Due to yesterday’s screwup with lineups I needed to make it easier to parse and present current roster value. Usually value of lineups does not change that much from day to day but yesterday it changed significantly throwing off the entire analysis. Had this been fed into an algorithm to predict a probability that probability would have been wrong.
The purpose of this blog is to be a prototype for an app that provides a window into this data model using the Cubs team as an example. With the app you could do this analysis with any team at any time. We choose weekly team status because not much changes in baseball from day to day. But if you wanted to see the current status of say the Minnesota Twins or their status on September 4, 1996 it’s a simple lookup. We’re working on the presentation using this blog.
Most baseball sites throw table upon table at you with so many numbers no one can make sense of it all. Much of it is utter nonsense (i.e. WPA, OPS, etc. etc.) which we may discuss later. This model discerns value into a single weighting factor and a rank amongst the entire league, both pitchers and batters. The weighting factor has a solid proof behind it, something that very few baseball stats can claim.
That said let’s take a look at the Reds’ entire 12 man hitting roster to see what happened yesterday.
This team has some great hitters and a couple of terrible hitters. By replacing Barnhart and Peraza with Cozart and Mesoraco, -5 value came off the books and around +1 was added meaning the lineup added more than +6 to the group’s value which is significant. Normally teams don’t have such extremes. A team can improve on current value just by getting rid of a negative performer. Sometimes they can’t either due to contract issues or that they need someone to play shortstop. Here is the value of Cubs’ hitters for comparison.
These values include yesterday’s game. The Cubs do not have extremes at both ends like the Reds even though their overall hitting value is lower.
Teams in playoff contention trade to upgrade their teams and teams not in contention trade to upgrade their minor leagues. The roster value can differ significantly from their BAT and PITCH runs above average in the team status line.
Below is a table listing all 30 MLB teams sorted by their real WAA=W-L value listed in the last column. All numbers below are total WAAs for hitters, pitchers, starting pitchers, relief staff, and total for the entire team. Theoretically if you added up all team totals it should add to zero. This model adds to zero but not all players listed in the 2017 player database are still playing. Bad players get washed out so the sum of the league will be well above zero for those still playing.
Highlighted in tan are the three teams in contention for NL Central and their total WAA value according to every player on the 25 man roster. The Cubs are +7, SLN is +4 so the Cubs are (7-4)/2 = 1.5 games in first place right now. The Cardinals have been improving their team and have a higher total player value on paper according to this data model. Both the Cardinals and Cubs have better pitching than hitting. From this table you can see which teams will probably end up in the NLCS and ALCS.
Also note the poor White Sox who got rid of all their good players have the worst total roster value in the entire league. They also almost have the worst record as well.
That is all for now. Had to write some scripts to compile the roster data but that’s very important at the end of the season and especially when we get into the playoffs. Tomorrow State of Tennessee Smokies and the top AA players from all 30 affiliates. Until then….