John Ward

[SABR Matt]

He never gets talked about at Fever.

I'd love to hear some commentary on Ward. Among the better pitchers, fielders, and hitters of his time...truly a master of many trades...but because he did all of those things some and none of those things exclusively...I never see him considered for HOF status here...he's among the biggest snubs still unheralded in the bigs.

[538280]

Matt, John Ward is in the HOF.

[SABR Matt]

I didn't find him anywhere on the official HOF website. He's in?

Somehow when I went and recorded every HOF player as a Y in my HOF column on the Master rankings table...I missed him.

In any event...I still think he should be talked about more here...LOL Not many people can claim to be as versatile and useful in any role as Ward...the closest I can think of is Deion Sanders back when he was both a very good corner and a skilled offensive threat in football (not to mention a halfway decent hitter in the major leagues...though hardly eye popping)...

Put John Ward anywhere on the field and he's an all-star calibar player...tryly unique even in the era of player-managers, pitcher/outfielders, and true utility players.

[538280]

Ward is talked about sometimes, Chisox stumps for him quite a bit, and I once talked about him when someone here tried to suggest he didn't belong in the HOF. I shot down that proposition extremely swifly and with a passion. Anyone who doesn't think Ward is deserving of his plaque doesn't get the true picture of the man (pitcher, shortstop, executive, pioneer). It could be argued no one had more influence on the game of baseball than Ward.

BTW, to prove he is in the HOF, here's his BBRef Page:

http://www.baseball-reference.com/w/wardjo01.shtml

[Gee Walker]

John Montgomery Ward is often listed as Monte Ward. That may be who you are missing.

[leecemark]

--He is listed as Monte Ward on the Hall of Fame Website. He has also been elected to the BBF Hall of Fame, although I think it was as a contributor rather than a player. IMO he comes up a little short as either a pitcher or player, but the combination is unique and deserving.

[SABR Matt]

AHH!! I wondered who the heck Monte Ward was...LOL

Now I feel stupid.

Anyway...yeah...as a pitcher he's not a HOFer...as a SS, he's not quite a HOFer...but the fact that you could say "he's close" when looking at either of those sides of his game, and he's got BOTH...that's a HOFer.

[jalbright]

We had a thread in June or July of last year about him in the HOF section debating whether he should go in as a player or contributor. I thought he should make it as a player (i.e. he didn't need his contributions outside the field of play to be HOF worthy), and wrote the following on his behalf:

His career divides into three parts, which makes it hard to get a handle on him. First he was a pitcher, and, according to baseball-reference.com, the most similar guy to him in that capacity is HOFer Addie Joss. He hurt his arm, so he moved to shortstop, and the most similar player to him there is a darned good, though not great, player in Bobby Lowe. Put the value of Bobby Lowe on top of Addie Joss, though, and that is definitely a HOF quality player. Eventually, he became a powerful force in the business of baseball as well. So if you somehow don't like him as a player, he's got to make it as a contributor.

In Black Ink, he has 8 as a hitter, 22 as a pitcher when 27 and 40 respectively are average for HOFers. In Gray Ink, he has 92 as a hitter and 142 as a pitcher when 144 and 185 respectively are average for HOFers. In HOF standards, he gets 28 as a hitter and 43 as a pitcher, when 50 is the mark for an average HOFer. He doesn't make the mark of an average HOFer in either capacity, but you combine those two unique pieces of his career, and it's hard to deny he was a great one. (Source for the above: baseball-reference.com). I don't have his exact career Win Shares total at hand, but it is over 400, which is a very high total, well within the realm of Hall of Famers.

[SABR Matt]

According to my research, he also rates as an above average fielder though not historically great...at short.

So...he could pitch like Addie Joss when he was young...hit like Bobby Lowe when he was old...and field like...say...Allan Trammell...when he was in between...that's truly astounding.

[538280]

According to my research, he also rates as an above average fielder though not historically great...at short.

So...he could pitch like Addie Joss when he was young...hit like Bobby Lowe when he was old...and field like...say...Allan Trammell...when he was in between...that's truly astounding.

DWS actually has him even better than you do then. They have him graded an A+ SS.

[SABR Matt]

I have problems...many problems...with James' defensive analysis. Not sure I buy that he was an A+ fielder...but it's entirely possible he's better than PCA thus far has seen for him.

[SABR Matt]

I have him as a career .292 PBA fielder (.270 being average...Ozzie Smith getting a .321)

[SABR Matt]

FWIW...

If you combine his pitching GI, fielding GI, and hitting GI (Offense -> 193.8, Pitching -> 142.3, Fielding -> 134.3), Ward winds up the third greatest SS of all time by the GI method...

Behind only Wagner and Robin Yount...just ahead of George Davis and Arky Vaughan.

[538280]

FWIW...

If you combine his pitching GI, fielding GI, and hitting GI (Offense -> 193.8, Pitching -> 142.3, Fielding -> 134.3), Ward winds up the third greatest SS of all time by the GI method...

Behind only Wagner and Robin Yount...just ahead of George Davis and Arky Vaughan.

I don't really think pitching contributions should count in my postion player rankings though....I've been thinking about that for some time. Maybe only count 1/4 of pitching contributions.

That's why I'm thinking more and more about moving Ruth down to #3, but if I did that that whole Fever house would jump on me scream, and complain (considering who I currently have at #3), so I'm holding off.

[SABR Matt]

So...a player's accomplishments as a pitcher don't have value? It's not important to you when a player displays skill in unusual ways?

BTW even if I drop the pitching GI from Ruth's career, he's still #1...though the gap between him and Cobb shrinks dramatically.

[baseballPAP]

I have him as a career .292 PBA fielder (.270 being average...Ozzie Smith getting a .321)
Matt, am I correct in assuming that the difference between average and Ward includes a LOT more players than the difference between Ward and Smith? In other words, is it a bell curve or a straight line?

[538280]

So...a player's accomplishments as a pitcher don't have value? It's not important to you when a player displays skill in unusual ways?

BTW even if I drop the pitching GI from Ruth's career, he's still #1...though the gap between him and Cobb shrinks dramatically.

The value is important to me, but I'm just not sure if it should count on my top 25 shortstops of all time or on my top 100 position players of all time. If we're counting pitching for our position player lists, couldn't we really say Walter Johnson deserves to rate in the top 10? (sum of pitching, hitting, fielding).

I just think pitching and fielding/hitting should be separate. I don't think it's valid to rate pitchers and postion players together.

[SABR Matt]

Matt, am I correct in assuming that the difference between average and Ward includes a LOT more players than the difference between Ward and Smith? In other words, is it a bell curve or a straight line?

It's more like a bell curve...the PBA statistic is calculated by placing a player's z-score for win scoring rate on the batting average curve. In other words...since the standard distribution of batting averages for players with 100 PA is .037...a player with a .307 PBA is 1 standard deviation better in win-scoring rate than one with a .270 (the average). There are in fact many more players between .270 and .292 than there are between .292 and .321.

[SABR Matt]

The value is important to me, but I'm just not sure if it should count on my top 25 shortstops of all time or on my top 100 position players of all time. If we're counting pitching for our position player lists, couldn't we really say Walter Johnson deserves to rate in the top 10? (sum of pitching, hitting, fielding).

I just think pitching and fielding/hitting should be separate. I don't think it's valid to rate pitchers and postion players together.

Well I have Walter Johnson in the low 20s I believe among the great PLAYERS of all time...including position players...this is because I include all elements...pitching, fielding, hitting...when I rank the entire class of players, because I think it's not correct that people assume pitchers represent 70% of the run prevention effort (it's closer to 55% by PCA). But I suppose I understand people choosing to separate the two groups...

I personally think if you're gonna separate along an axis like that...then you shouldn't count fielding...because fielding and pitching are related...not fielding and hitting.

And I personally think that not ranking pitchers is just a long-standing cop-out among historians who didn't want to touch the sticky question of how important pitching and fielder were compared to hitting in determining value.

[Chisox]

FWIW...

If you combine his pitching GI, fielding GI, and hitting GI (Offense -> 193.8, Pitching -> 142.3, Fielding -> 134.3), Ward winds up the third greatest SS of all time by the GI method...

Behind only Wagner and Robin Yount...just ahead of George Davis and Arky Vaughan.

Actually, I have him around #30 overall; around A-Territory and just ahead of Ripken and Yount, which would place him as the 2nd greatest SS of all-time on my list. I'd say that's pretty consistent.

[Chisox]

The value is important to me, but I'm just not sure if it should count on my top 25 shortstops of all time or on my top 100 position players of all time. If we're counting pitching for our position player lists, couldn't we really say Walter Johnson deserves to rate in the top 10? (sum of pitching, hitting, fielding).

I just think pitching and fielding/hitting should be separate. I don't think it's valid to rate pitchers and postion players together.

I agree with you, but I make three lists: position players, pitchers, overall.
Ward is in my top 25 SS, I think in my 50 P (used to be 30 or so, not sure where now), and around 30 overall, not sure exactly where.

[SABR Matt]

If you take out the pitching, Ward is 17th among SSs by the GI method right ahead of Jennings and behind Glasscock.

[Chisox]

If you take out the pitching, Ward is 17th among SSs by the GI method right ahead of Jennings and behind Glasscock.
I've got him in the low 20s, but that's considering the era (talent level) he played in. Does that system adjust for the time difference?

[SABR Matt]

No...PCA/GI makes no concrete era-adjustment, although it does account for the standard deviation of win rates in a league...it's not a full quality adjustment since league quality isn't fully captured in standard error estimates.

[Chisox]

No...PCA/GI makes no concrete era-adjustment, although it does account for the standard deviation of win rates in a league...it's not a full quality adjustment since league quality isn't fully captured in standard error estimates.
O.K. I assume that if era adjustments would be made, it would come out to be low 20s, so I feel pretty good in that assessment.

[2Chance]

Found the link that Jim was referring to.
John Ward - Player or Contributor? (http://baseball-fever.com/showthread.php?t=30242)
We had a pretty good discussion about him then.

[538280]

No...PCA/GI makes no concrete era-adjustment, although it does account for the standard deviation of win rates in a league...it's not a full quality adjustment since league quality isn't fully captured in standard error estimates.

Really? Well that changes my whole thought the results you give...I was under the impression they were full LQ adjusted. Is that why you have Cobb #2, because frankly I don't see that if you're making a good LQ adjustment.

[SABR Matt]

Well let's put it this way, Chris...

PCA does fully adjust for the differences in the win-scoring distributions of the players, which is all Schell did to create HIS era adjustment really...but I am of the impression that that is not really a complete measure of league quality.

However, I am also of the impression that a full LQ adjustment would be very small. If our AAA leagues register as a 90 on the league difficulty scale as Davenport claims (they are 90% as talented as the MLB), then I find it very hard to believe any major league after 1900 would fall below that range significantly with the possible exception of the 1900s NL and the 1930s NL. And if our modern Major League is the 100 figure then the 1970s through the mid 80s would probably exceed 100 but only slightly.

[jalbright]

Well let's put it this way, Chris...

PCA does fully adjust for the differences in the win-scoring distributions of the players, which is all Schell did to create HIS era adjustment really...but I am of the impression that that is not really a complete measure of league quality.

However, I am also of the impression that a full LQ adjustment would be very small. If our AAA leagues register as a 90 on the league difficulty scale as Davenport claims (they are 90% as talented as the MLB), then I find it very hard to believe any major league after 1900 would fall below that range significantly with the possible exception of the 1900s NL and the 1930s NL. And if our modern Major League is the 100 figure then the 1970s through the mid 80s would probably exceed 100 but only slightly.

I agree with your conclusion, as the History section's Collins versus Morgan debate would attest. I haven't read Schell's work (at least not yet), but I'd like you to address a few questions I have about the approach:

1) How do you and/or Schell deal with the issue of using Standard Deviations on a talent base which is not normally distributed? When we're talking major league caliber players, we're at the far end of a normal distribution. I would guess there's mathematical ways to deal with this, but do either of you do so? If you do, doesn't it involve more than a little calculus?

2) How would this approach handle a scenario like the following: Take a sim of any league, and give each team one of the all-time greats in his prime to add to its roster. I think it's clear that doing so would raise the quality of play. But shouldn't the extreme performances of these greats also raise the standard deviation of performances--and isn't that interpreted in this approach as evidence of a lower quality of play?

3) Especially given the distribution of talent, shouldn't variations in the number of regular jobs available make it difficult to compare one season to another? Given the statistical concept of seeding from the top, isn't the standard deviation approach effectively assuming that if the SDs are smaller for the greatest players, that the middle is therefore higher? But is that an accurate assumption when the number of jobs increases through expansion? I'd think a countervailing issue is drawing talent from a larger pool--but how do we balance these issues?

Jim Albright

[SABR Matt]

All excellent questions Jim.

The short answer to 1 is:

I don't.

And neither does Schell.

I realize that's a significant concern...the player talent distribution is pyramidal rather than normal and that causes problems to be sure. There is a way to handle the kind of distribution exhibited in the major leagues...but as of this moment I do not fully understand how to apply it. Someone did an analysis...and I've lost the link to the paper written about it...of HR frequency and essentially concluded that peak HR performances haven't really changed much in recent years and that most of the difference in HR rate per AB can be explained by a cluster of players having career years at the same time. But the reason I bring that up here is that that paper references a distribution which is used to measure greatness in human achievement in ANY field which is really what you need to use to measure the distribution of player talent in baseball.

2) The thought had certainly occured to me. One possibility would be to remove the greatest players in the sample and look at the CENTER of the win-scoring distribution only (those players falling within the inter-quartile range perhaps...or the inner 90 percent...or something like that)...and that is something I've considered doing.

3) Interesting point.

I don't know if you saw my research in the area of LQ adjustments wherein I looked at the distribution of game scores in each league and season...specifically and the skew and kurtosis of those game scores...working under the assumption that a more balanced league would include more balanced games..IOW if the league is deep...there won't be patsy teams who you can wallop 22-0 on a regular basis...and more to the point...if baseball is played ocrrectly...it's exceedingly difficult no matter who you're playing to score bunches of runs...many many mistakes have to be made to score a whole heck of a lot of runs. That's why softball leagues have average scores of like 29-18...that's why run scoring was so whacky in the 1870s. I found that the NL in the 00s and 30s was severely weak but the same was not true of the AL in those time periods...at least if I am to believe that the skew of the RS distribution can tell us something about league quality.

I also found that the widening talent pool and expansion washed each other out in the early 60s and that the game got better...not worse...as expansion occured in the 60s/70s to the point where it reached a peak in toughness in the early-mid 80s.

[jalbright]

I'd like to take the discussion of league quality adjustments and unify it with a past thread on that issue. The thread in question is here: http://baseball-fever.com/showthread.php?t=34915&page=6 I resurrected it at post 128 and included a link to the Collins/Morgan thread discussion of the issue as well.

Jim Albright