Saturday, October 18, 2008

Chasing the Colonels: Compiling the Early Stats

Today is Saturday, October 18, 2008.

I have just spent the last eight days developing a comprehensive roster listing for the 1903 Louisville Colonels. I’ve been doing nearly nothing but working on this. But that’s how long it takes. And there are portions of the record which remain incomplete. But more on that later.

Roughly 14 months ago I began the task of compiling the complete rosters for the American Association’s Louisville Colonels form 1902-1954. After completing the initial steps in the process, I was able to advance through the 1915 season before it was time to begin “fleshing out” the roster for each year. A few weeks ago I started with the 1912 season because I knew it would be one of the most demanding, considering that the team had nearly 50 players on its roster at one point or another (at this point I should define “complete roster” -- the approach I’m taking involves establishing the record for each active player, so anyone who did not actually appear in a box score isn’t part of this project; I am using a complete set of box scores I photocopied from Sporting Life microfilm). The process of completing the basic stat line for each player during the 1912 season occupied roughly 60 hours, and that was after completing preliminary work which involved going through the 1903 box scores to make sure I had all the names, positions, etc.

Proceeding in a rather roundabout way, I tackled the 1902 season next. This went fairly well; I was able to reconcile the sum of the pitchers’ wins and losses with those of the official team record with ease. I completed a separate data sheet for each individual player, entered the data into my database and made adjustments for all of the new information, relying upon the text covering the rosters of each American Association team by Marshall Wright and the newly accessible database of minor league rosters at

Then last Friday I started to complete the record for the 1903 Louisville team.

There were a lot of problems getting the numbers to justify for 1903, and I finally realized last night that I would not be able to reconcile the official number of the team’s wins because, apparently, there was a game played sometime between June 26 and July 2, 1903 that was never added to the published record, not in Sporting Life or in the Sporting News. There was also no record of any forfeit that I hadn’t already noted.

I was simply unable to account for this game.

The final official record for the second-place Colonels was 87-54 (they played 143 games in a 140 game season, two of them being withheld from the final record for whatever reasons); the final record I was able to come up with, by various means, was 86-54. Sporting Life had them at 87-53, so perhaps somewhere along the line they got it wrong. The Sporting News has a different record yet.

I’ve been over my game-by-game record file many times and I can account for 87 wins, but my pitchers account for 86. There were two forfeits to be included as wins for the Colonels, but that’s figured into the 87 wins.

So for now I’m leaving things the way they are.

My record is not perfect, but it fills in the blanks for each player for whom there never was a published record (principally because they didn’t play in enough games to get published by the guides of the time).

I was also able to verify changes in a few pitchers’ records, something I consider notable. For example, the league’s leading pitcher, Tom Walker, is listed with a record of 26-7 in standard sources. By going over (and over and over...) each box score I was able to substantiate that Walker actually finished the 1903 season with a record of 27-6 with the league’s highest winning percentage as well as number of wins. This compares with a 26-7 record.

Here is a summary of the step-by-step process I use for each season:

1. Print published roster (with stats).

2. Using collection of Sporting Life box scores, use highlighter to indicate all games for Louisville.

3. Complete “Season Index of Games” file to account for all games played in published box scores. This has proven to be an invaluable tool for cross-checking pitchers' won-loss info.

4. Compare pitchers’ #W/L with Team #W/L.

5. Compare Season Schedule of Games with published box scores.

6. Note games which were called or forfeited.

7. Create Listing of Players absent from published records.

8. Complete/Revise Record for Player Batting (for multi-team players and “less thans”) and add player position for each game.

9. If not completing a BATTING Record for a particular player, complete Fielding Record as necessary.

10. Complete/Revise Record for Pitchers, esp. games started and complete games and shutouts.

11. Complete Missing Data form for all blanks and question marks.

12. Enter data into Data Form for Base Roster.

13. Check for errors:

--compare Pitcher Record with Season Index of Games for each pitcher to help reconcile team/pitcher W/L record.
--look for blank spaces in the batter records and make sure any blank has a corresponding notation for either the “Trouble Board” or my list for requests to the libraries for research assistance for which a dedicated clipboard is prepared.

I’ve just started using a “Trouble Board.” It’s a dedicated clipboard I use to make make notations on outstanding questions. I started using it as a way to keep track of unresolved questions arising toward the end of the roster compilation process. The Trouble Board helps me cross-check information which was absent from the box score summary.

One thing I’ve found especially valuable as a cross-checking tool is the file I call the “Season Index of Games,” a game-by-game record of the season I use a spreadsheet for. It includes the date, opponent, runs scored by Louisville, runs scored by opponents, starting pitcher and winning pitcher. For the 1903 record, I used this file extensively to help me determine the date of the missing game I was searching for. I was able to determine that the missing game took place between June 26 and July 2, where Sporting Life’s won/loss record in the standings differed from my record. As of this writing that game remains a question.

Pertaining to the photo of the 1903 Louisville Colonels, most of the players are unidentified. But there is certainty for the following: back row center is Louisville outfielder Fred Odwell who was quite a slugger despite his boyish demeanor in the photo. He led the American Association in triples with 19 in 1903 and hit .317 while leading Louisville with 8 home runs.

In the center row, second from left is team manager and outfielder Bill “Derby Day” Clymer who was nearly 30 years of age at the time of the photograph. Second from right is the league’s top winning pitcher, Thomas William Walker who posted a 27-6 record (contrary to the published record of 26-7) in over 300 innings of work. It was his grandson, Ted Walker of Pennsylvania, who supplied this photograph and ID notes. To the right of Walker is Wish Egan who entered the Detroit organization and excelled as a scout, named such after his first name Aloysius. Despite Egan’s published 24-16 record, my efforts declare he won 25 while losing 15. This difference may be attributable to a difference in how the rules of the day were interpreted.

A photograph (not too good) from the microfilm of the Minneapolis Journal depicting the 1903 Colonels with each of the players identified. Each player shown is wearing a ball cap which is characterized by a colored stripe between the bill and the cap.

In the above photo there is no one wearing a ball cap. It’s amazing how a player’s persona, based upon their facial appearance, changes upon the wearing of a cap. The players appear older with a cap on.

The images of the remaining players in the above “cap-less” photo who remain nameless at this point are difficult identify in the semi-corresponding Journal photo because of their caps. However, it is plausible that the player lying to the right in the front is outfielder Dan Kerwin who led the league with 141 games played (a 140-game season) and in at-bats with 70. Kerwin hit .273 while swiping 33 bases and leading the Louisvilleans with 192 hits. Opposite Kerwin is likely the outstanding Louisville third-baseman Suter Sullivan who led the club with 33 doubles.