PECOTA projections peg 2020 Mariners for 66-96, 2nd-worst record in MLB

The first major projection system to release their expected standings is PECOTA, from Baseball Prospectus, which projects the Mariners to go 66-96, finishing last in the AL West, with the second-worst record in all of baseball. The system takes into account their player projections, playing time expectations, injury concerns, and each individual team’s schedule in the projection, putting out a range of likeliest outcomes over a thousand simulations, a practice for using an exceedingly high number of statistical tests to minimize randomness known as the “Monte Carlo method”. This means, while the single record above is a “likeliest” outcome, there were a range of results, which can vary from team to team. For the Mariners and the rest of the AL West, it looks like this:

Baseball Prospectus

The 66-96 number, therefore, is important, but it is there specifically because the Mariners most commonly were right there. They do, however, have a slightly wider range of outcomes, as you can see from the slight dip and flattening near the top on the right side of the mound. I don’t know if it grants all that much solace that Seattle also has a reasonable set of periods with low-70s wins in addition to mid-60s, but it’s a level of variance befitting a young club like this. PECOTA’s projections also extrapolate the record permutations into what chance a team has at a playoff spot and division title. The Mariners are listed at 0.0% for both.

The following paragraph goes into a bit more depth on the history and methodology of PECOTA, which has changed slightly this year, but if you’d like to skip it to get to more of the results, feel free.

PECOTA (Player Empirical Comparison and Optimization Test Algorithm) was created for Baseball Prospectus by Nate Silver, who now leads FiveThirtyEight, in the winter preceding the 2003 season. Baseball Prospectus (BP) continues to publish the PECOTA rankings, which Silver published himself through the 2009 season, and per their most recent update the algorithm itself is now an entirely new engine, free of any of Silver’s touches. PECOTA is a proprietary system, so how its process compares to other systems like Marcel, Steamer, and ZiPS is more a generality than a fully trackable comparison of equations, but it is venerable and well-regarded. Notably, PECOTA now fully integrates with BP’s “Deserved” metrics, like Deserved Runs Against (DRA) for pitchers and Deserved Runs Created (DRC+) for hitters. Those metrics take into account several variables, like quality of contact/quality of contact allowed, that are either unaddressed or intentionally left out of Fielding Independent Pitching (FIP) or FanGraphs’ Weighted Runs Created (wRC+), which we lean on often here. Naturally, that makes for discrepancies in projections and attribution of merit, some of which stand out dramatically on Seattle’s roster.

If all those acronyms hit you in the face like a wall of bricks, don’t sweat it, but also as always consider reading our Sabermetrics 101 series, listed above, which we hope to add to this spring once again. PECOTA’s projection is full of interesting tidbits for both the Mariners and league-wide. Of note...

• Seattle is expected to be better than the Orioles, but both the Tigers and Royals eke ahead at 69-93 and 68-94 respectively. Neither club’s roster is particularly impressive, but a more balanced AL Central, and a schedule involving playing one another often, makes a difference.
• The Angels have done enough to get the edge on the second Wild Card spot race, making the playoffs in just over 45% of the simulations. The Athletics are just two games behind Anaheim (87-75) at 85-77, chasing the Rays who have the lead on the first Wild Card spot.
• The Astros (98), Dodgers (103), and Yankees (99) are expected to run away with their divisions, and the Twins are the only other club projected for over 90 wins (93).
• Potentially flipping the trend of the past couple seasons, the AL is projected for nine teams with >.500 records, with only the M’s, O’s, Royals, Tigers, Rangers, and Blue Jays falling short.
• The NL, conversely, projects for a mollywhopping in the NL West, and just L.A., Cincinnati, Chicago, New York, Washington, and Atlanta coming out over .500. The records may be misleading, however, as 11 of the NL’s teams have at least 15% playoff odds, versus just nine at that threshold (and none of those above 3%) in the AL.
• Moreover, while there should be better records in the AL on the whole, both Houston and New York win their divisions in 88-89% of simulations, with Minnesota taking the Central in 77.4% of simulations. Maybe cheating fallout dogs the ‘Stros, but PECOTA is skeptical we see any major drama in the AL divisional races this year.
• While the Reds(!) are the projected NL Central winners, they don’t win the division in the majority of simulations, per Harry Pavlidis. That division should be tight yet again.

• PECOTA sees the Rays and Angels as slight favorites over Cleveland for the Wild Card spots, with Oakland and Boston trailing a few steps behind. By their projections, selling off Corey Kluber is the difference between playoff favorite and playoff contender, with the same holding true in greater measure for the Mookie Betts/David Price deal.
• The Cubs have talked a lot about trading Kris Bryant, which could put them in a similar situation, as PECOTA pegs them as narrow favorites for a playoff spot, with plenty of hope for a division title. That script changes significantly without Bryant, and might push the Cardinals and Brewers up significantly.
• The Phillies project for comparable playoff odds and an identical record to the Rockies, which will assuredly ruffle some feathers. vitally, Philly at least has a pathway to a division title, with the Mets a hair ahead of the Nationals at the moment, as well as what could likely be a highly contested NL Wild Card race once again.

The standings projections are based upon the player projections, which can be found and downloaded here for those with a BP subscription. The spreadsheet is both proprietary and gory, so I won’t share it in its entirety here, but a few notes that matter include...

• Mitch Haniger is the team’s best player, but only he and Kyle Seager are projected for 2+ WARP seasons.
• If you had Carl Edwards Jr. as the best projected pitcher on the Mariners, congratulations! He’s marked for 0.86 WARP, which should begin to explain some of the rest of the issues PECOTA sees with this club.
• DRA (and subsequently, PECOTA) do not think Marco Gonzales has something special going on. WARP graded him as an exactly replacement-level (0.0) player despite 200 innings in 2019, and expects about the same in 2020.
• Understandably, PECOTA doesn’t know what to do with Austin Nola, and projected him mostly as a 1B in 2020. That means minor league free agent Brian O’Keefe is currently projected as the backup catcher, which is unlikely to drastically impact expectations but is certainly worse.
• PECOTA doesn’t think Logan Gilbert is necessarily ready, but he is one of the team’s best present pitching options.