Sam Page's provocative post from this morning caused quite a stir, but most people weren't buying into his prediction (on a METS fan site, no less). Personally, I was skeptical as well. But the post kind of lingered, probably because I desperately hope he's right.
This year's Phillies, while seemingly formidable, are rather one dimensional: they have exceptional starting pitching. The rest of the team is merely pedestrian, particularly if Utley misses a lot of time. If the starting pitching lives up to expectations, it will paper over the team's weaknesses. But what if it doesn't?
I am extremely doubtful that the pitchers will not pitch up to their expected performance level; all four seem to be in their prime. Injuries, however, are a possibility.
So I figured, why not run some numbers?
First off, this is NOT a sophisticated analytical study, or anything like that. Run prevention isn't that smooth, or linear, but I feel like for the purposes of a back-of-the-envelope calculation, it's useful. Our underlying assumption is that Baseball Prospectus' PECOTA simulator of the Phillies is a plausible starting point for projecting where they finish; everything else stems from that.
According to BP's depth charts, PECOTA tells us that the Phillies will score 742 runs. That will be a constant in the study. We'll be tinkering with the runs allowed, though. PECOTA tells us that the Phillies will allow only 664 runs. Using Pythagenpat with those numbers, the Phillies would finish at 89-73, which would probably be good enough for the playoffs.
BP also provides Value Over Replacement Level (VORP) projections for every player based on their projected playing time. VORP is measured in runs, so 1 VORP = 1 run. I made a chart of the projected innings pitched and projected VORP for each pitcher, as per PECOTA. At the same time, I took their totals from 2010 and added those to the chart as well. PECOTA seemed rather pessimistic on the Phillies, so I felt like making an optimistic estimate for the sake of comparison.
So we have two versions of the future here. In PECOTA world, the Phillies allow 664 runs. In the more optimistic world of 2010, the Phillies allow only 612 runs. (That's 664 + PECOTA VORP - 2010 VORP).
The more interesting question, though, is about injuries. How many innings will the Big 4 pitch? Because we're assuming that run scoring is linear for the sake of the calculation, all we need to do, then, is divide VORP by IP and get a value of every inning pitched by the Big 4. (We'll take the Big 4 in aggregate, again for simplification. Obviously, losing Halladay would be much worse than losing Oswalt.)
PECOTA: 160.5/832 = .193 VORP/IP (call this VORPrate)
2010: 212.7/893.3 = .238 VORP/IP
Those numbers are, essentially, how many runs per inning the Phillies' Big 4 prevent in comparison to a replacement-level player (in other words, the type of player who would get those innings if one of them got hurt.) Now, we project out, based on innings pitched. The formula here is projRA + projVORP - (proj. IP * VORPrate). So, we have two different sets of inputs:
PECOTA: 664 + 160.5 - (proj. IP*.193)
2010: 612 + 212.7 - (proj. IP * .238)
And here's how it plays out:
|Proj. IP||Proj. RA (PECOTA)||Proj. Record (PECOTA)||Proj. RA (2010)||Proj. Record (2010)|
So, for PECOTA, we have an easy rule of thumb: start the Phillies from 73 wins (which feels about right intuitively for a Phillies team without the Big 4, right?). For every 50 innings the Big 4 pitches, add 1 win to that.
For 2010, the rule of thumb is a tad more generous: it's more like "add 1 win for every 40 innings the Big 4 pitches."
So from there, all we're left to do is to guess how many innings they'll end up with. You think they'll get 750? They're mean expectation is probably somewhere between 88 and 92 games. Only 600? Then they're probably between 84 and 87 wins. From the looks of this, if they don't get at least 700, they're in trouble. (Disclaimer: random variation may propel them to an outlying season, where they significantly outperformed their run differential or their batting components. Here, all we are doing is looking for an expectation.)
Now, you could take issue with some of the other constants. You could object that even my optimistic model undersells the Big 4, or that the offense will be better than 742 runs, or that the bullpen would be better than the BP estimate, or that the replacement pitchers would be better than replacement-level. But this is simply a model to show how critical those 4 pitchers are, and how fragile the Phillies' are; it's very much a stars and scrubs model of teambuilding (much along the lines of the Minaya-era Mets). But the Phillies' stars are pitchers, who are (generally) more injury prone.
So, what about fourth place? In a fairly competitive NL East, 82-83 wins might still be 4th place, particularly if you're bullish about the Mets. BP pegs the Braves at 87, the Marlins at 84, and the Mets at 81. If the Mets exceed expectation and the Braves and Marlins merely meet theirs, the division may well have 4 teams over .500. If the Big 4 is in the 400 to 500 innings range, Sam may well be right. For the Phillies to only get 500 innings out of the Big 4 would require that at least two miss a LOT of time, or three spend significant time on the DL. Is that unlikely? I think so. Is it impossible? Hardly.
In short, I don't think Sam's prediction is crazy. I do not think it is the most likely outcome this year, but I certainly can respect him for going out on a limb.