It’s called: “Who’s the Better Reliever?”
It’s a simple game, really. We’ll take numbers from the past two seasons – 2008 and 2009 – as they relate to Brad Lidge and Ryan Madson. Then you, the loyal group of five readers who read this thing, can vouch for your favorite. It’s not fair to compare saves and holds – both of which are disgusting stats in their own right – as we’re dealing with relievers used in wholly different situations. Counting Ryan Madson’s eighth inning blown saves is one of the most blindsided things you can do. You can’t “save” games in the eighth inning, you can only “blow” them. It’s a lose-lose to count Ryan Madson’s save percentage when he is dealing with a different situation.
Plus, not counting blown saves actually gives Lidge a pass this time. He has as many blown saves this year – 10, nearly 11 – than Madson has in the last four seasons, since the beginning of ’06. Still want to count blown saves?
Why only two seasons, you ask? Well, no one really cares about three-year-old performance. Do you care about 2007 Cliff Lee?
Anyway, shall we dance? All numbers can be cross-referenced with Baseball-Reference, Fangraphs, and TheHardballTimes.
Usually, this stat is pretty useless for a reliever. Someone who pitches one inning – rarely more, often less – per appearance should not be judged on a per nine inning basis. Everything gets skewed by one bad outing, and five good outings barely undo the damage.
That said, here are the cumulative ERAs for Lidge and Madson since the start of 2008.
Madson 1, Lidge 0, but it’s a shaky 1 given the stat.
WHIP is a marginally better stat for evaluating relievers. Walks plus hits per inning pitched give a bit of a better idea of what a reliever does in his appearances. The numbers, please!
Madson 2, Lidge 0.
Does a pitcher strike out more batters than he walks? He should. In fact, he should probably strike out about four times as many batters as he walks, if your name is Joe Nathan, or better if you’re Mariano Rivera. We can’t expect either of our guys to be that good, but I’m sure two-two and a half will do just fine. Behind door number one…
Lidge: 144 K to 66 BB, or 2.18 K:BB
Madson: 132 K to 43 BB, or 3.07 K:BB
Madson 3, Lidge 0.
K/9 and BB/9
Pretty good measures of control and stuff. Obviously, they’re not wholly independent, as balls and strikes rely on the umpires if there’s no swing. Usually, though, the margin of error added by umpires is negligible. The times when an umpire has an egregiously bad strike zone are far, far outweighed by the occurrences of a more “normal” strike zone. I think that’s something we can all agree on to an extent. Given that…
Lidge: 66 BB in 119.2 IP, good for a 4.96 BB/9
Madson: 43 BB in 147.2 IP, good for a 2.62 BB/9
Lidge: 144 K in 119.2 IP, good for a 10.83 K/9
Madson: 132 K in 147.2 IP, good for a 8.05 K/9
Madson 4, Lidge 1.
As mentioned in a previous post, xFIP is a great stat for figuring out just what a pitcher’s ERA should be. ERA, as we all know, is a stat flawed for relievers not just because of use but also because of the rules regarding earned and unearned runs. It’s still better than, say, a pitcher’s record, but not absolute. xFIP is defined as “an experimental stat that adjusts FIP and “normalizes” the home run component. Research has shown that home runs allowed are pretty much a function of flyballs allowed and home park, so xFIP is based on the average number of home runs allowed per outfield fly.” (Thanks again Hardball Times)
Basically, a pitcher’s home run component is removed from his ERA and rearranged to match what is average around the league. It’s not only good for seeing if a pitcher’s performance has been a fluke or unlucky – see: J.A. Happ for the former and Cole Hamels for the latter – as well as an indicator of future performance.
Lidge: 3.93 xFIP
Madson: 3.55 xFIP
It’s not a runaway here, but Madson 5, Lidge 1.
9th Inning Slash Stats
Let’s delve a little deeper. How does each of these guys perform in the ninth inning? A lot was made of Madson’s struggles while Lidge was on the DL a couple months back, mostly due to a tendency to give up HR late. But has he really been so bad in the ninth? Does he really not have the “mentality” or “what it takes” to close the door at the end of games?
Lidge: .238 opponent average, .331 on base percentage against, .388 opponent slugging
Madson: .231 opponent average, .276 on base percentage against, .394 opponent slugging
Madson 7, Lidge 2. It’s fair to note that Lidge was obviously better than Madson in all three facets in 2008. Alas, the disparity is so enormous in 2009 that Madson gets the edge in AVG and OBP, only losing slugging because he’s allowed four HR in 19 ninth inning appearances, while Lidge has, well, surrendered 11 in 54 ninth innings. Rate isn’t much different after all.
Wins Above Replacement
Now, we get to the meaty stuff. Long story short, this shows how many wins that, theoretically, a pitcher’s performance has been worth in relation to any Joe Schmoe you can call up from AAA and throw into your bullpen. A higher positive number is better. A WAR of 0 would indicate absolute league average, and no better than any other guy out there.
Lidge: 2.2 WAR in 2008, -1.0 in 2009, averaging out to 0.6 WAR
Madson: 1.3 WAR in 2008, 1.1 in 2009, averaging out to 1.2 WAR
Madson 8, Lidge 2. Again, it’s astonishing just how bad Lidge’s ’09 has been, enough so that Madson has essentially been twice the reliever Lidge has been over the past two seasons.
Runs Above Replacement
Lidge: 21.7 RAR in 2008, -10.3 in 2009, averaging out to 5.7 RAR
Madson: 12.4 WAR in 2008, 10.9 in 2009, averaging out to 11.65 WAR
Madson 9, Lidge 2. WAR and RAR are cumulative throughout a season, so those ’09 numbers will change before the end of the year, though not nearly enough to make up such an enormous gap. Also worth pointing out, as you can see from a difference in volume, one “win” is worth far more than one “run” as it pertains to these numbers.
Yup, you bet: even individual pitches can be evaluated and graded on effectiveness. Here, we’ll use the two best pitches from each reliever: for Madson, fastball and changeup; for Lidge, fastball and slider.
Lidge’s Fastball: -2.8 in 2008, -13.6 in 2009, good for a cumulative -16.4
Madson’s Fastball: 3.7 in 2008, 4.9 in 2009, good for a cumulative 8.6
Lidge’s Slider: 17.1 in 2008, -0.4 in 2009, good for a cumulative 16.7
Madson’s Changeup: 6.2 in 2008, 6.3 in 2009, good for a cumulative 12.5
Madson 10, Lidge 3. However, Madson cumulatively has 21.1 runs of value in his best two pitches, while Lidge stands at just 0.3.
Now, no manager these days will take notice of these advanced metrics, much less Uncle Cholly. Do not, not even for a second, diminish the importance of these numbers. They are not arbitrary statistics cooked up to kill time and not meant to be tossed aside as some sort of gibberish when, in fact, they are more meaningful than your typical, casual fan stats.
Brad Lidge had an excellent season in 2008. He is having a masterfully terrible 2009. Ryan Madson has been consistently great over two seasons. Too much weight is put on a small sample in 2009 for Madson, and one that ended two and a half months ago, at that! Brad Lidge has been terrible for five months, and some still assert that he’s a better option than someone like Madson where, clearly, that’s not the case.
Contracts and loyalty be damned, I want to win ball games.