Projections: Twins will improve by 20 games; Buxton will be their best player

Baseball Prospectus released its MLB-wide PECOTA projections for players Tuesday — basically, the projected stats various players will have in 2017 based on their career arc and historical data.

If you want to get really fancy, PECOTA is an acronym for Player Empirical Comparison and Optimization Test Algorithm. But you already knew that.

Much of this data — understandably — is paid content. BP puts out a book every year (foreword this season by none other than Twins pitcher Glen Perkins), which would make for nice reading during these dreary days. You can also subscribe to the web site (for a 17 percent discount).

For now, though, editor in chief Aaron Gleeman was kind enough to share the Twins' individual projections with me — info that, along with the free information on the site that gives us projections for all 30 teams' records, tells us a lot about what we could expect from 2017.

The number that immediately jumps out: 79.

Baseball Prospectus projects the Twins will win 79 games this season (against 83 losses). While that wouldn't be a "good" year in many senses, it would be a dramatic improvement from last year's Twins-worst 59-103 record — and more in line with getting back on the path the Twins thought they were on after their 83-79 season in 2015.

That record, per BP's projections, would be good for second in the AL Central. That should tell you the projections think the division will be quite bad — and indeed, Cleveland (92 wins) is slated to be the runaway winner. Rounding out the rest, after the Twins: Detroit at 78-84, Chicago at 77-85 and Kansas City at 71-91.

Now: even the best projections using the best data can be wrong. PECOTA has rather infamously whiffed on projections for the Royals in previous years. Trying to gauge how many wins a team will wind up with before injuries, slumps and other things take hold is a tough task. For now, though, it is very interesting to see the relative faith placed in the Twins.

To figure out why, let's look at a few key individuals — keeping in mind that the numbers aren't necessarily reflective of anticipated playing time but rather how those players would perform if given that time:

*The PECOTA projections see a clear progression for Byron Buxton, who checks in with an anticipated 17 home runs and a .730 OPS in 509 plate appearances in 2017. He's projected for 2.9 wins above replacement player — best on the team and indicative of the fact that he could very well be the Twins' best player as soon as this season. Miguel Sano is next with 31 homers and a 2.6 WARP. It's pretty safe to say the Twins would love it if those two were their best players in 2017.

Brian Dozier is next with 25 homers and a 2.2 WARP. Fourth in WARP? New catcher Jason Castro (15 homers, 1.8 WARP). Because you asked: Joe Mauer is projected to hit .275 with nine homers and a .360 OBP.

One has to imagine Buxton and Castro — the former having tremendous range in center and the latter coming in with a reputation for being a great pitch framer — helped fuel the projection that the Twins will finish 22.7 fielding runs above average in 2017. That's the second-highest projection of any American League team and the third-best in the majors. That would be quite the upgrade from 2016, when FanGraphs had the Twins finishing at minus-45.7 in defensive runs above average — second-worst in the majors.

Indeed, if the Twins improve by nearly 69 runs from last year to this year solely from their defense, that will account for a significant share of their projected improvement from 59 to 79 wins. Last year's Twins scored 722 runs and allowed 889. The projection this year from BP is 734 scored but only 752 allowed.

*The PECOTA projections see starting pitchers Tyler Duffey, Ervin Santana and Kyle Gibson as basically the same — all with 1.2 projected wins above replacement, all with ERAs between 4 and 4.3. That would mean a regression from Santana from 2016 to 2017 and a step back forward for Duffey and Gibson, who both struggled last season. Projections are also much higher on Jose Berrios (4.31 ERA) than Hector Santiago (5.06 ERA). It's safe to say the Baseball Prospectus numbers call into question the Twins' decision to keep Santiago on a one-year, $8 million deal — projecting a regression from relative past success.

Long story short: PECOTA projections see improvement from the Twins' young core, middling (but close to adequate) starting pitching and a significant defensive improvement. And those things add up to a projected 79-win season.

After the way 2016 played out, getting anywhere near .500 would be an accomplishment and a pleasant surprise.