Tyler O’Neill with normal luck

Much ink as been spilled about Tyler O’Neill in this offseason, and I apologize in advance, but you’ll have to suffer through at least one more. Because O’Neill has had a very strange start to his career. His career numbers of a 91 wRC+ with a .305 BABIP certainly suggest a much more normal career than he’s had. And it’s easy to see 1.9 fWAR in 450 PAs and conclude he’s probably pretty good!

But the way he’s gotten here is both encouraging and discouraging. His wRC+ has gotten worse each season he’s played in the majors, which is never a good sign. At the same time though, the player he was his rookie season was never going to be successful long-term. And for better and for worse, he’s not that player anymore.

So when looking over his stats - completely not intending to write another article about him mind you - I thought of an interesting premise: if you strip out luck, has Tyler O’Neill actually improved in his each season? This would be a strange phenomenon indeed. A player whose surface stats get worse each season because of luck, but who might possibly have improved each season despite this.

I won’t leave you in suspense because without even calculating the numbers, I can tell you with a certain degree of confidence that, no, he has not in fact improved each season. His 2019 season was nothing short of a disaster. There’s no way to see that as an improvement over his 2018 season, even if he made some necessary improvements to his game. But he was unquestionably a worse player in 2019, at least with what he showed on the field.

So interesting premise ruined, I am still curious about O’Neill’s stats when you strip them of luck. He has a career .305 BABIP, but he has never had a season with anything close to a .305 BABIP. So what if I gave him a .305 BABIP each season. How would that affect his stats? And what if I gave him his career HR/FB% instead of how each season actually turned out? And I’m going to do something special to his 2020 to try to account for how he got jobbed at the plate more than average.

First things first, it seems a little unfair to ignore that O’Neill did actually mash the ball in 2018 more than he did in the following two seasons. He had an average exit velocity of 92.1 mph compared to his career 89.6 mph. I don’t think an average exit velocity of 88 compared to a career 89.6 would necessarily indicate I should alter my BABIP rules but nearly 3 mph higher is simply impossible to ignore. So I’ll give him a slightly higher BABIP and a slightly higher HR/FB to account for that.

With 9 homers with 36 flyballs in 2018, I can either give O’Neill a higher number than I want for HR/FB% or a lower number than I want. He has a career 18.9 HR/FB%, so a 19.4% HR/FB% doesn’t seem like a big enough advantage to account for what seemed to be a guy who mashed the ball. But if I give him just one more homer, it’s all the way to 22.2%. I’m going to go with high though because what’s the point of accounting for exit velocity if you get a 0.4% bump. He had 66 balls in play, but I’m taking away a homer, and since it’s a flyball that doesn’t go over the wall, it’s probably an out. I originally calculated a .318 BABIP without accounting for the extra ball in play. If you add the out, his BABIP is now .313.

Without one homer, his ISO is .215 instead of .246. And taking away four hits puts his average from .254 to .223. Combining all this information together, O’Neill with relatively neutral luck would have hit .223/.275/.438. I’m going to take a shortcut for his wOBA and simply do OBP * 1.75 plus slugging which is not exactly accurate but definitely close enough for me to save a lot of time. His new wOBA is .306, down from .340. This moves his wRC+ in 2018 from a 116 to a 92 wRC+.

2019 is going to be rough, but there is one benefit - he had his career low HR/FB%. Another situation where more PAs would be much preferred, because I have another annoying decision to make. Giving him one more HR puts his HR/FB% at 17.1%, lower than I’d like. Giving him two though puts him at 20%. 2019 feels like a year where I should not give him the benefit of the doubt though, so I’ll go with the lower number. As for BABIP, well it was .386 that year. I’m removing a ball in play - it went for a homer - and with 82 balls in play, a .305 BABIP gives him nearly exactly 25 hits. That means I’m taking away six hits even with the added homer, and his average goes from .262 to .220. On the plus side, his ISO is now .178 instead of .149. HIs overall line is now .220/.272/.398.

His wOBA falls from .308 to .291. This causes his wRC+ to drop from a 91 to a 75 wRC+. Like I said above, there’s really no way to look at his 2019 as anything but a setback. Of course, he made a concerted effort to improve his plate discipline numbers - which he did - but it came at an expense. An expense which I think partially started to pay off in 2020.

So 2020. I’m going to try to account for O’Neill’s unfortunate tendency to get screwed by the home plate umpire. Ben Cerutti of Birds on the Black did the work for me. So I’ll let him do the talking here.

Furthermore, 21 of the 104 pitches that Tyler O’Neill took for strikes this year were (according to Baseball Savant data) outside of the strike zone. That’s just over 1 out of every 5 strikes that he took that were given to the pitcher. He was called out on strikes on 5 of those, another 4 of them were the first pitch of a plate appearance to take him to 0-1 instead of 1-0, and another 4 of them were on 1-1 counts making it a 1-2 count instead of a 2-1 count. Lastly, two of them took a walk away from Mr. O’Neill (one of them replacing it with a strikeout instead).

How does this compare to league average? Well he helps out there too. On average, 16.1% of strikes called should have been balls. Which means that somewhere between four and five pitches that were outside of the strike zone should have been balls if O’Neill were treated like the average player. I see really no reason why O’Neill would be a player who would specifically get treated worse by umpires, so I can’t look at this as anything but incredibly bad luck.

If I’m taking away four or five bad calls, how does that affect his numbers? Well I’ll try to be fair. I’m going to take away one bad call that resulted in a strikeout, one bad call from the first pitch strike, one from the 1-1 count, and one that would have resulted in a walk. The fifth is irrelevant. It’s one of the calls that doesn’t seemed to have affected the plate appearance. So we take away one strikeout and add a walk. The others don’t really affect my personal calculations - it might help him get a hit or a homer, but I’m doing that to his numbers anyway.

For HR/FB%, I’m going to give him a 20% HR/FB because it’s more interesting and because we’ll say more favorable calls for him lead to more favorable balls thrown to him. It’s just an extra homer. With BABIP, I’ve had to account for an extra homer and walk, and one less strikeout, and with these changes, his comes to 87 balls in play. Helpfully, I can give him a .299 BABIP with these changes. With eight extra hits, including one homer, his average jumps all the way to .245. With an extra homer, his ISO is now .215.

Where does this leave us? Well, his 2020 line - previously .173/.261/.360 - is now .245/.331/.460. However, I now feel like adding that extra homer did too much to his line to fit my neutral luck point of the post. I can already sense the O’Neill “realists” objecting to it, so I’ll play fair. I’m removing the homer. A hit gets taken away, in the form of a homer, and his BABIP falls to .295 and his ISO remains at his 2020 level of .187. So his line is now .237/.325/.424. His wOBA is now .331 instead of .271.

I think I might be a little off by taking the shortcut on wOBA here, but the most comparable player to that line is Maikel Franco, who hit .278/.321/.457, who had a .329 wOBA and a 106 wRC+. So hard to say with different ballparks and whatnot, but higher OBP, lower slugging - I think O’Neill’s line might be at about a 105 wRC+. It’s above average, whatever the case may be. So to recap:

2018 - 92 wRC+

2019 - 75 wRC+

2020 - 105 wRC+

O’Neill’s luck has roughly evened out over the course of his three years, which is why when you add those three numbers together, it comes out to a 91 wRC+. Which is literally his career line. But if O’Neill is coming off a 2020 season with a 105 wRC+, we’re all looking at the outfield situation quite a bit differently, no?

So despite the seemingly declining stats, I’d argue O’Neill is a vastly superior hitter the one we saw in 2019 and a better one that we saw in 2018. And I think it could be argued that 2019 was a very necessary growing pain he had to experience to improve overall. Because a hitter with 40% strikeouts and 5% walks will never be successful for very long. So I have to once again come to the conclusion that I am very optimistic about O’Neill’s future with the Cardinals.