As we near the halfway point of the season, I thought it would be useful to look at the departures from projections of individual players and what it implies for the team's record the rest of this year and next year. I'm going to put the departures from projections into 3 broad categories: 1) those that are "real" and can be expected to continue going forward, 2) those that represent good or bad breaks (you can book those but should not expect them to carry forward), and 3) noise.
I won't say much about category 3. It involves players out-performing or under-performing but not in a major way. In rare cases, all of the deviations in this category break in the same direction and cause the team's record to deviate from projections. I don't think this has happened with the Braves this year. Rather the important deviations have been in the first two categories.
Category 1 is the most important one because those are the kinds of developments that have more long-lasting implications. I think within this category there are three players who have performed in a way that would make me raise their projected production going forward (Freddie Freeman, Inciarte and Flowers) and one that would make me lower their projected production (Teheran). The net though is fairly significant, about 5 wins between those four. So if you think this was a 75-80 win team at the start of the season, with the changes in expected production of those four you would have to think this is now a 80-85 win team, assuming everything else is about equal with the rest of the roster.
Category 2 are the good and bad breaks that you can book but should not anticipate carrying forward. There have been two big ones in the first half. One was literally a break--of Freddie's wrist, which should have cost us about 3 wins over the 8-10 weeks he was out. But this has been largely offset by the acquisition of Matt Adams and his hot streak.
It could turn out that Adams' hot streak is more than a hot streak. The longer it continues the more likely it is something that should be classified as a Category 1 deviation from projections, something "real." I'm not quite ready to go there. It is instructive to compare the improvement we've seen in Adams and Flowers. With Adams a big chunk in the improvement is from ISO. With Flowers it is from the strikeout rate. An improvement in strikeout rate for the kinds of sample sizes we are talking about is more likely to be real or sustainable. The other difference is that the improvement in strikeout rate from Flowers has been going on for a bit longer than the improvement in ISO from Adams.
I don't mean to imply that there is a black and white distinction between what we've seen from Adams and Flowers this year. But I think there is enough of a difference to say that one is more likely to carry forward to a significant extent over the next year.
