None of what you are about to read is intended to suggest that the Phillies made a mistake in signing Jake Arrieta. From a strictly competitive standpoint, the move has little downside.

Even if Arrieta is abducted by one of those UFOs the Air Force keeps spotting and never throws a pitch, the Phillies will be far enough under the luxury tax threshold in each of the next three seasons that they could sign three free agents at $25+ million per year and still need to wildly overpay somebody before they are no longer able to sign whomever they want.

They have eight regulars, a fourth outfielder, two top-of-the-rotation starters, and a closer all under contract through at least 2020 at a total price tag that would top out at around $92 million even with some extremely generous salary projections for their arbitration-eligible players. And that doesn't include a minimum-salary guy such as Scott Kingery who could eventually replace somebody more expensive.

Granted, there's opportunity cost with any move, but the same is true with standing pat. Given the makeup of next year's free-agent pitching market, when the top three arms could be Drew Pomeranz, Dallas Keuchel, and Gio Gonzalez (assuming Clayton Kershaw and David Price do not opt out of their deals), this might have been the Phillies' best shot to use their checkbook to address a glaring weakness.

That being said …

In a vacuum-sealed world where all that matters is profit and loss, the Phillies are taking a considerable risk with a contract that guarantees Arrieta $75 million over the next three years.

The free-agent market is often a place of pyrite rather than gold. In the four offseasons leading to this one, 21 starters signed contracts of three-plus years that paid $10+ million per year. Nearly half — 10, to be exact — failed to reach 175 innings in the first year of their deal, and seven failed to reach 150 innings. In their walk years, those 21 pitchers averaged 192 innings pitched with a 3.39 ERA. In the first year with their new team, they averaged 163 1/3 innings with a 3.80 ERA. In Year 2, those numbers declined to 139 1/3 innings with a 4.16 ERA.

Arrieta will be older than all but six of those pitchers were in the first years of their deal. And there's plenty to suggest that he has already started down the side of the aging curve that, as we saw with Roy Halladay, can get steep in a hurry.

We're not just talking about end results, either. Arrieta pitched just 168 1/3 innings with a 3.53 ERA after averaging 213 innings with a 2.39 ERA in his previous two seasons. And a deeper look at his performance suggests something more was in play than year-to-year variance.

There has been a lot of attention paid to Arrieta's fastball recently, and for good reason. Velocity isn't necessarily a determinant of success, but it does offer us a baseline that can signal changes in a pitcher. Both the Phillies and Arrieta talked about it in his press conference Tuesday, with the club citing the righty's performance over the second half of last season as evidence that he can still be effective at 91-93 mph instead of 93-95 mph.

But Arrieta's diminished fastball wasn't the only conspicuous change in his process last season. In fact, there is some evidence that suggests the biggest difference between the pitcher who won the Cy Young in 2015 and the guy we saw last year was the sudden impotence of a piece of his arsenal that he has previously credited for his late-career resurgence. 

The cutter is something of a controversial pitch in baseball circles, given the strain it puts on the arm. There's data to suggest that pitchers who lean heavily on it are more prone to injury than others. Whatever the case, Arrieta threw one of the better cutters in baseball, a crisp, low-90s version that could induce bat-breaking contact or miss the thing entirely. In his dominant 2015 campaign, Arrieta recorded more outs with his slider/cutter than with any other pitch. That season, it accounted for nearly one of every three pitches he threw, a whopping 29.5 percent.

Over the last two seasons, though, his usage of the pitch has plummeted, as you can see in the table below, which features the percentage of fastballs and percentage of sliders thrown in both the first and second halves of last season compared to the previous three years.

Year %FB %SL
2014 47.6 27.6
2015 51.1 29.5
2016 65.0 18.7
2017, Apr-June 61.3 14.3
2017, Jul-Sept 67.2 13.6

That's a dramatic drop. In both his disappointing first half of 2017 and his excellent second half, Arrieta was throwing his slider/cutter 50 percent less often than 2014-15.

There's at least one logical explanation: The pitch wasn't as good as it used to be. In the table below, you'll find two outcomes that offer some indication of the effectiveness of a pitch: the number of swings-and-misses it generated, and the number of groundballs it induced. In 2015, opposing hitters whiffed on 17.2 percent of the sliders/cutters that Arrieta threw. They hit groundballs on another 11 percent. In the second half of last season, those two numbers were just 10.9 percent and 5.2 percent.

Slider Whiff% Slider GB%
2014 14.7 10.6
2015 17.2 11.0
2016 13.4 9.2
2017, Apr-June 16.0 8.3
2017, Jul-Sept 10.9 5.2

Again, the evidence suggests an explanation. In the table below, you'll find a quantification of the horizontal and vertical movement on Arrieta's cutter/slider as recorded by

Slider Velo (MPH) hMov (inches) vMov (inches)
2014 90.2 3.49 -25.77
2015 90.7 3.41 -26.11
2016 89.7 3.55 -27.38
2017, Apr-June 87.8 3.45 -29.17
2017, Jul-Sept 88.8 2.11 -25.56

In essence, during the first three months of last season, Arrieta's slider featured a lot more 12-6 action than it had when he was at his most dominant. That changed in the last three months, but at the expense of some of the horizontal break that the pitch once featured.

Granted, we're getting pretty granular here. And none of these numbers tells a complete enough story to form the basis of a definitive conclusion. But they are an indication that things are different from what they used to be. There are plenty of potential explanations that don't involve physiological decline. It's just worth noting that, in the physiological world, change moves in one of two directions, and is difficult to stop.