Filed under:

# Win percentage vs. BaseRuns win percentage

Which teams differed the most and least in the two metrics?

There is a significant amount of variance in baseball. There is what actually happened on the field, sequence and luck and all, and there is the true talent level. Sometimes, they're pretty much the same; sometimes, they're not. It is interesting to compare the two.

For those unfamiliar, this page on FanGraphs supplied the data to which I will refer throughout this article. The first section of that page shows the actual standings in 2014 along with some stuff about runs scored and runs allowed. The second section shows the win-loss record for each team as predicted by PythagenPat, which might be considered a "first order" win-loss record. PythagenPat works purely off runs scored and runs allowed to predict a team's win-loss record. Taken from Tom Tango's website:

W% = RS^x/(RS^x + RA^x)

Where (from Baseball Prospectus)

x = ((RS + RA)/G)^.285

This formula tries to strip out the effects of luck and sequencing to arrive at a win percentage more representative of a team's true talent level. However, this formula uses the actual runs scored and allowed to arrive at a predicted win percentage.

There is a third section on the FanGraphs page, which is the BaseRuns section. BaseRuns are considered "second order" wins. BaseRuns attempts to predict the number of runs scored and allowed based on the component offensive statistics of a team (walks, hits, etc). BaseRuns works to get closer to a team's true talent than PythagenPat.

Below is a Tableau viz to show the difference between a team's actual win percentage and BaseRun win percentage. The visual is a slopegraph; teams who have a lower win percentage predicted by BaseRuns compared to their actual win percentage will have a downward sloping line and vice versa. I included color to designate the magnitude of the difference; darker colors represent larger changes. Color is not necessarily associated with a negative or positive difference. Using the BR filter can show only those teams with a positive or negative difference between the two win percentages.

There are multiple filters and interactions to play with; the graph in the bottom half of the Tableau viz shows the runs scored and runs allowed compared against the BaseRuns scored and BaseRuns allowed. You can see what's driving the difference between the BaseRuns win percentage and the actual win percentage.

I also noted the team with the highest difference and the lowest difference. With all teams present, the Royals had the largest positive difference between their actual win percentage and BaseRuns win percentage. The Rockies had the largest negative difference between their actual win percentage and BaseRuns win percentage. The Royals won 8 more games than predicted by BaseRuns, while the Rockies lost 11 more games than predicted.

Diving further, small differences in the Royals' predicted BaseRuns scored and allowed are driving the difference. The Royals perhaps scored a little more than their talent would suggest, and they allowed fewer runs than their talent would suggest. The Royals had an elite defense and bullpen, which probably explains the majority of the difference between the actual runs allowed and predicted runs allowed. Having an elite bullpen seems to confuse projections a bit. On the offensive side, the Royals emphasize contact over all else, which means they rely on sequencing to produce runs. They got lucky. There were stretches in which the Royals' offense was unbeatable; there were stretches where fans wondered if the team would ever score a run.

For the Rockies, there wasn't really much of a difference between the actual runs allowed and the BaseRuns allowed. The difference was on the offensive side; BaseRuns predicted more runs than were actually scored. It is known that the Rockies face unique environmental circumstances, but it is unclear to me if their home park had any effect here.

Though the Royals may have played over their heads, I don't think any Royals fan really cared during their playoff run.

. . .

All statistics courtesy of FanGraphs.

Kevin Ruprecht is an Editor of Beyond the Box Score. He also writes at Royals Review. You can follow him on Twitter at @KevinRuprecht.