Just How Good is Archie Bradley?

Fans are frustrated with Archie Bradley right now. A former top prospect, Archie never really found his groove as a starting pitcher and was eventually relegated to the bullpen. In 2017, he was very good as a reliever - a 1.73 ERA over 73 IP, good for 16th in WPA among relievers and 7th in WPA/LI (accounting for leverage).

However, things have only gotten worse for Bradley. As Jack Sommers recently noted, since July 10th, 2018 through today, here are Archie’s numbers:

46.1 IP, 5.83 ERA, 4.11 FIP, -2.85 WPA, 18 shutdowns against 14 meltdowns

Those are not good numbers for a reliever, let alone your setup man and/or closer. That -2.85 WPA is the single worst among 732 relievers over this time frame. That is BAD.

But at the same time, it is only 46.1 IP, so there are some small sample size concerns at play. The 4.11 FIP suggests some regression, but a 4.11 ERA and/or FIP is still not good for your setup man.

So, this begs the question: Just how good is Archie Bradley?”

Let’s start with the good: over that bad stretch above, Archie has an 11.46 K/9, which is considerably better than his 2017 season.

And the bad: Well, just about everything else.

There are two things that stand out for Archie over this prolonged “slump”: he is getting hit hard (.405 BABIP and 1.36 HR/9) and he’s really lost his command. A huge part about what made Archie so good in 2017 was that he had solid command. From the start of 2017 through July 9th, 2018, he had a 2.47 BB/9 or 7.0% BB%. Since then, his walks have jumped to a 3.69 BB/9 or 8.6% BB%. This might be easier to see visually:

Something happened at the end of 2017. His walks shot up and they’ve never really recovered.

Not only that, but this seems to roughly coincide with a decline in velocity over roughly the same time period:

That’s not a good sign. It seems like there might be something mechanically wrong with Bradley, given the drop in velo and command. This could be a warning sign for a looming injury or it could be something else. Time will tell.

But let’s get back to assessing Bradley. We’ve seen what has been a clear decline from his 2017 season. But what was the real Archie Bradley? Let’s start with the strikeout and walk rates and compare them to xK% and xBB%.

xK%: 26.1% (Actual: 26.3%)

xBB%: 8.0% (Actual: 7.5%)

These aren’t too far from his actual performance, so there isn’t really much hope for regression there. That’s not too far from his “slump” numbers, either (which are 26.7% K% and 8.6% BB%). This seems to suggest the problem lies within the batted ball data, which is supported by the spike in BABIP and HR/9 during this slump.

To analyze this, let’s look at Archie’s pitches. We’ll start with his fastball.

Archie’s fastball was fantastic in 2017. It had a run value of +1.84 runs per 100 pitches (wFB/C on Fangraphs), which was 12th best out of 150 MLB relievers in 2017 with at least 50 innings pitched. That’s really good.

However, the fastball has only gotten worse. The run value dropped to +0.77 in 2018 (most of that coming in the first half) and is now at a massive -2.58 runs per 100 pitches in 2019. This is now the 22nd-worst fastball among qualified relievers (220 relievers, so bottom 10 percent). This is not where your setup man is supposed to be. The bottom of this list is generally reserved for bad long relievers, like Matt Koch. And yet, here is Archie, where he seemingly doesn’t belong.

What in the world has happened to Archie’s fastball? As seen above, there’s been a drop in fastball velocity, but there’s also been a drop in his spin rate on the fastball:

You can see in 2017 that Archie’s fastball spin rate is very close to his spin rate as a starting pitcher - this is to be expected. However, the spin rate just drops off terribly in 2018 and 2019. Once again, this is suggesting either a mechanical problem or an injury to Archie.

There is some mysticism to spin rate, but the physics are undeniable: a drop in spin rate means a drop in pitch movement. Fastball spin rates correlate with increased whiff rates as the RPM increases; on the other side, fastball contact rates go up but so do ground ball rates as RPM decreases (so a good sinker baller like Brandon Webb would have really low spin rates on his fastball). A pitcher wants to be on either extreme; moving towards the middle makes it easier for batters to hit.

This is exactly what’s happening to Archie: in 2017, Archie was around the 82nd percentile mark for fastball spin rate for relievers. In 2019, he’s around the 51st percentile. In other words, he’s moved to the middle of the pack in terms of spin rate and when you pair that with the drop in velo as well as losing command, his fastball is becoming far less effective. And as a result, batters are now teeing off on his fastball:

His fastball is now being hit harder AND more regularly than when he was a starter... and he was not a good starter. To further validate the fall of his fastball, Statcast’s xwOBA on his fastball tells a pretty clear picture:

2017: .294 xwOBA on fastballs

2018: .310 xwOBA on fastballs

2019: .342 xwOBA on fastballs

We honestly don’t need to bother looking into his curveball right now. Without his fastball, his curveball will naturally lose its effectiveness and as a two-pitch pitcher, batters can just sit on his fastball, which he throws over two-thirds of the time. A fastball-curveball combination isn’t particularly strong as the two pitches don’t tunnel well, making it all the easier to avoid the curveball if you’re sitting on the fastball.

At this point, it’s pretty clear that the current Archie Bradley is not what he was in 2017. He was legitimately good then and he’s not good now. He is due to perform better than he has to date, though the “regression to the mean” is still below average for a setup man, at least in the current state.

It seems pretty obvious that something is wrong with Archie. It’s hard to tell if it’s mechanical or an injury but the team needs to do something about it immediately. Archie’s peripherals (strikeout and walk rates) are good enough to be a good setup man in the future if he can fix his fastball and cure his batted ball problem. But until then, it’s looking increasingly likely that Archie Bradley is no longer a good reliever, yet alone a good setup man.