The drop from 2016 to 2017 for most of the lineup looks to be expected regression. I wouldn't bet on any "positive" regression from 2017 to 2018.
This list shows expected home run rate (derived from statcast data).
Mookie - 41.4% (actually dropped from 31 to 24)
Bogaerts - 49.5% (21 to 10)
Pedroia - 33.9% (15 to 7)
Bradley - 56.3% (26 to 17)
Hanley - 54.6% (30 to 23)
Young - 34.8% (9 to 7)
None of those drops are larger than the percentages would indicate. It looks like the team simply enjoyed a bit of home run luck in 2016. So if there's going to be internal improvement, it will have to come from a combination of Hyers's approach and some good luck.
Is the bolded true? I'm not very well-versed in Statcast data and this is my first read of that article, but he states that the AvgExHR% of all the players he included was 59.1%. Obviously, all 6 of our guys listed were below that, but I'm not sure how to use those numbers to try to predict a drop in HRs from year to year. It surely isn't a direct translation of expected HR total from 2016 to 2017. Otherwise, we would have been expecting 2017 MLB HR totals to be 60% of 2016's. My understanding is that the AvgExHR% is for those balls that have already been hit by the player, as compared to similar balls across MLB. It tells us how many HRs a player "should" have at a given point. In order to project it out, you'd have to assume the player (and MLB) continue to hit balls the same way. This is possibly useful in the short term or maybe within a single season (i.e. his Max Kepler example), but there are way too many variables across multiple seasons (I think) - changes in health, mechanics, ballpark, slumps, pitchers' approach - that would affect that batted ball data. It may tell us whether/how much a player has been lucky in his HRs, but I think it would have to be compared to the 2017 data in order to see if/how much the change in HR total is due to that luck. For example, if Mookie's AvgExHR% in 2017 was 75% while hitting far fewer HRs in roughly the same PA, what would that tell us? The 2016 data tells us he probably shouldn't have hit 31 HR, but it doesn't tell us he should hit more or less in 2017. That will be based off how he hit the ball in 2017. Maybe the AvgExHR% is pretty stable year to year, but I doubt it.
Just as an exercise for curiosity...The top ten players on that list -- % change in HR total from 2016 to 2017 -- % change in PA from 2016 to 2017 (to give an idea of how that HR total would be expected to change based solely on playing time):
Park ------ +16.7% -- +86.5% (all in AAA)
O'Brien -- +200% -- +468.7% (AA/AAA)
Souza ---- +76.5% -- +31.8%
Cruz ------ -9.3% ---- -3.3%
Davis ----- -31.6% -- -19.8%
Cabrera -- -57.9% -- -22.1%
McCann -- +8.3% --- +6.7%
Garcia ---- +50.0% -- +23.8%
Flowers --- +50.0% -- +13.8%
Carter ----- -58.5% --- -43.8%
So, 3 guys who hit significantly more HRs than I'd expect given the change in playing time, 5 who hit significantly fewer, and 2 who were about in line.
The bottom 10 (not including Pedroia and Young):
Hardy ------- -44.4% -- -31.7%
Escobar ---- +40.0% -- -31.4%
Ellsbury ---- -22.2% -- -33.4%
Cabrera ---- +21.4% -- +3.1%
Hill ----------- -80.0% -- -73.7% (AA/AAA/MLB)
Giavotella -- -16.7% -- +6.0% (AAA)
Gregorius -- +30.0% -- -1.3%
Brown ------- -80.0% -- +15.2%
Kepler ------- +11.8% -- +27.1%
Barnhart ------ 0.0% -- +0.7%
4 guys who hit more, 4 who hit fewer, and 2 in line.
And our guys:
Betts ------- -22.6% -- -2.5%
Bogaerts -- -52.4% -- -11.7%
Pedroia ---- -53.3% -- -33.7%
Bradley ---- -34.6% -- -14.9%
Ramirez --- -23.3% -- -10.8%
Young ------ -22.2% -- -21.6%
5 guys who hit fewer, and 1 who was in line (and from whom we also aren't hoping for improvement next year).
I realize this is back of the napkin and SSS, but if there's useful predictive data there, I'm not seeing it. But, if there is, 1) I think it would lend to the argument that we could expect our guys to bounce back in 2018, wouldn't it? and 2) fuck the yankees.