Strikeouts: the out of choice.[i] Current sabermetric analysis has defined a relationship between ballparks and strikeouts (SO or K): the ballpark SO factor (K PF). [ii] [iii] The K PF takes into account foul area (AF) and many other factors that might contribute to SO (e.g., altitude, humidity, wind, batter’s eye, etc), in a very global manner, while not quantifying the effects of any specific contributor. As hypothesized in our previous post, perhaps a historical effect of AF alone on K PF for all parks might be constructed based on the average square feet (30,500) and average # of SO / Sq. Ft. (0.0284) for the entire 1954-2000 period, or any part of it.
Using the data derived in the previous analysis indicates that for 1964-66 the average ballpark was 31,500 sq. ft. and Cleveland Stadium 43,700. By inference, it would also suggest the relationship identified for hitters should apply for SO by pitchers as well. Multiplying the average SO/sq. ft. for AL hitters during that period (0.0309) by the 12,300 sq. ft. difference suggests the staff may have struck out up to 190 additional batters at home yearly had AF been merely average, assuming: 1) a direct linear relationship between AF and SO; 2) the inverse relationship that increase in AF decreases SO accumulated by pitchers to the same extent shown for hitters; 3) other factors were not affecting the potential increase. Applying the same method, they may have SO an additional 59 batters yearly in 1967-68, when AF decreased to 35,000 sq. ft. Although the 3 requirements above are not plausible, it is of note that the numbers proposed are not vastly different from ranges of estimates offered by K PF analyses, where stadiums may vary up to 10-20% in total SO.[iv]
It is notable that the Cleveland Indians pitching staff of 1964-68 set a number of AL SO records, while pitching in the largest (1964-66) AF park in the AL, 43,700 sq. ft.. In a near-perfect experiment (for our purposes) of 1967, seats were added dropping it to 3rd in AF at 34.500 sq. ft. If the hypothesis were correct that SO/AF was dependent on more foul balls being caught on the field, catchers should have more putouts in stadiums with larger AF. In fact, compared to the ’64-66 average of 74 putouts, the Indians’ catchers’ non-SO putouts indeed decreased by 20 (37%) in 1967, and Indians’ pitchers’ SO increased to a new record high (1189 SO), a record that would last for 30 years until the steroid era.
Changes in numbers of foul pop-ups caught by the infielders in 1967 is not readily determined: an attempt to do so on Retrosheet showed it did not clearly distinguish between foul and fair pop-ups caught by catchers and infielders. Perhaps each side of the infield might contribute an additional 10-20 potential SO lost, or an even 50 season total. Where a staff is SO almost 1 batter/inning, 50 SO opportunities could lead to 10-15 SO/ season.
Whereas the estimates of SO based on this SO/ sq. ft. are likely exaggerated, and the data from box scores currently incomplete, a truer estimate might be achieved by examining the trendlines of figures previously published in posts here at BTB in the past 3 weeks, depicting rank order of stadiums/teams vs. average yearly SO (an example from the AL 1964-1968 again displayed below) .
Comparison of leftward point of the trendline to the most rightward ( SO by the clubs with the largest and smallest AF) might suggest the actual average maximum number of yearly SO AF might contribute for each time period: approximately 50-75, on average, based on the set of graphs previously posted, and analysis from 1969-2000 (not posted). Based on those comparisons, it is reasonable to believe the figure of 10-15 SO derived from the Cleveland experiment of 1967 might be underestimated. Fifty SO per season would be equivalent to approximately 5% of SO overall, a not-insignificant portion of total! Whereas the assumptions here are simplifications of a non-linear variable, application of this logic in this rudimentary analysis nevertheless supports the contention that Indians’ 1967 record total would then surpass the subsequent record numbers of 1208 by Seattle (1997), 1213 by Cleveland (2000), or even the Yankees’ single-season SO record of 1266 (2001), just before the steroid ban, had they not been pitching 81 games in Municipal Stadium. Their record numbers of 1964-68 would have even been higher, and would represent the largest differential between the SO leader and the rest of the AL ever.
SO…how is this relevant to other SO records and today’s staffs? We’ll address that in upcoming posts.
[i] Lederer R: A new Way to Measure strikeout Proficiency. http:baseballanalysts.com/archives/2006/02/a_new_way_to_me.php
[ii] Jacques K: Prospectus Toolbox. Non-Contact Part II: More on Strikeouts., http:www.baseballprospectus.cm/article?articleid=6559 , August 7, 2007
[iii] Carty D: Context adjustments: Park factors. www.Fandueal.com/insider/2011/03/21/context-adjustments-park-factors/[iii] http://www.answers.com/topic/fenway-park
[iv] Gassko D: Batted balls and park effects. Hardball Times, http://www.hardballtimes.com/main/article/batted-balls-and-park-effects/ March 20, 2008