Re-introducing 3WAR as a way to measure player value

No starting rotation in baseball provided more value as measured by Fangraphs’ fWAR since 2013 than the Cleveland Indians’. This isn’t due to a couple of years of total dominance; no matter which range you choose to measure within that time span — 2014 to 2016, or 2013 to 2015, for example — the statistic rates the Indians’ starters as at least the fourth most valuable in that range. fWAR may not be every fan’s favorite flavor of WAR, not is wins above replacement the only measure by which we can assess the value of players. If for some god-forsaken reason you still think pitcher wins are the best indicator of a starting pitcher’s worth, you can take comfort in the fact that the Tribe’s starters have accumulated the largest number in that category, too.

That doesn’t mean there aren’t other statistics that might tell a little bit of a different story. One thing that I believe we get a little bit fast-and-loose with in the infinite information age is which statistics we use to provide evidence for our arguments. If fWAR doesn’t support our viewpoint quite as well as Baseball Reference’s bWAR or Baseball Prospectus’s WARP, I think we’ve all been guilty of quietly tossing it aside and returning to it again only when it resumes being useful to our goals.

These publicly-available stats all try to measure the same thing: how many more wins was player X worth to a team than the type of player they could easily find to replace him? Despite that, we occasionally get wildly different answers from the different flavors of WAR. This is especially true when it comes to pitchers. Aaron Nola’s 2018 is a wonderful example.

fWAR: 5.6
bWAR: 10.5
WARP: 6.6

In other words, Aaron Nola pitched well in 2018. But how well? Depending on which metric you choose he was either a pretty good All-Star or had the best season of any pitcher since Randy Johnson in 2002...and the 13th best in the last 100 years.

For this reason I’ve referenced 3WAR in the past, which is just the average of the WAR total given by the three most popular data sources. Similar to a poll-of-polls in politics, the goal of this is to find a more accurate representation of a given value by averaging different answers. 3WAR suggests that Nola provided 7.5 wins above replacement. For comparison’s sake, Jacob DeGrom was worth 8.8 3WAR, and Max Scherzer was worth 7.9 3WAR.

Putting aside the question of why Baseball Reference thinks Nola had a legendary season, I think it is clear that the 3WAR gives us a more reasonable assessment of Nola when compared to his peers than any of the three flavors alone.

What does it say in that regard for the Indians’ 2018 starting pitchers?

The Chart

Before we get any further I want to note that there are some relief innings in here for Plutko and Tomlin. For our purposes at the moment that doesn’t really matter. Now to the fun!

• We might want to pump the brakes a little bit on the “Trevor Bauer was better than Corey Kluber in 2018” narrative. On a rate basis, yes, Bauer was the most valuable pitcher. In terms of overall value to the team, however, they end up with the exact same value. I don’t want to yell too much about whether or not the difference in innings matters in this case. On one hand, Bauer pitched better when he was in the game; on the other hand, there is definite value in “eating innings,” especially when the Indians’ bullpen struggled at times this season.
• Josh Tomlin did not have a good baseball season and may want to consider not having a baseball season at all in 2019.
• My impression of Adam Plutko’s 2018 is more in line with what Fangraphs and Baseball Reference report. Baseball Prospectus did not like him one bit. It’s difficult to draw many conclusions based on 70 IP mixed between the bullpen and the rotation, but it’s something too keep an eye on next season.
• I am surprised to see Clevinger close the gap between himself and Carrasco in 3WAR, and overall the top four are more equal in value than I expected to find.
• This is a great starting rotation regardless of how you slice the numbers. 5 WAR is roughly equivalent to All-Star value for starters and position players, and 3WAR suggests that the Indians essentially have four of them. There aren’t really any breaks for a hitter to look forward to when facing the Indians’ starters.

Next time: Projections

Now that we’ve seen what 3WAR says about the starting rotations 2018, I’d like to take the same approach to the various projection systems available and see what 2019 might look like. Assuming everything goes swimmingly I’d like to keep this up and do it for the bullpen and then the position players, too.