By: Chris Weikel
Old school baseball thinking and team building almost always dramatize the necessity of the “Team Captain” in an effort to lead young players, create a clubhouse culture and develop team cohesion. These “Glue Guys” have always been plentiful in baseball, but their actual impact on team success has been almost impossible to identify or quantify. With the new landscape in baseball — in which every team has embraced analytics and various forms of performance evaluation — teams have to work much harder to create an edge on the field. That is where “Glue Guys” come in. They are said to create an advantage outside the box score that many analytics-heavy teams will fail to pick up on. This could create a point of differentiation for any team that finds a way to quantify these players and could be a huge first step toward analyzing the impact of clubhouse culture on teams. I have conducted initial research to measure the impact of these players. While the analysis is full of noise, it provides a broad view of their effectiveness.
The first step of the research was choosing 10 players who played in the last 30 years and are known around the league for their leadership and teamwork skills. On top of this, the players must have played for at least three teams, to try and separate the impact the player made on the team rather than the entire franchise culture as well as obtaining a diverse sample. The list finally settled on were these ten players: David Ross, Adrian Beltre, Dexter Fowler, Hunter Pence, Carlos Beltran, Curtis Granderson, Jonny Gomes, Jim Thome, Ivan Rodriguez and Jeff Francouer. This is a diverse group featuring players with different positions, quality of teams played for, and amount of time played which should help to even out what will always be a slightly skewed sample group.
On the most basic level, the two things that “Glue Guys” are supposed to do are make players around them better and help their team win close games. For this research the effort to quantify this skill was achieved through this method; looking at each time one of these players changed teams and measure the average difference in WAR for all players who had at least 300 PA in both the previous season and the first season with the “Glue Guy”. Below is the graphical representation of the average impact each player made on his new teammates.
The results are rather erratic, which is expected with each player’s sample only being around 25 players. But some players so seem to have an impact on player development. Carlos Beltran, for example, brought growth to all 5 of his new teams. All but 6 of his 26-player sample from his almost 20-year career experienced a better season once he was brought aboard. There is a massive amount of noise, which is impossible to avoid when attempting to measure intangibles. But when players gain on average almost one WAR across five different teams Carlos Beltran does appear to be at least a tiny bit responsible for this gain. On the flipside, Hunter Pence is linked to a .74 WAR loss for his new teammates. His case does show the flaws of working in this small sample size as he was tanked by both the aging 2012 Phillies roster he joined as well as the “Odd Year-Even Year” shenanigans of the 2013 Giants team.
While using this data at the individual player level is interesting, the sample size is too small to have any significant findings. But if we look at the sum of all the players in the sample, we do gain enough of a sample to have significant data. When all players’ samples are combined into one dataset, each player on average gained .15 WAR just for having one of the players join their team. While on an individual level .15 WAR isn’t a massive amount, when amassed from all position players on the roster, a team in theory could add around 2 extra WAR to its roster just for having one of these players. This, of course, does not directly translate because this is a very simple and noise-infused study. But on the most basic level, it illustrates that “Glue Guys” may have at least a slight impact on the surrounding players’ WAR.
The other facet of the value supposedly added by “Glue Guys” is the leadership skills that help a team scrape together more wins than they would have. This studies way of measuring this was by examining every team one of these players played for and taking the difference between their actual record and their Pythagorean record. For those of you that don’t know, Pythagorean record is a Bill James-created method for determining how many wins a team should have gotten based on its runs scored and runs allowed. When you analyse the difference between this record and actual record, it creates a makeshift clutch/close game wins-added that in this research will be used to try to quantify the intangibles brought by the Glue Guys. Here is the chart:
Again this chart is very erratic due to the noise and small sample size but when the players are summed together, it does show an increase in wins over Pythagorean wins. The average increase each year is only .134 wins but when measuring such an erratic and not well-understood skill this could be significant.
The overall goal of this study was not to create a perfect way to measure impact of “Glue Guys,” but to instead determine on the most basic level whether they actually achieve any of the things that they claim to do. While both the data and the way I measured it have flaws, it does seem that these particular players did appear to make the players around them slightly better as well as help their teams win extra games. This could have been due to randomness of the data or the impact of many other variables, but in this study these Glue Guys did appear to make an ever-so-slight difference. This test is in no way definitive, but is an initial effort toward quantifying the unquantifiable.