Nap Clarke - Deadball Era Manager

1900-1910+ decade AIM micromanager Nap Clarke. Although designed essentially for early twentieth century replays (ca. 1905), Clarke can easily handle later replays up through the 1919 season, or what is historically considered the end of the dead ball era. He can, then, be used as competition for the other dead ball manager currently available, Ned Jennings.   Clarke is unique among twentieth century micromanagers in that he is designed to manage pitchings staffs while the replayer employs 19th century recovery rates, i.e., pitchers can pitch consecutive games and never tire. In order to prevent, however, overuse or underuse of pitchers, Clarke selects relievers based on a "weighted randomization" criteria. Essentially, this randomization is determined by: pitching grade, score, inning, opponent (strong or weak), part of the batting order faced (e.g., heavy or weak parts of the lineup), relief appearances, and total relief appearances of the team.

For example, when facing a strong opponent (winning percentage of .550 or more), in a tie game with any of the top four hitters coming up, Clarke will likely use his team's best reliever. If, however, the opponent was a weak one (under .450 winning percentage) with the other conditions identical, Clarke will likely choose the second best reliever. By choosing relievers based on this series of changing circumstances, you should see relievers compiling relief appearances close to their historic totals. Without  this randomization, however, you would see two or three relievers garnering excessively high relief stints and resulting in a highly inaccurate replay.   

 Because a number of statistics for the era may be incomplete or missing, Clarke will calculate artificially some of this data. This is most important with stolen base numbers, specifically caught steals, which for many seasons may be absent. Using the player's steal rating, Clarke calculates total steal attempts - a figure that is vital in determining base stealing - based on the following formula, as devised by Bill Staffa:

SB/(StealRating-2)/0.02778

This formula should give you extremely accurate replay SB statistics.

Additionally, for some teams or season disks, sacrifice hit totals may be absent. In such cases, another formula has been devised that considers the strikeouts per at-bat (so player SO totals MUST be included) of the player plus hit hit run rating AND the position in the order that he is batting. Actual SH totals, as well, are adjusted downward by roughly 25% to offset any sacrifice flies that were defined as sacrifices before 1930 or so.

 Some other strategies to note include:

1) Clarke does have a limited substitution strategy for blowout or lopsided games.

2) His offensive substitution pattern is, as was the style, very conservative and generally only used late when behind or for non-regular type players in extra inning games

3) He will use pitchers as pinchhitters and will generally only use players who did pinchhit during the season as pinchhitters (this is defined as players with more offensivegames than defensive ones).

4) Clarke adjusts all of his pitchings strategies based on the type of staff he is handling. A key ingredient in this is the total relief appearances of the bullpen or staffhe is handling, with teams having relatively high totals leading to a quicker hook than lower ones. Note: teams with total relief appearances equal to or under 15 will see Clarke bringing in a new reliever ONLY if the current  pitcher has been injured or ejected. Otherwise, he will go nine innings even if permitting 20+ runs. He will, however, pinchhit for the pitcher in the ninth or later if losing.

5) Clarke does not - and cannot - take advantage of either the bunt for hit strategy or the corners in strategy. He does, however, a limited one out "semi" bunt for hit approach that will only be called with baserunners on.  Other replay managers for this era of interest include:  Ned Jennings, Nixey O'Brien, Hughie Collins