Was Rostering Ranger Suárez the Correct Decision?

This past Sunday, there were just a few available middle relievers who seemed to be better additions than some starter, so my co-owner Fred Zinkie settled on Ranger Suárez. Right now, Ranger Suárez has a 0.93 ERA, 0.72 WHIP, and 8.1 K/9 and is ranked as the 52nd pitcher according to our auction calculator. While his strikeout rate isn’t the highest, he does average 1.7 IP for each appearance and accumulates 1.5 K in each of those appearances. Also, he has been able to get 3 Wins so far this season (18% Win Rate).

The other option was Giovanny Gallegos who has also been great this season (2.53 ERA, 0.70 WHIP, and 10.3 K/9). He’s averaged throwing 1.2 IP per appearance with 1.4 K per appearance. He’s been able to get 5 Wins during that time (14% Win Rate).

Since the pitcher was supposed to take a starter’s spot, the hope was to get good ratios, some strikeouts, and possibly vulture a Win. After historically finding pitchers with these traits, our two choices were fine with a major flaw. We ignored recent usage.

To find middle-relievers to target, I took all the relievers (GS/G <= 0.05) and grouped them by Win% (W/G). I included pitchers from 2017 to 2019 who had at least 10 appearances. Here are their average stats.

Now, these stats can be massaged to get the desired number of candidates. My goal was to have around 10 choices and add that week’s best option.

I started with several other stats (including all the Leverage Index’s), but these few had distinct dropoffs and reasons to be used.

• Team Win% >= 45%: The team needs to be competitive to get a Win
• gmLI >= 1.1: The pitcher’s manager must have enough confidence to use the pitcher in High Leverage (i.e. close) games.
• IP/G >= 1.0: Has average more than one inning.
• K/G >= 1.1: Used to make sure some strikeouts are accumulated.
• Saves/G < 25%: To remove closers since they were already rostered.

While trying to back or forward test these values, I found changing reliever usage messed with the final criteria (reason for only using 2017 to 2019 data). Here are the number of pitchers who might the above requirements each season (min 10 games) since 2020.

Not one pitcher met the thresholds a couple of decades ago with strikeout rate being the major obstacle. I’m sure the game will change in some way going forward and the above benchmarks will need to be massaged.

So, using these thresholds, here are the pitchers from 2020 and 2021 who met the requirement (min 5 G).

By per luck, both Suarez and Gallegos meet the statistical requirements. The one issue that wasn’t meant was their usage. Both threw on Saturday, and other middle-reliever should have been cheaply rostered as a fill-in.  The workload in the previous few games needs to be monitored.

Also, the upcoming schedule needs to be checked for maximum games. The two initial options had seven upcoming games with Gallegos having four of those games in Colorado (three runs allowed in his first Colorado appearance). Probably a better option would have been one of the Yankees (vs LAA, NYM) with seven upcoming games and they didn’t throw on Saturday or Sunday.

Finally, use some common sense. Just because a pitcher meets the requirements, make sure they are still being used in high-leverage situations. Also, make sure there hasn’t been any talent degradation (e.g. velocity loss).

And that’s how I’ll start finding the pitcher subset when I’m wanting to add a middle reliever. The filters help to find the talent and high-leverage usage. The past usage and potential future usage are the only other variables to consider.