As I write this, the Pittsburgh Pirates are tied for the second best record in baseball. They also happen to be tied for second place in their own division, because the Cardinals are the only team with a better record while the Reds have matched Pittsburgh’s 33-20 start, making the NL Central the most competitive and most interesting division in the sport right now. The Cardinals are Reds are both excellent teams, and we should expect both to continue to win at a good clip over the rest of the year, but what about the Pirates? Is this another first half mirage that will lead to a second half collapse, or do Pittsburgh fans finally have a contender to root for?
I think the answer to both of those questions is probably yes; the Pirates are playing over their heds and will likely regress over the next four months, but their strong start and their overall talent level should keep them in the race to the very end.
Let’s start with the playing-over-their-heads aspect of things. Pretty much any team that is on pace to win 101 games is probably having a few things go their way, so we could look at nearly any team at the top of the standings and say that they should be expected to play worse over the rest of the year. This is just how regression to the mean works. Find someone or something that is doing better than anyone else at that thing and suggest it won’t keep doing that thing as well as it has been; you’ll be right more often than not.
But, what we really care about isn’t whether or not the Pirates will regress, but how far they will regress. It’s the magnitude, not the direction, of the regression that really counts. So, how far over their heads have the Pirates been playing?
The most common way of answering that question is to look at pythagorean expected record, which judges a team based on runs scored and runs allowed rather than wins and losses. By RS/RA, the Pirates “should be” 30-23, not 33-20, as they’ve wracked up an extra three wins because of the timing of when they’ve scored and allowed their runs. However, I’m not really a fan of using pythagorean record as some kind of indicator of how many wins a team “should have had”, since it really only goes halfway in stripping out unsustainable sequencing. If we’re going to acknowledge that the timing of runs is mostly random, why not also acknowledge that the timing of the things that lead to runs are also mostly random, and use those individual events rather than the sequencing-included runs scored and runs allowed totals?
If we really want to strip out timing and just focus on the actual events that a team has been involved with, we’re better off going all the way down to the value of the individual plays, rather than stopping at RS/RA and deciding that the runs scored and allowed are a good measure of luck-free performance. And, perhaps the easiest way to sum up all the of events a team has been involved in without taking any sequencing into account is to just look at their wOBA differential. We did this last week with the Cubs, but let’s focus it on the Pirates this time.
(Run Differential is on a per game basis, by the way.)
|Team||Batting wOBA||Pitching wOBA||wOBA Differential||Run Differential||Winning %|
By run differential, the Pirates are hanging out with the A’s, Rockies, and Diamondbacks, and they rate #9 overall in MLB. And maybe that’s the group that it feels like, based on pre-season forecasts, they belong in. None of those teams were expected to make the playoffs based on most forecasts, and each one seems to be playing a bit over their heads at the moment.
By wOBA differential, the Pirates don’t move that much — jumping from #9 to #8 — but their company changes. Now, they’re in the mix with the Rays and Braves, teams that were expected to be contenders, and are generally seen to have playoff caliber rosters. This isn’t a case where the Pirates run differential overstates how many extra wins they’ve earned through timing, but I do think it’s helpful to know that, in terms of the plays the teams have been involved in, the Pirates have performed in a roughly similar manner to teams that everyone believes can keep on winning.
By either run differential or wOBA differential, the Pirates have played like a team that should win about 57% of their games, now 62%, so, again, regression is almost certainly coming. And, of course, we shouldn’t just regress a team’s winning percentage in two months back to their underlying performance over the first two months of the season, since even things like wOBA are subject to sample size issues. For instance, the Pirates lead the league in wOBA allowed at .288, but a large part of that is based on holding opponents to a .265 BABIP.
Even if we think that the Pirates terrific defense is a big reason why they’re turning so many balls in play into outs, that’s the kind of number that is less likely to be sustained over the rest of the season than, say, the Tigers starter’s strikeout rate. The Pirates have a pretty decent pitching staff — and an excellent if perhaps overworked bullpen — and a good group of defenders, but that .288 wOBA allowed is probably going up.
That’s why, on our standings page, we don’t use season-to-date numbers to forecast a team’s projected record over the rest of the year, but we instead lean on the rest-of-season projections from ZIPS and Steamer and playing time forecasts from updated depth charts. These numbers take 2013 performance into account, but also adjust for a player’s historical norms and where he is on the aging curve, which allows for a better future forecast than just looking at two months worth of data.
Here, you can see the expected coming regression. The Pirates offense is likely to perform a little bit better, jumping from 3.92 runs per game up to 4.18 runs per game, but the run prevention gets a lot worse, rising from 3.40 runs per game to 4.08 runs per game. Overall, those forecasts see the Pirates going just 56-53 the rest of the way, if they don’t make any changes to their roster.
But, here’s the thing; because the Pirates are already 33-20, going 56-53 the rest of the way would cause them to finish with 89 wins, and the full season forecast on the standings page has them ending the year with the sixth best record in all of baseball, and their final record would be good enough to earn them a spot in the Wild Card play-in game against the Reds. And these records don’t take into account the future upgrades that contending teams will make this summer, so you can probably add some additional wins to the teams at the top of the pile. If I had to guess a final record with the expectation that the Pirates will be buyers this summer, I’d probably pick them to finish with somewhere in the 90-92 win range.
In other words, we can regress the Pirates early season performance heavily, note that they’re playing well over their heads, and that their pitching can’t keep up their current levels while still also acknowledging that they’ve put themselves in pretty good playoff position. Right now, the Pirates should probably be favored to join the Reds as the NL Wild Card teams.
It’s been a long time since Pittsburgh had a winning season. It’s been a long time since Pittsburgh had a baseball team as good as this one. They won’t keep winning at their current pace, but this team should not only be good enough to break the streak of 20 consecutive losing seasons, it might just break the playoff drought as well.