The last week has delivered unto Braves fans a veritable cornucopia of prospect-related lists, rankings, and similar information, which has got to be at least somewhat exciting for Braves fans. Who among us doesn't enjoy hearing that their favorite MLB team has a smorgasbord of young, controllable, exciting talent in the pipeline? But, amidst all the future-focused excitement, another wave of releases today speaks to the brussels sprouts-esque 2016 season that Braves fans will need to endure before gulping down the delicious rebuilt dessert that awaits further down the line.
In case heavy-handed, food-related metaphors aren't your cup of tea, what I'm talking about are team projections for the 2016 season. In short order, we've seen releases from USA Today, PECOTA (courtesy of Baseball Prospectus), and one Nevada-based oddsmaker, and none of them are very sanguine about the Braves in 2016. (Of course, if you've been paying attention, that's hardly surprising.) Let's run through these one-by-one, but let's start with one set of projections that's already been available for a while.
The Fangraphs projections have been out for a while, and currently rely on Steamer projections along with Fangraphs staff input on the depth charts and projected playing time for each team's assortment of players. Eventually, these will be supplemented with ZiPS, but I don't think that's the case yet.
Steamer is not kind to the Braves, projecting the team to have the second least-valuable suite of position players in 2016, and the league's least-valuable pitching staff. Overall, only the Phillies are projected to be worse, and the Braves' projected 68-94 record is only a game better than where they finished last season. However, it's also a lot better than the Braves' Pythagorean record from 2015 (61-101), so at least there's some expectation of a sequencing-adjusted improvement from last year.
A few notes:
- The inclusion of ZiPS will potentially improve this projection a bit, as Steamer is generally more pessimistic about Braves' pitchers (mainly Julio Teheran) than ZiPS is.
- Currently, the Braves are projected to score 609 runs and give up 727 runs. The 609 runs is an improvement over their 2015 total (573), but not a large one, as even had the Braves scored 609 runs last season, they'd still be last in the majors in that regard. The pitching staff is projected to have a more substantial improvement, from last year's 760 runs given up (4th-worst in majors, 3rd-worst in NL) to 727 runs, which would've been only 11th-worst in the majors last season.
- Despite the central tendency of projection systems to want to lump most teams towards the middle as a "least wrong" estimate, the Braves, Phillies, and Brewers are the only teams projected to have fewer than 70 wins, and run differentials of below 100 runs over the course of the season by Fangraphs.
Now, your first thought when reading the above subheading is, "What is USA Today even doing, releasing projections?" And, well, I don't have much of an answer to that, and, to be fair, it looks like neither do they, as they're calling them "Semi-Scientific" predictions. This leads me to think, what exactly does semi-scientific actually mean? Is it kind of like semi-correct? Semi-chocolate? Semi-Sammy Sosa? I don't know.
Per the article in which these "projections" were unveiled, the above records are said to be derived from "equal doses of probability, pragmatism, and passion." I have no idea how you can add passion and pragmatism into something to derive a win-loss record, but hey, it's their national news magazine world and I'm just living in it.
The Braves are projected at 67-95, which, you know, was their record last season. So, clearly, the formula included an "If(team=Braves, wins_2016=wins_2015" or something. Looks like I figured out the trick. On a more serious and descriptive note, USA Today is actually comparatively bullish on the Braves by hating a bunch of other teams much more: 67-95 doesn't look quite so bad when the cellar is also filled with the Phillies, Brewers, Reds, Rockies, and Athletics, each of whom is projected to finish with fewer wins than the Braves. On the one hand, the idea that there may be six teams with sub-70 win records is probably a more reasonable assumption than the centrally-pulled Fangraphs projection; six teams had sub-70 win records in 2015. But, on the other hand, USA Today has the White Sox winning 90 games, so... yeah. (Also, the Giants ahead of the Dodgers? Huh?)
There aren't any other tidbits to go off of, short of a tiny blurb that says that USA Today expects the Braves to finish ahead of the Phillies "thanks largely to the track records of their position players." Still, even with those track records being what they are, the semi-scientific guesstimate has the Braves towards the bottom of the league. Even USA Today isn't cray-cray enough to suggest otherwise, I guess.
Another projection, another 2016 Braves win total in the same range. PECOTA, courtesy of Baseball Prospectus, has the Braves at 68-94 for 2016. So, a different projection system than that used by Fangraphs for player projections, aggregated up across players, and the same result.
There's not much that differs substantially here from the Fangraphs projection. PECOTA expects the Braves to score 602 runs and give up 721, which is within 10 runs of each total as currently projected by Fangraphs. As PECOTA is considerably more bullish on the Brewers (77 wins instead of 69 via Fangraphs/Steamer), the Braves and Phillies are left alone in the sad-because-we-have-fewer-than-70-wins club.
I've never been a stalwart fan of PECOTA as a system, but some of the team projections are certainly interesting. The Indians are 92 wins per PECOTA, and the Rays are at 91. And, of course, they have the Royals last in the AL Central at 76 wins, which will certainly continue to fan the very particular flames of "why do the projection systems hate the Royals arghhh." Those are some fun flames, let me tell you. (But, if you want to argue about the Royals, you should consider reading this great thing from Sam Miller, because it is great and a thing.)
No, Atlantis is not the name of a new projection system. It's a casino based in Reno, Nevada, that deigned to publish its over/under win totals for the MLB season.
There's a lot of interesting stuff there, and if you cross-reference the above with some of the Steamer/PECOTA projections, you can maybe see a strategy for maybe making some theoretical money, if you feel confident in the projection systems. Even if you get burned on one team, you should come out ahead on average, right?
The Braves, though, are in a pretty interesting 65-win split range. The goal of oddsmakers when doing these sorts of exercises is to find a place where the betting population is likely to be evenly split to some extent. To that end, I can see where they were going with a lot of these, but 65 seems perhaps somewhat low for the Braves. But, on the other hand, with the potential for a young pitching staff to bleed win after win, and another deadline selloff of any contributing major leaguers, there's a real chance for an ugly August and September that plunges the win total down. (Of course, you may also believe based on your own reasoning that the Braves are worse than a 65-win team, which makes the under a fine bet for you.) I'd take the over, though.
A few final notes:
- As I said on the podcast, I think the Braves will likely end up at 71 wins, but I also realize that even that prediction is bolstered by a thought that the Braves have to be able to kludge four wins (not Wins, but wins above replacement) out of their non-Teheran starters, because if they can't, then something has gone horribly wrong. (One-win starters aren't really that good, or that impressive.) If they can't, then I agree that 67 wins or so feels more natural, but I'm erring on the side of "some value from the rest of the rotation" rather than "no value."
- Projection systems have been getting worse over time for various reasons, nicely detailed here. There are a lot of reasons for this, including young players driving a lot of team production and being harder to project, more aggressive deadline trading due to the (stupidly) expanded playoff format, internal use of team analytics which may suggest different paradigms for amassing actual wins on the field (though I have no idea what these paradigms might be), and just good ol'-fashioned random variation leading to large performance variation spikes in certain situations. (If you figure that in 2014 and 2015, leverage performance splits for some teams heavily upset the projection apple cart, the overall trend in projection value from 1996 through 2013 looks pretty stable to me, and there's an outside chance that even in a stable system a 5% outcome would occur in consecutive data points.)
- From the above, you could argue that the Braves' projections have more "give" towards more wins, given the relative youth of their team and the fact that some young prospects could start contributing right away. On the other hand, the major league team as projected for 2016 really isn't all that young given some of the stopgap contributors being employed. Additionally, there's been some research that shows that the tendency to defy the central "smushing" of projections tends to exacerbate a team's distance from .500 by about 33% in the appropriate direction. (That is, if a team is projected to be an 84-win team, they may be more likely to garner 85 wins, i.e., 3 games over .500 x 33% = 1 additional win.) This is due to stuff like teams changing players at the trade deadlines, and depth spirals for bad teams who suffer injuries and have no suitable replacements, both of which reinforce the abilities of good teams to reap more wins at their expense. So while there may be a lot of upside "give" from 67 or 68 wins, I have to think there's some downside "give" as well due to the possibility of a sell-off or just a disaster of major league performance with no suitable replacements. I'll stick with 71 wins for now, given the absence of any other information.