Thursday, January 05, 2006

The 1918 Season Recompiled

The 1918 Season Recompiled

That brings us to 1918, the season where the Coast League threw in the towel at mid-season. In the Coast League Cyclopedia, we used the pitching records as compiled by Robert C. Hoie, of San Marino, California, a noted expert on the PCL. He assured us that it was one of the best and most accurate pieces of work he had done, and we had no reason to dispute that, given his reputation.

Last year we decided to compile the batting side of the 1918 season. We felt that there was a good chance— considering that only individual batting stats made it to print— that those batting statistics had been sloppily compiled. At the same time we compiled our batting stats, we compiled pitching stats in order to see that runs and hits made by batters matched runs and hits allowed by the pitchers. We expected some minor discrepancies between our compilation and the Robert C. Hoie compilation, and a great number between our batting stats and the batting record published the following spring in the baseball guides.

Again, we couldn’t have been further off the mark! As it turned out, we found relatively few discrepancies with our compilation and that of the league statistician on the batting side of the equation. With Mr. Hoie, it was quite a different matter. In the four categories (games pitched, games started, complete games, shutouts) where it’s virtually impossible for one box score to differ from another, we found fully 20% of pitchers’ records in the league to be in error in the Hoie compilation. And, overall, we found 119 “discrepancies,” many of which we believe cannot be explained away by the use of different box scores.

To give the reader some idea of what we call “discrepancies,” we will give the details on three that we find instructive:

1) While Mr. Hoie lists Hi West with 78 strikeouts, we were only able to find 28. While a strikeout or two might be explained away by a different source newspaper, fifty seems to be somewhat beyond that type of explanation.

2) Mr. Hoie credits Jack Picus Quinn with 13 victories, yet Quinn pitched complete games in 14 contests where his team scored the most runs. And in five of those contests (on April 9, May 15, June 16, June 26, and July 9), the opponents did not score, which— according to Mr. Hoie— gave pitcher Quinn four shutouts! (Additionally, it did not apparently change the win-loss record of the pitchers he opposed. Another perplexing turn of events.) We checked every one of Quinn’s games pitched in the following papers: San Francisco Chronicle, San Francisco Examiner, Sacramento Bee, Los Angeles Times, The Sporting News, and still were unable to come up with an adequate explanation for Mr. Hoie’s “discrepancies.” Considering that Quinn’s ERA set a PCL record, you would think Mr. Hoie would at least gone the extra mile, but apparently not.

3) Spider Baum is one of the great PCL— and minor league— pitchers of all time. Mr. Hoie lists Baum with 7 losses. We have 8. In five of those losses, Baum was the only San Francisco pitcher, so there can be no dispute about those games. In two of the other three games, Baum gave up all the runs, so there cannot be much of a dispute about who lost. In the third, Baum gave up 4 of the 7 runs scored, and his team never held the lead. Three of the four papers we consulted credited him with the defeat in the box score summary, the LA Times box not crediting any pitcher with the loss. Additionally, if this were the game that produced the discrepancy, then presumably the other pitcher, Tom Seaton, would have his record affected, but both ours and Mr. Hoie’s compilations list Seaton pitcher with identical 11-8 records. What Mr. Hoie was doing is beyond me.

While Mr. Hoie might be a fine historian— though we have found only two articles with his name on them— it appears to be nothing more than a dilettante when it comes to compiling league averages. The point being: It’s not that easy to compile averages from box scores, and— by extension— a book like this, with more than a half-million pieces of data is bound to have its share of errors, though we have made every effort to keep them to a minimum. And so we ask an reader who finds any possible error to contact us. After all, they are still making corrections to the major league record in the encyclopedias some thirty-five years after the first Macmillan Baseball Encyclopedia came out.

One additional note on 1918: In compiling our pitching averages, we came across several games where earned runs (called runs responsible for by the PCL) did not appear in any of the box scores we consulted. What we did was use the weekly pitching stats issued by the league, and published in several papers around the league every Tuesday, that gave wins, losses, and runs responsible for to determine earned runs for the few games in which they were missing. When we didn’t have the RRFs for a game, we subtracted what we had for each pitcher involved in the game from the weekly RRF totals. In no case was the pitcher involved in more than one “missing RRF game” during the week in question.

Our 1918 standings are also at variance with the official published record, as indicated below:

We believe the following will demonstrate why the published league standings are in error:

By 1918 standings were carried on a daily basis in most— if not all— the newspapers in league cities. Therefore, we immediately knew the date when the errors took place. With the standings issued after games played on Sunday, June 9, we found the one of Vernon’s two wins over Sacramento and San Francisco’s loss at Salt Lake were not counted for some reason in the four papers consulted (SF Chronicle, SF Examiner, Sacramento Bee, LA Times). After games played on July 13, the LA Times listed the Vernon game at Salt Lake City as a rainout; however, the box score for the game appeared in the three other newspapers consulted (The Sporting News, SF Chronicle, SF Examiner). The published standings in the latter two papers did not reflect the Vernon win and the Salt Lake City loss.

At the league meetings that winter, the 1918 season had been long forgotten, and the owners only concentrated on the coming 1919 season, and the expansion of the league to eight clubs. We found no discussion of protested games, or even about compiling pitching stats to go with the batting stats that were issued by the league on a weekly basis during the season. No pitching averages were ever issued, so we believe that was the reason those games were never picked up by the league statistician. Had the league statistician— who did such a marvelous job on the batting— finished compiling the pitching statistics, we believe the standings would have been found to be in error and corrected, especially since the official batting statistics apparently included those games.


Post a Comment

<< Home