When is a Walk Rate Too High?

Today’s post is brought to you by Vince Velasquez who is schedule to face the Marlins. When I decide which pitchers to target, I look for who is facing some of the weaker offenses and Velasquez popped for being interesting (3.68 ERA, 11.4 K/9) and available. I noticed his 15% K% (5.8 BB/9) and moved on. He broke my simple rule of not rostering a player with a walk rate over 10% BB%. But is 10% the right value? After diving into the numbers, it’s not a bad rule, but I’ll try to shrink that number down.

I use 10% because it’s simple to remember. The equivalent is 3.9 BB/9 which could be rounded up to an even 4.0 BB/9. The deal is that while simple rules help, I’d like to have every advantage. For example, I used to use a 50% GB% to show an advantage. The deal is that between 50% and 55% GB, the pitcher doesn’t gain any advantage in double plays and additional high flyballs (i.e. outs).

The other reason I picked 10% is that only a few pitchers reach this total. Of the 68 qualified starters this season, nine have a walk rate of 10% or higher.

The first step was to decide on my data set. I used only starters who had no relief appearances. I know this cost me a few guys, but I really wanted true Strikeout and Win numbers per start. Also, I used pitchers from the past 10 seasons who made at least five starts. I wanted to remove the guys who got called up for spot start or two. Finally, I focused the talent spectrum by limiting my sample to pitchers with a strikeout rate between 7.5 K/9 and 10.5 K/9. In all, 541 pitchers are included.

An additional step is that I wanted the value to be in BB/9 because is more frequently used across various websites. Finally, I divided up the pitchers into 0.5 BB/9 intervals and here are the results.

The first key is that the K/9 stays relatively constant with a range of 0.3 K/9. All the other stats change as expected and do so linearly.  I expected some stats to deteriorate as the walk rate increased like the Win% (for not making it five innigs) and ERA. It was just a steady decline.

To set a reference for the WHIP and ERA values, here this season’s median values for the NFBC Main Event (15 teams) and Online Championship (12 teams).

The corresponding walk rate to the above WHIP and ERA values is in the 2.0 BB/9 to 3.0 BB/B range. This range is the middle ground and with some top-tier starters rostered, this middle range can be fudged higher.

The best reasoning for increasing the values would be that the starter is facing a weak opponent, like with Miami in the original example. For more details on how to value matchups, check out this recent article. To expect acceptable results from a subpar starter, he has to be facing a team with wRC+ at or under 80.

Focusing on the original table, I will lower my acceptable walk threshold a bit. I went and average the 20 values around a 3.5 BB/0 (9.1% BB%) and they work out to 1.32 WHIP and 4.02 ERA. Both seem like reasonable thresholds for the above two league teams. The limit would need to be lowered even for a 10-team mixed league or increased for an AL or NL-only league.

Going back to my thoughts on starting Vince Velazquez, I’m fine with not considering him, especially since his walk rate is 50% higher than my limit. Of the qualified pitchers, nine are over the 3.5 BB/9 limit**. This group is still being allowed to start because they’ve been a bit lucky (.256 BABIP) with an average ERA under 4.00. For an example of the expected regression, I took the staters from the original sample with a 3.5 or higher BB/9 and sub-4.00 ERA in the season’s first half. They saw their ERA jump 0.61 points in the second half. And these were the pitchers whose team allowed them to keep starting.

While a walk rate over 3.5 BB/9 doesn’t mean I’ll immediately remove a starter from consideration, it’s a huge red flag that needs to be overcome. A combination of a higher groundball rate (>55%), a strikeout rate over 10 K/9, and a weak opposing lineup and starting pitcher could allow me to reconsider my stance. While I no longer have a nice round 10% BB% to use, I’d prefer to have a better spot in the standings as buckle down on the starters I’m willing to roster each week.

* German Marquez, Austin Gomber, Lance McCullers Jr., Ian Anderson, Erick Fedde, Jose Urena, Lucas Giolito, Freddy Peralta, Casey Mize, Taijuan Walker, and Mike Minor.

Interesting note on these pitchers, three of them (Gomber, McCullers, and Urena) have a groundball rate over 55% (my threshold mentioned earlier). Also, Marquez (54%) and Mize (53%) are almost there. A high groundball can help mask other flaws.