Just as in the halcyon days of the Buck Showalter era, when preseason forecasts were locally mocked and dismissed for being too light on the Orioles’ win total, those pesky projection systems are at it again.
Only this time there’s no summary judgment to make on the validity of data analysis and statistical modeling — the Orioles we saw win 101 games last season were built on the principal that such forecasts should be adhered to above all else.
So what, if any, solace can be taken when any forecast for the Orioles features a meaningful drop in wins from last year’s magical run? As the Orioles found out last year, the playoffs can be random and volatile. And, as long as these projections put them in the postseason, the club is starting in a good place.
First, a primer on how these work. By modeling future player performance based on how those who performed similarly in the past to a current player, these systems run millions of projections to create a range of outcomes for how the season might go, and they provide something of a midpoint for public consumption.
Baseball Prospectus’ PECOTA forecast has the Orioles’ average record in its simulations at 86.7 wins against 75.3 losses — which while still good enough for a playoff spot would put them third in the division behind the Yankees and Blue Jays, and basically 14 games worse than last year.
At FanGraphs, the Orioles are forecast to win 85 games, with a 57.5% chance of making the playoffs — both fifth best in the AL. The ZIPS forecast on the site is a bit more favorable, projecting a 90-win Orioles team atop the AL East and with an overall 74.9% chance of making the playoffs. (The writer, Baltimore’s own Dan Szymborski, believes this is because his system values the team’s offensive depth.)
Read More
That these projections are notably higher than last year’s for the Orioles — PECOTA’s average win total for the 2023 Orioles was 74 — means the obvious. The team and talent base have improved significantly, and even stripping out the year-to-year volatility that can make a team drastically outperform or underperform these forecasts, there’s a pretty high floor for this team.
The reasons the numbers are lower than you might expect for a team that won 101 games and is mostly back intact are relatively consistent with years past, though. Even with Corbin Burnes’ addition and some steps forward on the mound from their inexperienced rotation last year, the repeatability of that for a full year isn’t guaranteed in any projection. Neither is the quality of the bullpen, with Félix Bautista out and Craig Kimbrel replacing him as closer. Nor is the team’s success in winning close and late — it was 30-16 in one-run games last year and won 48 games from trailing positions. Those are hard to replicate, and yet the projections show, even with regression in any of those categories, this is a very good team.
If none of this matters to you because you let what you see on the field tell the story, fair play. The reason any of it should matter to an Orioles fan is that, unlike those mid-2010s glory days, there’s a much more meaningful relationship between these forecasts and what the Orioles actually do this season than back then.
That’s because, as we know, they rely heavily on statistical modeling and forecasting for pretty much everything they do — roster construction through free agency and trades, evaluating internal and external talent, making draft picks and more. The Orioles’ platform has different outputs than the public-facing models, with heavier or lighter weighting given to aspects that the club finds more or less important, allowing it to produce the best forecasts for what it values. But you hear often these days from people inside the game about how there’s more alignment than you would expect between analytical clubs’ valuations and projections, so it’s fair to assume the same can be said about public and proprietary ones, to some degree.
That means the Orioles likely view themselves as a good team and, as we saw last week with the Burnes trade, they’re willing to take advantage of opportunities to enhance that. This stands in stark contrast to the 2022 trade deadline when, in line with public models, general manager Mike Elias said it wasn’t a probability the team would make the playoffs and subtracted instead of added.
It’s also a bit different than last winter, when the Orioles made modest short-term free agent deals but didn’t execute a major move to upgrade the team. Throughout spring training last year, the buzz around the club was that expectations were too high from a fan base that saw the 2022 team improbably post a winning record but was overlooking the fact that most projections had it as a sub-.500 team for 2023. As such, the front office didn’t exactly lean hard into upgrading the club.
It’s safe to say things are different now. Elias said so himself when explaining the Burnes trade in the context of the Orioles’ processes, noting that they made decisions in the early years of the rebuild geared toward enhancing playoff possibilities for years in the future and that this version of the team warrants those enhancements now.
The projections we have now back that up. The Orioles are at least a playoff team in many of them, and if the front office — presumably bolstered by midseason with additional resources from the new ownership that’s incoming — is operating on similar forecasts, this is a team they’ll fortify to get to the postseason and prepare for better October results than 2023.
So, yes, the projections released this week are a bit light considering local expectations for the team. We already know what matters most from them: that this has the potential to be a playoff team yet again, and a front office that believes in data likely has plenty to motivate it to support this team’s ambitions because of that.
Comments
Welcome to The Banner's subscriber-only commenting community. Please review our community guidelines.