I was very tempted to make this a joke review. I wouldn’t have written an intro or offered any insight and would have instead simply left a bolded “Grade F… no explanation needed.” Qualls was the most maligned Phillies player this season and was out the door by the end of June. He blew a number of games in a limited amount of time and his entrances became associated with doom.
However, upon reflecting on his season, I realized that Qualls is a solid proxy for a discussion of certain statistical and analytical concepts. Consider this a hybrid review — I’m still giving him an F, so let’s get that out of the way right now, but he is worth discussing for important reasons that go beyond his cringe-inducing performance in a Phillies uniform.
Qualls’ time with the Phillies serves as the perfect foray into discussions of: incorporating context, ERA and run prevention, the difference between per-9 and per-plate appearance statistics, and the need to dig deeper and consider all pertinent information in any analysis. At least he gave us something.
While this seemed like a low-risk move, given Qualls’ minimal salary on a one-year deal, his turning into a pumpkin shouldn’t have come as a surprise.
Qualls posted a 7.32 ERA two years ago, in part because he located his pitches poorly and gave the opposition a better chance to make solid contact. Last season, his ERA was beautified by the friendly confines of PETCO Park. However, his FIP was 10% worse than the league and he finished with -0.2 WAR.
On paper, relievers with solid strikeout, walk and groundball rates should be heavily pursued. Qualls is still a 55% groundball pitcher, but his strikeout and walk rates have worsened in recent years. While his 2007 production should have no bearing on what the 2012 Phillies should have expected, contextually, it’s important to note that his peripherals were declining.
Before digging deeper into these concepts, let’s first relive his Phillies tenure. Qualls logged 31.1 innings over 35 games and finished with a 4.60 ERA and 5.65 FIP. Of his total -0.4 WAR on the season, -0.6 came with the Phillies. He posted respectable K/9 and BB/9 rates of 5.5 and 2.6, respectively, but served up home runs at a ridiculous pace of 2.01 per nine innings. Though he kept the ball on the ground at a 54.5% clip, thereby reducing the number of flyballs he allowed, those flyballs turned into homers at a 25% rate. Both his HR/9 and HR/FB rate with the Phillies would have been among the very worst in the sport.
Among relievers with 30+ innings, only Francisco Cordero (2.06) and Livan Hernandez (2.00) would have been in the same HR/9 stratosphere if Qualls’s time with the Phillies appeared on the leaderboards. His 25% HR/FB would have only been topped or matched by Hisashi Iwakuma (27.3%) and Marc Rzepczynski (25.0%).
An obvious response to his astronomical home run rates is that his strikeout and walk rates were stable and the home run rate would regress, but that ignores the other side of the equation: Qualls’ pitch location was terrible. Players weren’t golfing excellent pitches into the stands. They were belting elevated sinkers thrown middle-in. Players can’t ‘regress’ in the way we commonly think of the concept if they continually put themselves in positions to fail.
His poor run prevention, largely fueled by his propensity for serving up dingers, manifested itself in blown leads, blown ties, or allowing the opposition to extend their lead and keep the Phillies out of the game. Baseball-Reference player game logs have two columns at the end that describe the situation a pitcher entered into, and the situation when he exited. Qualls blew five leads for the Phillies in his 35 appearances. There were also three games in which the Phillies were down by a run or two when he entered, and down by four or five runs when he exited, which is almost just as bad.
Getting back to his strikeout rate, I would even argue that it was less impressive than it was generally perceived. Contextually, his strikeouts per plate appearance was low for his K/9. The two might seem like different representations of the same stat, but there is more than meets the eye here. Using plate appearances as the denominator more accurately reflects production than using innings.
If we have two pitchers, both of whom pitch one inning and strike out one batter, they both finish with 9.0 K/9s. What if the first pitcher, in addition to striking one batter out, allowed two hits and a walk, while the second pitcher has a 1-2-3 inning? Suddenly, both pitchers have the same strikeout rate on a per-inning basis, but they certainly didn’t achieve that production in the same manner. From a per-PA standpoint, Pitcher #1 faced six batters and has a 16.7% strikeout rate. Pitcher #2 is at 33%.
Why does this matter for Qualls? Well, because his 13.6% K/PA was low for someone with his K/9. The average K/PA among all relievers with 30+ innings and a K/9 between 5.0 and 6.0 was 15.1%. If the range is between K/9s of 4.5 and 6.0, the average is still 14.4%, well above Qualls’s 13.6% with the Phillies. This is an example of both digging deeper and incorporating context, which is imperative lest we accept inaccurate determinations of the value of certain data points.
Qualls pitched poorly for the Phillies and was a major reason for the poor bullpen stats the team posted as a unit. However, that gets to another topic for another day — using full-season stats to evaluate the Phillies’ bullpen isn’t truly accurate, as some of the worst pitchers were gone by the middle of the year, while their stats weren’t.
Qualls gets an F for his partial season with the Phillies, but his season is an A+ example of why certain concepts are very important to consider when evaluating performance. That supposedly clever turn of phrase also gets an F.