Chicago Cubs: A statistical analysis of the National League Central

ST. LOUIS, MO - SEPTEMBER 27: Jon Jay #30 of the Chicago Cubs celebrates after winning the National League Central title against the St. Louis Cardinals at Busch Stadium on September 27, 2017 in St. Louis, Missouri. (Photo by Dilip Vishwanat/Getty Images)
ST. LOUIS, MO - SEPTEMBER 27: Jon Jay #30 of the Chicago Cubs celebrates after winning the National League Central title against the St. Louis Cardinals at Busch Stadium on September 27, 2017 in St. Louis, Missouri. (Photo by Dilip Vishwanat/Getty Images) /
facebooktwitterreddit
Prev
2 of 5
Next
Chicago Cubs
PHOENIX, AZ – AUGUST 11: Manager Joe Maddon #70 of the Chicago Cubs in the dugout during the MLB game against the Arizona Diamondbacks at Chase Field on August 11, 2017 in Phoenix, Arizona. (Photo by Christian Petersen/Getty Images) /

A consistent way to evaluate the teams

Before we can properly assess the teams in the National League Central, we must first set the ground rules. Hopefully, this leads to an objective assessment. In this particular analysis, we will look at each of six functional positional groupings. These include catcher, infield, outfield, starting pitcher, relief pitcher and closer.

Using a series of standard and sabermetric statistics, we can rank each NL Central team on a 1-to-5 scale. (Note: 5 being the best, 1 being the worst).

More from Cubbies Crib

Identifying the “right” analytical factors is where the magic really happens. Sure, it’s great to have a catcher with speed or a starting pitcher with a great eye at the plate, but neither trait better prepares the player to be a solid contributor at his position of choice.

There is no ‘I’ in team

Rather than focusing on overall, individual attributes – which are mired in subjective opinion and “beauty-in-the-eye-of-the-beholder” bias – this exercise takes a more objective approach. We first identify what defines “good” as it relates to a unique position group and then assess each team’s ability to satisfy that definition.

The result is a set of data that objectively frames a relative valuation of the six position groups. These results do not necessarily project that a given team excels at a given position. It is impossible to make such a determination from an analysis that did not include teams outside of the division. But it does clearly frame if Team A is better than Team B at Skill 1 and, in some cases, also frame how much better Team A may be at Skill 1.