This piece is offered as follow-up to my previous article and commentary on the nature and quality of the work offered by Rob Lifson and REA on the 1938 Lou Gehrig New York Yankees Road jersey that I evaluated. Today I want to focus on process and product without the unnecessary background noise that comes with REA’s involvement in the research and evaluation of this jersey. I am going to take you through the essential relationship that exists between exclusionary and imagery analysis as it pertains to definitive photographic attribution (“Photo Matching”). “Photo Matching” is a phrase I have come to detest and I hope by the end of this article, you will see why. Additionally, I also want you use this jersey as an instructional vehicle on process, and by taking you through this process in what I feel is a responsible manner, also help you to see this product (1938 Lou Gehrig New York Yankees Road Jersey) in a more informed light; nothing more and nothing less.

SUBJECT: 1938 Lou Gehrig New York Yankees Road Jersey Imagery & Exclusionary Analysis of PLATES A-I

My previous related research efforts on Yankee uniforms from this time frame indicate that the likely annual population of uniforms potentially available for each player would in the range of 3-5. This is based on an ordering/inventory pattern of:

1-carried over from previous season

2-ordered for the season

1-2 for World Series

Notice I did not say 5 uniforms were ordered for each player each season. What I am saying is that you have to have some idea of how many uniforms are in play before you start playing around with your “photo matching”. It has been my experience that most folks find something specific they are looking for and stop their research when they have convinced themselves they have found it. If this is your “photo matching process”, then as a trained analyst, I have problems with this and I think you will see why in short order.

In looking at various images, I have identified what I feel are at least three (3) separate Lou Gehrig road jerseys from 1938. This is based off a combination of characteristics including:

-Sleeve length

-Font spacing

-Lettering alignment with the button line

-Presence of the World’s Fair Patch

It is important to understand upfront that that sleeve length is only a distinctive characteristic under certain circumstances. Sleeve length can be changed or altered once a jersey has been produced; but the sleeve will only get shorter and not longer. In looking at the aspect of sleeve length and other characteristics over time, it is possible to use it as a way to distinguish one jersey from another, thus permitting the establishment of a known or likely population of jerseys for a particular player in a particular season. This becomes important since when you have a fairly well defined and accountable population of jerseys for a particular player in a particular year, claims of direct photographic attribution (“photo match”) can enhanced or refuted by being able to identify, account for and exclude other possibilities. This is referred to as exclusionary analysis. The logic trail for the relationship between exclusionary and imagery analysis as it relates to direct photographic attribution (“photo matching”) runs along these generic lines. By way of illustrative example, let’s say that:

-Elmer Fudd likely had 3-5 jerseys available for wear throughout the entirety of the 1956 season based on what we know or believe about ordering/inventory patterns.

-Through imagery analysis, we have identified and accounted for at least 3 separate jerseys worn by Elmer Fudd in 1956.

-Imagery analysis has permitted us to exclude 2 of the 3 jersey worn by Fudd from being the one offered for consideration.

-The one remaining jersey is consistent with images available of Elmer Fudd from 1956.

-While the probability might be high that the offered jersey is the one Fudd is wearing in photographs from 1956, no definitive claim of direct photographic attribution is made because there are other possibilities that have not been accounted for an excluded.

I take the time to lay all of this out as a both an educational and cautionary message to this hobby/industry because before someone runs off and claims a “photo match”, it is essential that they first identify what they know about the known or likely population and then physically demonstrate how these various jerseys have been identified and distinguished from one another, and then “excluded” from consideration. When this is not done, any claims of direct photographic attribution (“photo matching”) are not objectively or statistically defendable.

What follows is a combination of imagery and exclusionary analysis as it relates to the 1938 Lou Gehrig New York Yankees road jersey being offered by Rob Lifson and Robert Edwards Auction (REA). I think it very important to note that this work includes analysis of images I did not have or had not seen at the time my original opinion was provided. The overwhelming majority were located and provided to me by Mr. Peter Nash and for that I am extremely grateful.

Jersey #1: This jersey is identified as being separate from Jersey #2 based on the lettering alignment with the button line. Jersey #1 also cannot be Jersey #3 because of the sleeve length. Although sleeves can be shortened, the images of Jersey #1 predate those of Jersey #3. The jerseys shown in PLATES A & B are considered the same jersey at this time as I can find no readily apparent disqualifiers. It should be noted that while PLATE B bears a date of 9/26/1938, this is not the date of the photograph. Rather it is date the image was provided for use with the caption and credits. I confirmed this by looking at the 1938 New York Yankees schedule for the period in question:

September 7-8: Boston

September 9-11: Washington

September 13: Cleveland

September 15-17: Detroit

September 18-19: St. Louis

September 20-22: Chicago

September 24-26: Home against Boston

In my opinion, it is not unreasonable to presume this image is from one of the previous road games (although it could be from any time prior to the 23rd of September as well). This jersey is also consistent with the image from April 16th of 1938, so I do not disqualify them from being the same jersey. You will also notice that this jersey does not feature the World Fair Patch worn by the New York Yankees, New York Giants, and Brooklyn Dodgers in 1938.

Jersey #2: This jersey is excluded from being the offered jersey based on the button alignment. Although this observable characteristic may appear to artificially vary in images based on angle of photograph and fabric folds/body position, the image angle and rather upright body position of Gehrig in the photograph in question seem to negate any appreciable impact those factors might have on this characteristic. Since Jersey #2 is neither Jersey #1 nor Jersey #3 based on sleeve length/lettering alignment, it will not be discussed further in any material manner. (PLATE C)

Jersey #3: Jersey #3 refers to those images that I either believe to be or can be shown to be from the 1938 World Series. I attribute the 1938 team photo to the World Series based on the fact that the Yankees are appearing in road uniforms and the 1938 World Series opened in Chicago. This would not be an uncommon event for a team to pose for such a picture. The composite details (both physical and those affected by angle of image) of the jersey identified as Jersey # 3 are such that I cannot exclude them from being the same jersey. Additional images of the team photo from different points of reference confirm that the sleeves Gehrig is wearing in this picture are elbow length, which is not consistent with Jersey #1 or Jersey #2. Longer sleeve length appears consistent with the other images that can be placed to the 1938 World Series. (PLATES D thru G)

In addition, and upon closer comparison, Jersey #3 is excluded from being the offered jersey based on the spacing between the letters “E” & “W” in NEW YORK. While the general alignment with the button line is consistent, it does appear to feature a spacing between the letters “E” & “W” in NEW YORK that is clearly less than one button width. This relies on the assumption that the jersey in the team photographs is the same one as the ones from the World Series.(PLATES H, I)

I also exclude Jersey #3 or any other jersey with long sleeves and the World’s Fair Patch from being the same jersey as Jersey #1 or the offered jersey. The fabric cuts that I found on the offered jersey are in a location that would limit the application of that patch to have being placed on a jersey of longer sleeve length since the World’s Fair Patch was sewn to the bottom of the sleeve.

Based on the fact that I have not been able to account for other possible jerseys, and then exclude them, I offer no definitive claim of photographic attribution that the offered jersey is the same as Jersey #1. What I feel this imagery analysis in the PLATES does show is that:

-Both the offered jersey and Jersey #1 share a common sleeve length.

-Sleeve length is a discriminator based on the chronology of the photographs.

-Both the offered jersey and Jersey #1 can be seen without the 1939 World’s Fair Patch.

-Both the offered jersey and Jersey #1 share common and consistent font alignment with respect to the button line and consistent spacing between the letters “E” & “W.”

What a combination of imagery and exclusionary analysis indicates to me is that the offered jersey could be Jersey #1. In addition, since we now have images of Gehrig wearing a jersey without the World’s Fair Patch, it is quite conceivable that no patch was ever applied to the jersey and that the location of the fabric cuts is a coincidence. This would not have any appreciable impact on my overall opinion of the jersey and would only affect the overall grade assigned MEARS by +.5 since the A9 opinion provided by Dave Bushing did not deduct points for this. In my opinion (A5), I deducted -1 for the patch. Thus a revised grade of A6 from me would leave the aggregate grade at A7.5 (average A6 & A9) as opposed to A7.

This article has been about two separate but related issues; process and product. My hope is that the hobby/industry now has a better understanding of the essential relationship between imagery and exclusionary analysis when it comes to supporting a position of direct photographic attribution (“photo matching”). When the collecting community or auction houses conduct or subsidize substandard work, process, and analysis, then we should not be surprised when it results in “photo match” claims that are not objectively or statistically defendable. Please know for my work in this area, I leverage no special tools or technology. If you are looking at the images on a computer screen and care to invest about $1.00 in a clear ruler you can lay on the screen, then you are armed with the same tools as me. This only leaves differences in experience, training, and process.

This article has also been about a product or the ability for the collector to see and know as much as possible about a particular jersey. In the case of this 1938 Lou Gehrig New York Yankees Road Jersey, I maintain my original position that “based on my physical examination of this jersey and supporting data, it is my opinion that this jersey posses all of the characteristics you would expect to find in a 1938 New York Yankees road uniform supplied by Spalding for use and wear by Lou Gehrig during this period”.

This is an opinion based on:

-Imagery analysis.

-Comparative trend analysis with other Lou Gehrig and period New York Yankee jerseys I have examined.

-Comparative fabric analysis with other period and year specific Spalding major league grade/quality uniform products.

-Exclusionary analysis to determine a likely population and offering data to support either the possibility of exclusion or inclusion with respect to the issue of direct photographic attribution.

-A physical examination of the jersey utilizing UV lighting, a light table, illuminated magnification, and the use of a digital microscope.

I also maintain my previous position of that “if you are serious buyer then take this jersey very seriously. Although there are some condition issues, it is one of only a handful of legitimate Lou Gehrig jerseys out there and the likely population of good Gehrig jerseys can only to get smaller. This population has been cut down significantly as I have evaluated three others that were found to be problematic for any number of reasons. If you are in the market, then buy the jersey for what it is based on what you can objectively see and know. I think you’ll be thrilled and lucky to have it as the display appeal is incredible and the conservation work in the sleeves is museum quality”.

As always, collect what you enjoy and enjoy what you collect.

Dave Grob

For questions or comments on this article, please feel free to drop me a line at DaveGrob1@aol.com.

This piece is offered as follow-up to my previous article and commentary on the nature and quality of the work offered by Rob Lifson and REA on the 1938 Lou Gehrig New York Yankees Road jersey that I evaluated. Today I want to focus on process and product without the unnecessary background noise that comes with REA’s involvement in the research and evaluation of this jersey. I am going to take you through the essential relationship that exists between exclusionary and imagery analysis as it pertains to definitive photographic attribution (“Photo Matching”). “Photo Matching” is a phrase I have come to detest and I hope by the end of this article, you will see why. Additionally, I also want you use this jersey as an instructional vehicle on process, and by taking you through this process in what I feel is a responsible manner, also help you to see this product (1938 Lou Gehrig New York Yankees Road Jersey) in a more informed light; nothing more and nothing less.

SUBJECT: 1938 Lou Gehrig New York Yankees Road Jersey Imagery & Exclusionary Analysis of PLATES A-I

My previous related research efforts on Yankee uniforms from this time frame indicate that the likely annual population of uniforms potentially available for each player would in the range of 3-5. This is based on an ordering/inventory pattern of:

1-carried over from previous season

2-ordered for the season

1-2 for World Series

Notice I did not say 5 uniforms were ordered for each player each season. What I am saying is that you have to have some idea of how many uniforms are in play before you start playing around with your “photo matching”. It has been my experience that most folks find something specific they are looking for and stop their research when they have convinced themselves they have found it. If this is your “photo matching process”, then as a trained analyst, I have problems with this and I think you will see why in short order.

In looking at various images, I have identified what I feel are at least three (3) separate Lou Gehrig road jerseys from 1938. This is based off a combination of characteristics including:

-Sleeve length

-Font spacing

-Lettering alignment with the button line

-Presence of the World’s Fair Patch

It is important to understand upfront that that sleeve length is only a distinctive characteristic under certain circumstances. Sleeve length can be changed or altered once a jersey has been produced; but the sleeve will only get shorter and not longer. In looking at the aspect of sleeve length and other characteristics over time, it is possible to use it as a way to distinguish one jersey from another, thus permitting the establishment of a known or likely population of jerseys for a particular player in a particular season. This becomes important since when you have a fairly well defined and accountable population of jerseys for a particular player in a particular year, claims of direct photographic attribution (“photo match”) can enhanced or refuted by being able to identify, account for and exclude other possibilities. This is referred to as exclusionary analysis. The logic trail for the relationship between exclusionary and imagery analysis as it relates to direct photographic attribution (“photo matching”) runs along these generic lines. By way of illustrative example, let’s say that:

-Elmer Fudd likely had 3-5 jerseys available for wear throughout the entirety of the 1956 season based on what we know or believe about ordering/inventory patterns.

-Through imagery analysis, we have identified and accounted for at least 3 separate jerseys worn by Elmer Fudd in 1956.

-Imagery analysis has permitted us to exclude 2 of the 3 jersey worn by Fudd from being the one offered for consideration.

-The one remaining jersey is consistent with images available of Elmer Fudd from 1956.

-While the probability might be high that the offered jersey is the one Fudd is wearing in photographs from 1956, no definitive claim of direct photographic attribution is made because there are other possibilities that have not been accounted for an excluded.

I take the time to lay all of this out as a both an educational and cautionary message to this hobby/industry because before someone runs off and claims a “photo match”, it is essential that they first identify what they know about the known or likely population and then physically demonstrate how these various jerseys have been identified and distinguished from one another, and then “excluded” from consideration. When this is not done, any claims of direct photographic attribution (“photo matching”) are not objectively or statistically defendable.

What follows is a combination of imagery and exclusionary analysis as it relates to the 1938 Lou Gehrig New York Yankees road jersey being offered by Rob Lifson and Robert Edwards Auction (REA). I think it very important to note that this work includes analysis of images I did not have or had not seen at the time my original opinion was provided. The overwhelming majority were located and provided to me by Mr. Peter Nash and for that I am extremely grateful.

Jersey #1: This jersey is identified as being separate from Jersey #2 based on the lettering alignment with the button line. Jersey #1 also cannot be Jersey #3 because of the sleeve length. Although sleeves can be shortened, the images of Jersey #1 predate those of Jersey #3. The jerseys shown in PLATES A & B are considered the same jersey at this time as I can find no readily apparent disqualifiers. It should be noted that while PLATE B bears a date of 9/26/1938, this is not the date of the photograph. Rather it is date the image was provided for use with the caption and credits. I confirmed this by looking at the 1938 New York Yankees schedule for the period in question:

September 7-8: Boston

September 9-11: Washington

September 13: Cleveland

September 15-17: Detroit

September 18-19: St. Louis

September 20-22: Chicago

September 24-26: Home against Boston

In my opinion, it is not unreasonable to presume this image is from one of the previous road games (although it could be from any time prior to the 23rd of September as well). This jersey is also consistent with the image from April 16th of 1938, so I do not disqualify them from being the same jersey. You will also notice that this jersey does not feature the World Fair Patch worn by the New York Yankees, New York Giants, and Brooklyn Dodgers in 1938.

Jersey #2: This jersey is excluded from being the offered jersey based on the button alignment. Although this observable characteristic may appear to artificially vary in images based on angle of photograph and fabric folds/body position, the image angle and rather upright body position of Gehrig in the photograph in question seem to negate any appreciable impact those factors might have on this characteristic. Since Jersey #2 is neither Jersey #1 nor Jersey #3 based on sleeve length/lettering alignment, it will not be discussed further in any material manner. (PLATE C)

Jersey #3: Jersey #3 refers to those images that I either believe to be or can be shown to be from the 1938 World Series. I attribute the 1938 team photo to the World Series based on the fact that the Yankees are appearing in road uniforms and the 1938 World Series opened in Chicago. This would not be an uncommon event for a team to pose for such a picture. The composite details (both physical and those affected by angle of image) of the jersey identified as Jersey # 3 are such that I cannot exclude them from being the same jersey. Additional images of the team photo from different points of reference confirm that the sleeves Gehrig is wearing in this picture are elbow length, which is not consistent with Jersey #1 or Jersey #2. Longer sleeve length appears consistent with the other images that can be placed to the 1938 World Series. (PLATES D thru G)

In addition, and upon closer comparison, Jersey #3 is excluded from being the offered jersey based on the spacing between the letters “E” & “W” in NEW YORK. While the general alignment with the button line is consistent, it does appear to feature a spacing between the letters “E” & “W” in NEW YORK that is clearly less than one button width. This relies on the assumption that the jersey in the team photographs is the same one as the ones from the World Series.(PLATES H, I)

I also exclude Jersey #3 or any other jersey with long sleeves and the World’s Fair Patch from being the same jersey as Jersey #1 or the offered jersey. The fabric cuts that I found on the offered jersey are in a location that would limit the application of that patch to have being placed on a jersey of longer sleeve length since the World’s Fair Patch was sewn to the bottom of the sleeve.

Based on the fact that I have not been able to account for other possible jerseys, and then exclude them, I offer no definitive claim of photographic attribution that the offered jersey is the same as Jersey #1. What I feel this imagery analysis in the PLATES does show is that:

-Both the offered jersey and Jersey #1 share a common sleeve length.

-Sleeve length is a discriminator based on the chronology of the photographs.

-Both the offered jersey and Jersey #1 can be seen without the 1939 World’s Fair Patch.

-Both the offered jersey and Jersey #1 share common and consistent font alignment with respect to the button line and consistent spacing between the letters “E” & “W.”

What a combination of imagery and exclusionary analysis indicates to me is that the offered jersey could be Jersey #1. In addition, since we now have images of Gehrig wearing a jersey without the World’s Fair Patch, it is quite conceivable that no patch was ever applied to the jersey and that the location of the fabric cuts is a coincidence. This would not have any appreciable impact on my overall opinion of the jersey and would only affect the overall grade assigned MEARS by +.5 since the A9 opinion provided by Dave Bushing did not deduct points for this. In my opinion (A5), I deducted -1 for the patch. Thus a revised grade of A6 from me would leave the aggregate grade at A7.5 (average A6 & A9) as opposed to A7.

This article has been about two separate but related issues; process and product. My hope is that the hobby/industry now has a better understanding of the essential relationship between imagery and exclusionary analysis when it comes to supporting a position of direct photographic attribution (“photo matching”). When the collecting community or auction houses conduct or subsidize substandard work, process, and analysis, then we should not be surprised when it results in “photo match” claims that are not objectively or statistically defendable. Please know for my work in this area, I leverage no special tools or technology. If you are looking at the images on a computer screen and care to invest about $1.00 in a clear ruler you can lay on the screen, then you are armed with the same tools as me. This only leaves differences in experience, training, and process.

This article has also been about a product or the ability for the collector to see and know as much as possible about a particular jersey. In the case of this 1938 Lou Gehrig New York Yankees Road Jersey, I maintain my original position that “based on my physical examination of this jersey and supporting data, it is my opinion that this jersey posses all of the characteristics you would expect to find in a 1938 New York Yankees road uniform supplied by Spalding for use and wear by Lou Gehrig during this period”.

This is an opinion based on:

-Imagery analysis.

-Comparative trend analysis with other Lou Gehrig and period New York Yankee jerseys I have examined.

-Comparative fabric analysis with other period and year specific Spalding major league grade/quality uniform products.

-Exclusionary analysis to determine a likely population and offering data to support either the possibility of exclusion or inclusion with respect to the issue of direct photographic attribution.

-A physical examination of the jersey utilizing UV lighting, a light table, illuminated magnification, and the use of a digital microscope.

I also maintain my previous position of that “if you are serious buyer then take this jersey very seriously. Although there are some condition issues, it is one of only a handful of legitimate Lou Gehrig jerseys out there and the likely population of good Gehrig jerseys can only to get smaller. This population has been cut down significantly as I have evaluated three others that were found to be problematic for any number of reasons. If you are in the market, then buy the jersey for what it is based on what you can objectively see and know. I think you’ll be thrilled and lucky to have it as the display appeal is incredible and the conservation work in the sleeves is museum quality”.

As always, collect what you enjoy and enjoy what you collect.

Dave Grob

For questions or comments on this article, please feel free to drop me a line at DaveGrob1@aol.com.

“My name is Dave and I am a collector.” The reply “Hi Dave” comes back from the assembled group. I spend the next few minutes describing why I search E-Bay, flea markets, yard sales and any number of other venues looking for “baseball stuff.” I recount the find of the 1950s Ted Kluszewski cracked model U1 gamer in a Cincinnati area Play It Again Sports store in the late 1990s. I can see the look in their eyes…they too have dreamed of stumbling across “the find of the century.” If only there was a Collector’s Anonymous.

Collectors like the hunt to be sure. At times as we stalk piles of dusty and cracked bats, gloves, uniforms etc… we find something that appears to hold great promise, only to have our hopes dashed to find out it’s not a gamer. Other times we find a gamer that turns out to something other than what we had hoped for…still other times we find something of with great promise, and despite doing our best are left with a “could very well be”. This later scenario is the subject of this article.

As I was looking for bats from the Cincinnati Reds 1919 World Series, I came across this bat. It is a Hillerich and Bradbsy block name Williams bat from the general labeling period of 1916 to 1921. This bit of potentially Legendary Lefty Lumber was listed by the very experienced and reputable vintage bat dealer Art Jaffe. Art had the great fortune and foresight to purchase hundreds of bats years ago from Louisville Slugger. This bat was listed as:

Player: Williams, (Lefty?)

Label: H&B

DDD:

Era: circa 1919

Teams: Black Sox? Banned with Joe Jackson?

Side Written: Williams

Team Side Written

Art listed just enough information to peak my interest and I felt the bat, while not cheap, was priced right especially if I could put it in the hands of the White Sox pitcher who lost 3 games in the 1919 Series. Art had previously sold another such block letter Williams with a similar description, only being from the 1911-1916 general labeling period. I liked this one better given the range of the labeling period and that it included the World Series year of 1919.

I had only seen one other bat besides these that was said to have been possibly linked to Claude “Lefty” Williams. That bat was offered by Robert Edwards Auction in March of 2001 (Lot # 409; 36” 42 oz). The bat was attributed to Lefty Williams since it was said to be the same model and dimensions of a previously seen Swede Risberg bat, a team mate of Lefty’s. That Risberg bat was featured as lot # 510 in the Barry Halper Auction and is listed as being 36” and 37 oz. This Williams bat is 36” and 38 oz.
While I am not calling this attribution into question of the REA offered Williams based on the Risberg bat, I decided to take a different approach by beginning to look into the Williams’ who played major league baseball during this time frame and what their status was with to respect Hillerich & Bradsby.

Players whose last name was Williams who played in the majors during the labeling period of this bat produced these results. (Source being www.baseballreference.com). With the names I came up with, I also cross walked it against a list of Hillerich & Bradsby endorsers.

Cy Williams 1912-1930 (H&B Contract endorser) Cy Williams can also be found with retail offerings in the Spalding catalogs from 1917-1924. His bats in their catalogs were only offered in even whole inch (32,33,34,35, etc) lengths.

Denny Williams 1921-1928 (H&B Contract endorser) 7 AB’s during this period and only in 1921. Did not play in 1922 or 1923.

Ken Williams 1915-1929 (H&B Contract endorser)

Rip Williams 1911-1916, 1918 (Non-H&B Contract Endorser). Only 28 games in 1918.

Claude “Lefty” Williams 1913 -1920 (Non-H&B Contract Endorser) 195 AB’s during this period.

Marsh Williams 1916(Non-H&B Contract Endorser). Appeared in only 10 games in 1916.

These are some things I know:

Fact: Players with endorsement contacts with Hillerich and Bradsby had their signature names burned onto the barrel of the bat. Besides Rip and Marsh Williams, Lefty Williams is the only other one of the five Williams from this period who did not have an endorsement contract with Hillerich and Bradsby.

What complicates this is that Cy Williams appears to have been under contract with Spalding as well. Based on what I wrote about last year with respect to player endorsement contracts, I am not sure Cy might not have also been a bit sly in his business dealings.

Fact: This bat features a different style barrel from a Ken Williams signature model H&B bat. The Ken Williams’ signature model bat is 34” and 34.3oz. This bat is 36” and weighs just over 38 oz. This Williams bat is hand turned with rasp markings on both the barrel and knob ends. This bat happens to be the same knob style, length and approximate weight as the Lefty Williams bat in the June 2001 REA Auction.

I have been looking for a Cy Williams bat to do a similar comparison with as with Ken Williams. I got some help from Mike Specht when I asked him about this. While Mike did not have any production information on Cy Williams either , he did point out that:

Charles Gehringer records– On 2/1/1926 order as “Joe Kelly Toronto 5/20/25 — same as Cy Williams 34″ / 35 oz” — This (Joe Kelly 5/20/25) was also ordered on 7/21/26 and that he (Mike Specht) would interpret the notations penned above to indicate that Williams bat is indexed at a length other than the 34″ and the 34″ notation being a departure, although the way it is written could suggest that the 34″ is the same as Williams’ length which would be a departure from the Kelly index length.”

As mentioned previously, Cy Williams can also be found with retail offerings in the Spalding catalogs from 1917-1924. His bats in these catalogs were only offered in even whole inch (32,33,34,35, etc) lengths. An autograph model of one of his Spalding bats has turned up recently in the length of 33 ½ inches, so Mike might be right about Cy Williams using a bat other than 34” inches as this would the case at 33 ½”.

I have no idea at this time how many bats were ordered by any of these six individual Williams. Even if you take Cy and Ken Williams out the mix by either endorsement relationship or bat type/model/length, then how to we assess the overall likelihood of the bat being one of Lefty, Denny, Marsh or Rip Williams?

The general labeling period of the bat is 1916-1921. During this period, consider the number of seasons each one of these men played.

Lefty Williams: 4 of the 6 seasons.

Rip Williams: 2 of the 6 seasons.

Denny Williams: 1 of the 6 seasons.

Marsh Williams: 1 of the 6 seasons.

This leaves Lefty Williams as having had 4 times the number of seasons in which to have had a bat made for him than either of Marsh or Denny Williams and twice as many as Rip. Based on the number of years involved, the number of bats likely to have been made each year, and considering these factors are likely to have an impact on the chances availability and survival as a function of number, then this bat has a much higher probability of being one of Lefty Williams then either Rip, Marsh or Denny Williams. None of this makes this bat a Lefty Williams bat, but rather indicates that it has a much better chance of being one of his as opposed to Rip, Marsh or Denny Williams.

An additional point must be made before going any further. There is also the possibility that the bat was made for a Williams during this period that never made it the major leagues (either a minor or negro league player). There is no side writing on the bat indicating it was sent in by a minor league, negro club or any other player. What I am left here with is the question that has haunted intelligence professionals since the dawn of time, how do you prove a negative?

One thing that may be working in favor of this bat as Lefty Williams offering, might be that while the bat is cracked, the bat was not repaired for later use. This may be an indicator that the bat had some special significance when discarded. Clearly Lefty Williams was a more significant player than Rip, Marsh, or Denny.

I also decided to take a look for information that could be contained in period images. It came as no great surprise, that I could not find images of Rip, Marsh, or Denny Williams.

Cy Williams:

I found a picture of Cy Williams and Rogers Hornsby from 1927. Although the photograph was undated, the year could be identified since this was the only year Hornsby played with the NY Giants (and yes, I realize it is a posed shot).

Cy Williams was listed as being 6’, 2” in height or a total of 74 inches. In this shot he is not standing completely erect. Neither is Hornsby who is listed at 5’, 11” in height.

In the image Williams measures some 122mm. The half way point of his height in this image is marked in blue. Even if his slouched height is 72”, the half way point would be some 36”. Based on the bend of his wrist, I am assuming he has the knob of the bat in his palm (marked in a broken red line). You can see that the bat Williams is holding is very likely shorter than 36” and may be more in line with the shorter Spalding bat referenced above. I also found a picture of Cy Williams with Hack Wilson and the bat Williams has appears to be the same size (proportionally) as the one in the Williams/Hornsby picture.

Ken Williams:

In an image from August of 1922, Ken Williams can be seen with a bat that has tape on the handle. This Williams bat has no tape nor signs that it ever did. Yes I know Ken Williams could be looking at this bat in the picture and wondering why a teammate chose to use this type of bat. The other thing to consider is that Ken Williams may have taped bats at certain times and not at others.

At the end of the day, or actually many days of research, I am not sure this bat was one used by Lefty Williams, but I am sure it could have been. This would have been the same answer I would have given a collector if they sought my opinion on this Williams bat. You can’t have one opinion for your items and another opinion or standard of proof for the items of others…and yes, I see this all the time from collectors who are looking to make their own items look good at the expense of other items and other collectors. An item “is what it is” no matter who owns it and don’t let anyone every try to convince you other wise.

The purpose of this article has been to highlight a seldom used form of analysis that seems to be overlooked, especially by today’s “photo matching” crowd. The type of work I am referring to is called exclusionary analysis. The premise is that at times, one theory might have a higher probability over other options because there is something in the other options that tends to exclude them or diminish their likelihood. I used this type of exclusionary analysis in my re-look of the 1932 Babe Ruth “Called Shot” jersey.

If you are looking at items of your own or those you are interested in acquiring, consider looking at them in a similar manner that I have done with this bat. In the end, you will know more about you have or might have and why. You will also be armed with a logical methodology to question the work and opinions of others. In my opinion, this is something sorely lacking in this hobby/industry today.

As always, collect what you enjoy and enjoy what you collect.

Dave Grob

For questions or comments on this article, please feel free to drop me a line a DaveGrob1@aol.com

“My name is Dave and I am a collector.” The reply “Hi Dave” comes back from the assembled group. I spend the next few minutes describing why I search E-Bay, flea markets, yard sales and any number of other venues looking for “baseball stuff.” I recount the find of the 1950s Ted Kluszewski cracked model U1 gamer in a Cincinnati area Play It Again Sports store in the late 1990s. I can see the look in their eyes…they too have dreamed of stumbling across “the find of the century.” If only there was a Collector’s Anonymous.

Collectors like the hunt to be sure. At times as we stalk piles of dusty and cracked bats, gloves, uniforms etc… we find something that appears to hold great promise, only to have our hopes dashed to find out it’s not a gamer. Other times we find a gamer that turns out to something other than what we had hoped for…still other times we find something of with great promise, and despite doing our best are left with a “could very well be”. This later scenario is the subject of this article.

As I was looking for bats from the Cincinnati Reds 1919 World Series, I came across this bat. It is a Hillerich and Bradbsy block name Williams bat from the general labeling period of 1916 to 1921. This bit of potentially Legendary Lefty Lumber was listed by the very experienced and reputable vintage bat dealer Art Jaffe. Art had the great fortune and foresight to purchase hundreds of bats years ago from Louisville Slugger. This bat was listed as:

Player: Williams, (Lefty?)

Label: H&B

DDD:

Era: circa 1919

Teams: Black Sox? Banned with Joe Jackson?

Side Written: Williams

Team Side Written

Art listed just enough information to peak my interest and I felt the bat, while not cheap, was priced right especially if I could put it in the hands of the White Sox pitcher who lost 3 games in the 1919 Series. Art had previously sold another such block letter Williams with a similar description, only being from the 1911-1916 general labeling period. I liked this one better given the range of the labeling period and that it included the World Series year of 1919.

I had only seen one other bat besides these that was said to have been possibly linked to Claude “Lefty” Williams. That bat was offered by Robert Edwards Auction in March of 2001 (Lot # 409; 36” 42 oz). The bat was attributed to Lefty Williams since it was said to be the same model and dimensions of a previously seen Swede Risberg bat, a team mate of Lefty’s. That Risberg bat was featured as lot # 510 in the Barry Halper Auction and is listed as being 36” and 37 oz. This Williams bat is 36” and 38 oz.
While I am not calling this attribution into question of the REA offered Williams based on the Risberg bat, I decided to take a different approach by beginning to look into the Williams’ who played major league baseball during this time frame and what their status was with to respect Hillerich & Bradsby.

Players whose last name was Williams who played in the majors during the labeling period of this bat produced these results. (Source being www.baseballreference.com). With the names I came up with, I also cross walked it against a list of Hillerich & Bradsby endorsers.

Cy Williams 1912-1930 (H&B Contract endorser) Cy Williams can also be found with retail offerings in the Spalding catalogs from 1917-1924. His bats in their catalogs were only offered in even whole inch (32,33,34,35, etc) lengths.

Denny Williams 1921-1928 (H&B Contract endorser) 7 AB’s during this period and only in 1921. Did not play in 1922 or 1923.

Ken Williams 1915-1929 (H&B Contract endorser)

Rip Williams 1911-1916, 1918 (Non-H&B Contract Endorser). Only 28 games in 1918.

Claude “Lefty” Williams 1913 -1920 (Non-H&B Contract Endorser) 195 AB’s during this period.

Marsh Williams 1916(Non-H&B Contract Endorser). Appeared in only 10 games in 1916.

These are some things I know:

Fact: Players with endorsement contacts with Hillerich and Bradsby had their signature names burned onto the barrel of the bat. Besides Rip and Marsh Williams, Lefty Williams is the only other one of the five Williams from this period who did not have an endorsement contract with Hillerich and Bradsby.

What complicates this is that Cy Williams appears to have been under contract with Spalding as well. Based on what I wrote about last year with respect to player endorsement contracts, I am not sure Cy might not have also been a bit sly in his business dealings.

Fact: This bat features a different style barrel from a Ken Williams signature model H&B bat. The Ken Williams’ signature model bat is 34” and 34.3oz. This bat is 36” and weighs just over 38 oz. This Williams bat is hand turned with rasp markings on both the barrel and knob ends. This bat happens to be the same knob style, length and approximate weight as the Lefty Williams bat in the June 2001 REA Auction.

I have been looking for a Cy Williams bat to do a similar comparison with as with Ken Williams. I got some help from Mike Specht when I asked him about this. While Mike did not have any production information on Cy Williams either , he did point out that:

Charles Gehringer records– On 2/1/1926 order as “Joe Kelly Toronto 5/20/25 — same as Cy Williams 34″ / 35 oz” — This (Joe Kelly 5/20/25) was also ordered on 7/21/26 and that he (Mike Specht) would interpret the notations penned above to indicate that Williams bat is indexed at a length other than the 34″ and the 34″ notation being a departure, although the way it is written could suggest that the 34″ is the same as Williams’ length which would be a departure from the Kelly index length.”

As mentioned previously, Cy Williams can also be found with retail offerings in the Spalding catalogs from 1917-1924. His bats in these catalogs were only offered in even whole inch (32,33,34,35, etc) lengths. An autograph model of one of his Spalding bats has turned up recently in the length of 33 ½ inches, so Mike might be right about Cy Williams using a bat other than 34” inches as this would the case at 33 ½”.

I have no idea at this time how many bats were ordered by any of these six individual Williams. Even if you take Cy and Ken Williams out the mix by either endorsement relationship or bat type/model/length, then how to we assess the overall likelihood of the bat being one of Lefty, Denny, Marsh or Rip Williams?

The general labeling period of the bat is 1916-1921. During this period, consider the number of seasons each one of these men played.

Lefty Williams: 4 of the 6 seasons.

Rip Williams: 2 of the 6 seasons.

Denny Williams: 1 of the 6 seasons.

Marsh Williams: 1 of the 6 seasons.

This leaves Lefty Williams as having had 4 times the number of seasons in which to have had a bat made for him than either of Marsh or Denny Williams and twice as many as Rip. Based on the number of years involved, the number of bats likely to have been made each year, and considering these factors are likely to have an impact on the chances availability and survival as a function of number, then this bat has a much higher probability of being one of Lefty Williams then either Rip, Marsh or Denny Williams. None of this makes this bat a Lefty Williams bat, but rather indicates that it has a much better chance of being one of his as opposed to Rip, Marsh or Denny Williams.

An additional point must be made before going any further. There is also the possibility that the bat was made for a Williams during this period that never made it the major leagues (either a minor or negro league player). There is no side writing on the bat indicating it was sent in by a minor league, negro club or any other player. What I am left here with is the question that has haunted intelligence professionals since the dawn of time, how do you prove a negative?

One thing that may be working in favor of this bat as Lefty Williams offering, might be that while the bat is cracked, the bat was not repaired for later use. This may be an indicator that the bat had some special significance when discarded. Clearly Lefty Williams was a more significant player than Rip, Marsh, or Denny.

I also decided to take a look for information that could be contained in period images. It came as no great surprise, that I could not find images of Rip, Marsh, or Denny Williams.

Cy Williams:

I found a picture of Cy Williams and Rogers Hornsby from 1927. Although the photograph was undated, the year could be identified since this was the only year Hornsby played with the NY Giants (and yes, I realize it is a posed shot).

Cy Williams was listed as being 6’, 2” in height or a total of 74 inches. In this shot he is not standing completely erect. Neither is Hornsby who is listed at 5’, 11” in height.

In the image Williams measures some 122mm. The half way point of his height in this image is marked in blue. Even if his slouched height is 72”, the half way point would be some 36”. Based on the bend of his wrist, I am assuming he has the knob of the bat in his palm (marked in a broken red line). You can see that the bat Williams is holding is very likely shorter than 36” and may be more in line with the shorter Spalding bat referenced above. I also found a picture of Cy Williams with Hack Wilson and the bat Williams has appears to be the same size (proportionally) as the one in the Williams/Hornsby picture.

Ken Williams:

In an image from August of 1922, Ken Williams can be seen with a bat that has tape on the handle. This Williams bat has no tape nor signs that it ever did. Yes I know Ken Williams could be looking at this bat in the picture and wondering why a teammate chose to use this type of bat. The other thing to consider is that Ken Williams may have taped bats at certain times and not at others.

At the end of the day, or actually many days of research, I am not sure this bat was one used by Lefty Williams, but I am sure it could have been. This would have been the same answer I would have given a collector if they sought my opinion on this Williams bat. You can’t have one opinion for your items and another opinion or standard of proof for the items of others…and yes, I see this all the time from collectors who are looking to make their own items look good at the expense of other items and other collectors. An item “is what it is” no matter who owns it and don’t let anyone every try to convince you other wise.

The purpose of this article has been to highlight a seldom used form of analysis that seems to be overlooked, especially by today’s “photo matching” crowd. The type of work I am referring to is called exclusionary analysis. The premise is that at times, one theory might have a higher probability over other options because there is something in the other options that tends to exclude them or diminish their likelihood. I used this type of exclusionary analysis in my re-look of the 1932 Babe Ruth “Called Shot” jersey.

If you are looking at items of your own or those you are interested in acquiring, consider looking at them in a similar manner that I have done with this bat. In the end, you will know more about you have or might have and why. You will also be armed with a logical methodology to question the work and opinions of others. In my opinion, this is something sorely lacking in this hobby/industry today.

As always, collect what you enjoy and enjoy what you collect.

Dave Grob

For questions or comments on this article, please feel free to drop me a line a DaveGrob1@aol.com