batted ball to be @ .10 seconds less; so the reaction time between dead ball and modern era on artificial turf is 1.65 seconds [dead ball]; 1.35 [live ball, grasses]; and 1.20 [live ball, artificial turf].

This also takes into consideration the fact that most modern grass surfaces are hardy blends that thrive on closer cuttings than bent and blue grasses, making for a firmer, rich, but faster "carpet."


And bunts which might have died for hits on natural grass may get the batter thrown out.

Indeed. However, the bunt is a dying skill; and most metrics weigh it as a losing proposition in the attempt. On turf, the negative possibilities are magnified. To that extent, your argument, while physically valid, is diluted by the practical odds against success for the strategy.

I doubt many, if any, 3B suffered because of poor execution on bunt attempts on artificial turf.

[leewileyfan]

Duplicated post deleted by poster.

[leewileyfan]

How does one best account for the time era differences ? ...... Particularly for the lack of detailed box scores, equipment shortcomings and field conditions or turf, especially for infielders.

Honus Wagner has limited box score data available, Marty Marion never played on Astro turf, Louie Aparicio has limited TV film to analyze, while Jimmy Rollins has, probably every game on video.

While I responded pretty fully to the first part of this question from jaxxr, I didn't pursue in much detail players mentioned above.

Of the SS named in the quoted bit above, I have them rated, career, at the following DR saved [+] lost [-]:

Marty Marion +158
Louis Aparicio +137
Jimmy Rollins -10
Honus Wagner +30

In another portion of the thread, jaxxr mentioned Ozzie Smith, whom I have @ +280 DR.

[leewileyfan]

Just for what it's worth: I'm incorporating [or trying to] a catcher defensive breakdown of skill sets, taking on separately the elusive game calling skill - as isolated as possible from base runner management; plate blocking and protection, and overall athleticism.

This is not an evaluation of greatest catcher ever, since that would incorporate both offense and defense. To satisfy my own curiosity, I am trying to get a handle on the best game-callers, the guys pitchers love to work with, or whom they are forced to respect because of their savvy with staff weaknesses and strengths.

First observation: Great game calling evolved with greater consistency in the N.L over the A.L. during the dead-ball era, which I consider 1901-1920. The "new" ball was introduced almost a decade before 1920; but it was never fully exploited until 1921.

Second observation: The very best game caller in the game [1901-1920] seems to have been George Gibson of the Pirates, starting with 1908. He was among the elite handlers from 1908 through 1914.

Third observation: The best A.L. game caller of the 1901-1920 era was Ray Schalk. His excellence was consistently exhibited starting in 1915.

Edit: Added Note Johnny Kling was no doubt a superb catcher and handler of pitchers. His best streak was 1904 through 1907, after which he did more actively pursue business investments and competitive billiards. He did take a leave of abesence a couple of years later and was not at his former level.

The period covered is 1901-2009.

Edit: Additional notes:

For anyone who may be interested, I'll be dividing catchers into 15 season "generation:"

1901-1915
1916-1930
1931-1945
1946-1960
1961-1975
1976-1990
1991-2005
2006-through present

Of course any other questions, not relayed to catchers only - like questions on defender comparisons, etc. are most welcome.

[leewileyfan]

In the post immediately above this one, I outlined a plan to segregate "generations" of great catcher game-callers in blocks of 15 years. With all the overlaps, a 25 - 30 year span seems better suited; so here goes [with rankings in approxiamte order, by generation [B]NOT[/B adjusted for playing time]:

1901-1930

Frank Snyder
Ray Schalk
Bob O'Farrell
Walter Schmidt
George Gibson
Bill Killefer
Cy Perkins
Johnny Kling

1931-1960:

Buddy Rosar
Ray Berres
Gabby Hartnett
Jim Hegan
Bill Dickey
Mickey Cochrane
Muddy Ruel
Gene Desautels
Ray Mueller
Sherm Lollar
Harry Danning
Paul Richards
Frank Pytlak
Del Crandall
Sherm Lollar
Rick Ferrell
Rollie Hemsley
Roy Campanella
Johnny Edwards
Lou Berberet
Al Lopez
Hank Foiles
Yogi Berra
Clay Dalrymple
Wes Westrum
Johnny Roseboro
Gus Triandos
Bob Swift

1960 - :

Jim Sundberg
Elston Howard
Gary Carter
Johnny Bench
Bill Freehan
Rick Dempsey
Frank House
Charles Johnson
Randy Hundley
Lance Parrish