Fangraphs fans like their 81-win projection for the O's

[Frobby]

Based on individual players projections from ZiPS and Steamer, fangraphs projected the standings and came up with this for the AL a East:

BOS 93-69

TOR 86-76

TBR 82-80

BAL 81-81

NYY 81-81

Now they've asked their readers to vote on whether those projections are too high or too low.     So far, 38% think the O's projection is correct within one game, while 31% think it's too high and 31% think it's too low.    So on balance, the readers think it's just about right.    

For the other teams, a large number of readers think the Blue Jays and Red Sox projections are about right, but a majority of those who disagree think the Sox are a little low and the Jays a little high.   A majority of all voters think the Rays are too high, and the Yankees too low.    So, I think the "consensus" would be the Red Sox winning by a comfortable margin, the Jays and Yankees battling for 2nd place, the Orioles in 4th at .500 and Tampa in last.    

[Frobby]

Link: http://www.fangraphs.com/blogs/your-stance-on-the-team-projections-american-league-2/

[FlipTheBird]

A lot of FanGraphs readers are folks who totally buy into the FanGraphs approach to projections, so it makes sense. The only division they don't seem real keen on is the AL West, where they've firmly voted the Angels' and Athletics' projections as too high, and the Mariners' too low.

[crawjo]

It would be incredible if every team in the division finished at .500 or higher. Extremely unlikely that will happen. 

[Frobby]

Fangraphs has an article up now comparing their projections to the readers'.    The fan ratings of the O's were 2nd closest to fangraphs of all the teams, behind only San Francisco.   http://www.fangraphs.com/blogs/here-is-what-you-think-of-our-team-projections/

[fearthenoodle]

FWIW, that 81 win projection is based on a lot of assumptions we know (at least to a significant extent) to be wrong. That projection includes Mancini as the primary DH (and getting 420 PAs), Trumbo getting more PAs in RF than any other player (and only 420 total PAs for Smith as a result), 350 PAs for Flaherty, 105 for Dariel Alvarez, nothing in there for Bourn or Gentry at this point.

On the pitching side, they project 150 combined innings from Wilson/Wright, 75 relief innings from Ondrusek/Drake/Jason Garcia, Bleier to have the most relief innings behind the big 4 relievers (and more than Nuno), and nothing for Ynoa.

None of those are egregious mistakes (beyond maybe Mancini who doesn't appear to have a path to the roster unless there are injuries at this point), and ultimately a more accurate projection would only potentially move the needle a win or two at most, but nonetheless it was interesting to see how they project playing time right now.

[Phantom]

Projections are like referees. You only hear about them when they are wrong.

[Frobby]

27 minutes ago, fearthenoodle said:

FWIW, that 81 win projection is based on a lot of assumptions we know (at least to a significant extent) to be wrong. That projection includes Mancini as the primary DH (and getting 420 PAs), Trumbo getting more PAs in RF than any other player (and only 420 total PAs for Smith as a result), 350 PAs for Flaherty, 105 for Dariel Alvarez, nothing in there for Bourn or Gentry at this point.

On the pitching side, they project 150 combined innings from Wilson/Wright, 75 relief innings from Ondrusek/Drake/Jason Garcia, Bleier to have the most relief innings behind the big 4 relievers (and more than Nuno), and nothing for Ynoa.

None of those are egregious mistakes (beyond maybe Mancini who doesn't appear to have a path to the roster unless there are injuries at this point), and ultimately a more accurate projection would only potentially move the needle a win or two at most, but nonetheless it was interesting to see how they project playing time right now.

There's a lot of guesswork that goes into this.    Each year when I do my chart of the different Orioles OPS projections, I usually only list the top 13 guys or so and then have a PA allocation for "other."   The number of plate appearances that in fact get absorbed by "other" varies a lot from year to year, depending on injuries, performance and mid-season trades.   One thing fangraphs does is assume that everyone will miss some time -- e.g., they have Manny, Jones, Trumbo and Davis each playing 150 games.    If they're healthy, they'll all probably play more.    But rather than randomly picking one of them to get hurt and miss a significant number of games, they just give each guy a modest ding.   For some reason, they've got Schoop at 133 games, which is very unlikely barring injury.    They have Hardy at 133, which is a pretty good estimate.

[Camden_yardbird]

Projections will be very interesting this year as I believe the variation due to injury could be staggering.  I know a lot of people like Boston to be the best of the best, but that offense can't afford any major injuries or they might be trying to win a lot of 2-1 games.