The 90th Percentile Cubs

Last week, the PECOTA projections were released and while few would have expected another hundred-win season projected for the Cubs, the 91 wins they’re currently sitting at on the BP projected standings page is underwhelming to say the least.

To make it appear even more like PECOTA has something specifically against the Cubs, the Dodgers are projected for a major league-leading 99 wins, a remarkably aggressive number for the system. Zack Moser already broke down where the system forecasts a step back for the team, and while conservative, there aren’t many outright egregious projections. Kyle Hendricks is perhaps unsurprisingly underrated; Addison Russell’s glove isn’t given the credit it deserves.

In any case, I’m not here to suggest what PECOTA might be wrong about. Instead, I’m going to look at the system’s view of the Cubs through the most rose-tinted glasses. Everyone sees the win totals and the projections for each individual player, but these are just based on the mean projection, the 50th percentile outcome as determined by the system. If a reader was to pay a visit to a Baseball Prospectus player card, they would in fact find nine separate lines of PECOTA projections, ranging from the depressing tenth percentile to the glorious ninetieth percentile. PECOTA considers that there is a non-zero, albeit small chance that players will reach this level of production. Let’s get greedy and find out what the Cubs look like if they all hit their 90th percentile at once. We’ll start with the hitters, sorted by WARP:

Kris Bryant 652 105 32 3 39 114 82 167 11 4 0.303 0.404 0.580 0.341 70.9 7.9
Anthony Rizzo 699 105 37 2 36 114 81 113 9 5 0.297 0.394 0.550 0.329 56.1 6.4
Kyle Schwarber 653 109 25 3 38 102 91 166 5 3 0.284 0.389 0.547 0.324 63.3 6.3
Jason Heyward 650 87 32 3 20 78 71 103 17 5 0.286 0.370 0.457 0.297 37.8 5.9
Ben Zobrist 686 94 37 4 17 74 92 89 8 5 0.289 0.386 0.452 0.301 51.3 5.5
Addison Russell 644 81 33 2 25 89 57 148 6 3 0.273 0.346 0.468 0.288 39.6 4.8
Willson Contreras 568 74 30 3 20 80 56 108 4 3 0.304 0.384 0.500 0.314 51 4.6
Javier Baez 489 70 22 1 22 67 30 136 16 5 0.277 0.332 0.481 0.287 28 3.5
Jon Jay 485 54 21 2 6 47 39 80 5 3 0.298 0.370 0.403 0.283 25.9 2.8
Albert Almora 270 31 15 2 7 32 11 43 3 1 0.294 0.324 0.456 0.280 13.3 1.6

Hardly unreasonable, for the most part, which is sort of the point. This group is collectively projected to score 810 runs, two more than 2016’s 808. That also doesn’t account for the fact that if every player hit their 90th percentile projection, there would be a lot more run scoring. Any players projected for more limited playing time, such as Tommy La Stella and Matt Szczur, have also been left out, so the overall total would be higher than this, quite conceivably over 900. No one would be surprised to see this Cubs team score more than 800 runs again; 900-plus would be spectacular.

For Rizzo and Bryant, this isn’t really asking much other than to repeat what they’ve been doing. It’s a slight step forward for Bryant in terms of walk rate and power, while Rizzo’s projected home run increase of four is a combination of slightly increased plate appearances and some fly balls going over the fence.

Schwarber’s is the line that jumps out the most. His 50th percentile projection is .248/.346/.477, with 31 home runs. That still represents a fairly optimistic outlook for a player who still doesn’t have a half-season of major league playing time, yet it pales in comparison to this, which is essentially Rizzo from last season with even more home runs. To give some idea of how Schwarber’s limited major league sample and profile makes his outlook volatile, his 10th percentile projection has him putting up a .252 TAv, with 24 home runs, 72 points of TAv and 14 home runs below the top of the range. Rizzo’s is .275 with 27, a difference of 54 and 9 respectively.

What a boost it would be for Heyward to get back to being a six-win player. This line in combination seems like the one of the most unlikely outcomes. We have seen Heyward hit more than 20 home runs, but only when he also struck out over 150 times in 2012. Given his efforts to get back to his swing of that year, an uptick in power could be forthcoming, but to expect it with a near .290 average is extravagant. FRAA hasn’t been included here because PECOTA doesn’t really project a range of fielding outcomes in the same way, but despite rating poorly by the metric last year, Heyward is projected to go right back to being a plus-15 defender. That seems like the one projection that we can be confident in.

Zobrist’s numbers look like normal Zobrist with a touch of batted ball luck. Russell’s line is exactly like what we might expect from a highly talented 23-year-old continuing to make strides. Russell clearly doesn’t get enough credit for his glove from FRAA which prevents him from being more like a six win player here, but both DRS and UZR back up the scouting and rate him as one of the league’s best defenders. Twenty home runs and a solid walk rate are eminently possible for Contreras. The .300 average takes a little more stretching, although it’s not so far away from 2016’s .284.

Combined, this group is worth close to 50 WARP. BP’s replacement level for WARP is currently set at such a point that a team of replacement level players would win just over 50 games. Just adding all these totals together on top of the replacement level to get a win total isn’t the most mathematically sound approach, not least because for all of these players to hit their 90th percentile plate appearance totals, some others would likely have to not do so. Nonetheless, it’s clear that even with a replacement level rotation, this is a lineup that should comfortably win the division.

What about the rotation? Here are the 90th percentile numbers for the six starters projected for the bulk of innings:

Jon Lester 14.7 8 0 29 29 221 164 56 228 22 0.256 1 2.73 49.2 5.3
Jake Arrieta 14.5 8.2 0 30 30 219.2 149 69 222 19 0.237 0.99 2.77 47.6 5.2
John Lackey 13.6 8.6 0 29 29 213 164 58 196 24 0.248 1.04 3.18 36.8 4
Kyle Hendricks 11.8 7.7 0 29 29 183.2 138 46 163 17 0.247 1 2.95 35.8 3.9
Mike Montgomery 8 6 0 36 19 138.2 108 44 112 12 0.250 1.1 3.25 21.1 2.3
Brett Anderson 4.8 3.7 0 13 13 87.3 77 24 64 9 0.263 1.15 3.44 11.1 1.2

The first notable aspect is the continuation of the remarkably low BABIP across the board. Whether the Chicago defense is good enough to sustain this, and indeed whether the team is doing something in particular to keep BABIP down, remains to be seen. There are some reasons to think it can continue for at least some of the staff. Arrieta has two straight years of a sub-.250 BABIP. Hendricks is a master of soft contact, leading the league in inducing weakly-hit balls.

As Zack noted in his PECOTA piece, Hendricks’ 90th percentile projection is still worse than his actual performance was in 2016. He’s projected for more walks and fewer strikeouts, with a dip in innings and a decline of more than a win. It’s unreasonable to expect the same level of performance from the 27-year-old, and yet at the high end of what should be considered possible, it also seems unreasonable that he’s given a line that’s worse than either of his last two actual MLB seasons.

Lester has pitched at essentially the same level as his 90th percentile over the last two years and has topped five WARP in five different seasons. Arrieta was even better than this in 2014 and 2015, but given his drop-off in 2016, this seems more appropriate as a ceiling than it would have done a year ago. Lackey is capable of pitching 213 innings, a total that would give him around 4 wins if he repeated his 2016 performance otherwise. However, at the age of 38, with a manager who’s not afraid to go to the bullpen in the fifth or sixth inning, this seems like one of the most unlikely scenarios here.

It’s impossible to separate Anderson from his injuries. 1.2 WARP might be too conservative for 87 innings of a fully healthy Anderson. It could also look like a rather extravagant innings total when he’s struggled his way through another injury-filled campaign. Montgomery as a starter is still a relative unknown. He’s flashed the skills to succeed in the role but the Cubs would still be delighted with an outcome like this from a fifth starter.

The rotation provides just under 22 wins at the 90th percentile. Add a moderate five WARP for the bullpen and the staff comes out with 27 WARP overall. That’s four wins higher than they put up in 2016, which certainly is getting greedy, but then that was the point of the exercise.

While it’s fun to imagine a Cubs team that’s worth 77 WARP, the team obviously won’t win 127 games, an outlandish total even considering their 115-win second-order total of 2016. It’s not wrong to look at the projections for Bryant or Lester and think that they can considerably outperform them because they’ve done so in the past, and PECOTA considers it a possibility too. Expecting it to happen is something else altogether. It’s a perspective that doesn’t take in the downside.

PECOTA does both of these things all the time. It can be a good reminder that things can go wrong, that skills fluctuate, that luck plays more of a role than we might like to admit. There are things that projection systems don’t know that we do – like Kyle Hendricks being the second coming of Greg Maddux – but they also prove more accurate than human predictions in many cases by taking a balanced view.

All things considered, 91 wins seems too low; many would say much too low. Remember there is an version of PECOTA that agrees, in which Jason Heyward and Kyle Schwarber are six-win players and the whole Cubs rotation repeats its record BABIP performance. It just isn’t the most likely version. Bring on the 127-win season.

Lead photo courtesy Mark J. Rebilas—USA Today Sports

Related Articles

2 comments on “The 90th Percentile Cubs”


Darius, this is really excellent. And from a Giants fan… in the UK! I think Hendricks has completely broken PECOTA. And Heyward isn’t going to hit .290 with twenty dongs. But it is generally fun to look at the 90% projection on that offense, as ridiculous as it is.

Darius Austin

Thanks Atticus! Yeah, it is definitely ridiculous – I thought it might be particularly fun given the composition of this team. I certainly don’t think PECOTA is equipped to deal with the extremes like Hendricks. While I expect regression, this is a fairly unsatisfactory projection for the 90th percentile.

Leave a comment

Use your Baseball Prospectus username