Analyzing the Buchter-McFarland bullpen shuffle

The Oakland Athletics signed left-handed reliever T.J. McFarland to a one year deal earlier this month after claiming him off waivers from the Arizona Diamondbacks in early November. The same day, the A’s non-tendered left-handed reliever Ryan Buchter, who had been working out of the Oakland bullpen since 2018, primarily on short assignments and lefty-on-lefty matchups.

The move essentially constitutes a payroll-neutral bullpen roster substitution, with McFarland’s $1.8 million contract approximating what Buchter was expected to receive. So, why did the A’s pick one of them over the other? Let’s zoom in on their stats for a more extensive consideration of what might be gained—or lost—in the shuffle. First their career numbers:

Buchter, career MLB: 257 G 214 IP 2.86 ERA 4.01 FIP 4.74 xFIP 1.8 fWAR

McFarland, career MLB: 262 G 380.1 IP 4.07 ERA 4.03 FIP 4.02 xFIP 1.1 fWAR

Their most recent season:

Buchter, 2019 OAK: 64 G 45.1 IP 2.98 ERA 4.96 FIP 5.08 xFIP -0.1 fWAR

McFarland, 2019 ARI: 51 G 56 IP 4.82 ERA 4.48 FIP 4.33 xFIP 0.0 fWAR

And their 2018:

Buchter, 2018 OAK: 54 G 39.1 IP 2.75 ERA 3.54 FIP 4.53 xFIP 0.6 fWAR

McFarland, 2018 ARI: 47 G 72 IP 2.00 ERA 3.63 FIP 3.78 xFIP 1.1 fWAR

The most prominent difference between the career lines is that Buchter’s ERA is considerably lower than McFarland’s. Buchter has also put up more fWAR in about the same number of appearances. One could fairly easily take that into consideration along with his much higher strikeout rate and his superior pitch velocity and quickly conclude that a rational front office should favor Buchter.

However, when we control for team defense and ballpark conditions by comparing xFIP, McFarland looks considerably more favorable. Buchter’s escalating ERA/FIP/xFIP series suggests he is actually a below-average run preventer benefiting from a pitcher-friendly park and excellent defense. McFarland is a bit above average across the series, at least according to post-2015 standards: MLB average ERA, FIP, and xFIP were 4.51 last season and have been above 4.00 since 2016.

Furthermore, McFarland’s 2018 line is unquestionably the best of the bunch, and while his 4.82 ERA in 2019 is by far the worst in the set, his xFIP was much better than Buchter’s the same season. If you have confidence that FIP and xFIP effectively abstract externalities in evaluating a pitcher’s performance, then this analysis serves as a pretty good example of how misleading an indicator ERA can be on its own.

Beyond those big-picture stats, here are three advantages that McFarland might have over Butchter.

LOOGYs and the three-batter rule

Buchter has primarily been used as a LOOGY, and it remains to be seen how he’ll adapt under the three-batter minimum rule that takes effect next season. McFarland is capable of working in long relief and has averaged 1.45 innings per career MLB appearance while Buchter has averaged just 0.83. In terms of number of batters faced, McFarland’s career average is 6.37 and Buchter’s is 3.42.

Both of them have big platoon splits, but McFarland just hasn’t been used as a LOOGY. He made 141 appearances for Arizona over the last three seasons, and only 20 times did he face fewer than three batters. In those same three years, 51 of Buchter’s 189 outings came against fewer than three batters. Around 60% of McFarland’s batters were righty in those years, compared with around 51% of Buchter’s. During that span from 2017-19, McFarland’s FIP was 4.03 and Buchter’s was 4.40, even though McFarland was working longer outings and facing a lot more righties already.

McFarland fits in better with the league’s new rules, and also seems able to handle a bigger workload.

Walks

When relievers come into pressure situations and start pitching into scary counts and walking batters, matters can get out of hand very quickly. Putting a welcome mat at first base doesn’t preserve leads well and can potentially wear out one’s welcome. Granted, Jake Diekman was re-signed for two years and $7.5 million after posting 7.07 BB/9 with the A’s last season, but we’ll ruminate on that elsewhere.

Buchter has a career walk rate of 4.04 BB/9, which is substantially worse than league average. McFarland’s rate is 3.03 BB/9, which is average or a little bit better. Buchter’s walks were especially bad in 2019 (4.57 BB/9), and maybe it wasn’t a complete coincidence that his blown saves increased from zero to five the same year his walks went up.

Grounders

Ultimately, these decisions may well have hinged on the ground game. Are Forst and Crew pinning their hopes on McFarland’s astronomical ground ball rate to play well with the support of the A’s star-studded Gold Glove infield?

Setting aside Marco Estrada’s small sample, Buchter’s 25.2% grounder rate was the lowest among A’s pitchers in 2019. His career mark is 26.7%, so last season was hardly an outlier in this respect. McFarland, on the other hand, has a career ground ball percentage of 63.0%, which is a good deal higher than Brett Anderson’s team leading mark of 54.5% for the A’s in 2019. It’s also one of the highest in the entire majors.

Anderson himself is a good example of a pitch-to-contact grounder guy succeeding on this Oakland team. He kept balls on the ground with solid results for the A’s last season, finishing with an ERA of 3.89 and a FIP of 4.57. Lots of factors go into beating his FIP like that, but one of them is an excellent infield scooping up the abundance of ground balls he induced. It would make sense that McFarland’s top-tier ground ball rate would fit in well in the same way.

Like McFarland, Anderson also strikes out far fewer batters than today’s average big league pitcher. Last season, Anderson posted just 4.60 K/9. Given the similar “get lots of ground balls and let your defense get the outs” profile shared by these two, it’s easy to imagine that Anderson’s success last season served as a model of what to expect from McFarland. Perhaps the Oakland A’s are the perfect team to nurture the best possible results from their new bullpen lefty. If this proves to be the case and McFarland repeats his 2018 level of success, he will be quite a steal indeed.

But is a pitcher with such a low K/9 well suited for high leverage relief assignments? McFarland’s lesser ability to get the big strikeout in an inherited jam might make him better suited to mop up duty—or at least taking over in calm conditions—than to putting out fires.

So does the move make sense? It isn’t a terribly exciting one, but it seems like the A’s may have deftly struck an opportunistic bargain, or at least made a relatively low stakes wager with more potential upside than downside. Adding McFarland alone won’t solve the A’s bullpen problems, but he appears to be a better fit for Oakland than Buchter would have been in 2020. He shouldn’t be affected by the new rules, he can keep the walks at a reasonable rate, and his ground ball percentage is exemplary. You never know with relievers so it remains to be seen which one will have the better year, but Oakland could be a good fit for McFarland’s particular skillset.

What do you think?