PECOTA, among the first public baseball projection systems to grace the internet, is somewhat of a modern-day Ship of Theseus: still harbored at Baseball Prospectus, but sharing little to none of the original equations whirring within Nate Silver’s initial design. The system has proprietary functionality, as do ZiPS, Steamer, and most other public systems, yet BP’s primary projection system and Dan Szymborski’s ZiPS (via FanGraphs) are thus far the only two projection systems to which the general public has full access. PECOTA’s player projections are available to paid BP subscribers, so I won’t be spoiling the entirety of their work; however, their early unofficial standings highlight some interesting facets worth exploring.
PECOTA simulates the 2022 season a myriad of times, offering us ranges of outcomes from worst case to best case, collected in increments of 10 percentiles, as well as the 1st percentile (the worst) and 99th percentile (the best). The “base” projection then is the 50th percentile outcome for a player and/or club, but the range varies from player to player and team to team. For the Seattle Mariners, they presently shake out around .500, with an 81.2-80.8 record in PECOTA’s simulation as MLB’s lockout drags on. Lagging well behind Houston (96.9 wins), and a game or two behind the Angels and Athletics, they’re far from a preseason favorite. A few signings could shift that expectation, but for now they are merely peering in through the window on the playoff hunt, and certainly the divisional chase. A few notable projections could also easily be made or broken.
Thank goodness for Robbie Ray. Seattle’s rotation may return to a traditional five-slot setup in 2022, but PECOTA is wary of Seattle’s rotation no matter if they throw four, five, six, or twelve folks into a starting role. Both Ray and Logan Gilbert grade out as slightly above-average pitchers by BP and PECOTA’s core pitching evaluation metric, Deserved Run Average (DRA), or more specifically, DRA-. You can read more about the metric here, but the gist is that it works much like ERA- or wRC+ by creating 100 as “average”, and each number away from 100 is a percent better or worse than “average”. In the case of DRA-, lower is better (hence the minus). So how do Seattle’s projected potential starters grade out, including their Wins Above Replacement Player (WARP)?
Robbie Ray - 172.0 IP, 3.63 ERA, 4.42 DRA, 95 DRA-, 1.9 WARP
Chris Flexen - 169.0 IP, 4.50 ERA, 5.40 DRA, 114 DRA-, -0.1 WARP
Marco Gonzales - 160.3 IP, 4.10 ERA, 5.02 DRA, 108 DRA-, 0.7 WARP
Logan Gilbert - 137.2 IP, 3.51 ERA, 4.38 DRA, 94 DRA-, 1.6 WARP
Justus Sheffield - 79.1 IP, 4.94 ERA, 5.50 DRA, 118 DRA-, -0.1 WARP
Justin Dunn - 63.3 IP, 5.10 ERA, 5.50 DRA, 118 DRA-, -0.1 WARP
Matt Brash - 51.0 IP - 4.49 ERA, 4.97 DRA, 107 DRA-, 0.2 WARP
George Kirby - 32.1 IP, 4.15 ERA, 4.87 DRA, 105 DRA-, 0.2 WARP
This concludes the list of pitchers projected to start for Seattle, leaving the M’s with an uphill battle in terms of finding effective innings if things bear out this way. League average ERA in 2021 was 4.27, so Seattle’s favorable home confines seem likely to assist them a bit (as may a quality defensive group), but a great deal of skepticism appears targeted for Chris Flexen, in particular. Seattle’s innings leader in 2021 ran a fabulous 3.61 ERA last year in 179.2 frames, making all 31 starts asked of him and generally stabilizing a group that spent much of the year suffering bullpen games or abbreviated outings. Yet Flexen was dinged last year by DRA, earning a 5.21 DRA and 112 DRA-. Flexen’s contact management style worked out well, but PECOTA appears wary he can repeat his league-leading 22 double plays turned behind him to dance out of trouble. Of course, it also accounts for his previous struggles, so if we see Flexen as a genuinely shifted true talent starter compared to his pre-Korean Baseball Organization self, there’s reason for more optimism. All the same, another strong rotation arm should be a high priority for Seattle to make themselves a more serious playoff contender.
Layers of Decency
Whereas the rotation is easy to conceptualize, the Mariners’ current lineup is extremely difficult to parse or project. Currently, things expect to shake out thusly: J.P. Crawford at shortstop, Mitch Haniger in right field and at DH, Ty France at first base, some combination of Abraham Toro and Adam Frazier between second and third base, and Jarred Kelenic in either center or left field. All six players are projected to be at least average players. Cal Raleigh and Tom Murphy are likely to be the primary backstop duo and they too collectively project to be an average tandem.
So what’s the issue?
Seattle’s strength and struggle is that almost every position has a reasonable shot at average production, but few have a high likelihood of stardom. That’s where Seattle making a free agent addition or two (say, Trevor Story and/or Kris Bryant and/or Seiya Suzuki) would do wonders. Unfortunately, because Seattle has so many ~okay players and not many clear standouts or cesspools, it’s a bit more challenging to firmly upgrade any one position with absolute confidence. Seiya Suzuki is almost certainly better than Jake Fraley, but is he better than a healthy Kyle Lewis? Abraham Toro is not exhilarating, but how much of an upgrade over the Québécois Quoif do Bryant or Story offer? Likely a significant one, on the whole, as depth matters, and improved baseline expectations matter, but it is a challenge Seattle must navigate, or perhaps consolidate via trade.
PECOTA puts Jerry Dipoto’s feet to the fire. If the M’s see significant overlap with their internal evaluation and PECOTA, they’ll see 21 year old Julio Rodríguez, nary a game above Double-A to his name, as one of the M’s best hitters day one. PECOTA projects Rodríguez for a 108 Deserved Runs Created Plus (DRC+, BP’s wRC+ equivalent with contact quality metrics factored in for predictive purposes), trailing only Ty France’s 115 DRC+ projection for anyone in the organization. There’s something to be cautioned about this with Seattle’s mostly-average lineup, but it is mostly a staggering credit to what the system sees in the precocious youth.
By WARP, the projection is initially middling for the 50th percentile—just 0.3 WARP—but that is in merely 17 games, or 63 plate appearances. Stretch that to nearly a full season of around 600 PAs and you’re looking at a 2.9 WARP projection. Another way of putting it is that PECOTA thinks he may already be the best position player on the Mariners roster. His close player comps are blissfully tantalizing - rookie season Giancarlo Stanton, Eloy Jiménez, and Domingo Santana, with a .267/.339/.423 projected line that feels simultaneously tantalizing and tough to truly believe until we see him thriving at T-Mobile Park.
Often “projection system” is misread as “prediction system,” which over-simplifies and under-sells the work done by these models. Just as the ZiPS standings from earlier this month remind us of the work the Mariners have yet to do, so too is the case here. What these models can do, though, is help fans focus their off-season wish-list targets. Obviously, the lineup has question marks, but if we’re not all singing along to Big Daddy Kane and Biz Markie rapping the intro to “Me and Julio Down at the Schoolyard” on Mariners Vision like Boston fans screech “Sweet Caroline” every seventh inning by July, something has gone terribly amiss. While the Mariners delivered the PS5 of Robbie Ray this off-season, they need to supplement that gift with Lee Carvallo’s Putting Challenge in the form of a mid-level signing to bolster the rotation. A Tyler Anderson three-wood, anyone?