I like Lance Lynn as a human being. By all accounts, he’s a great representative of not only the Cardinals, but the game of baseball as a whole. What’s more, he competes the way you want all players to compete, and he’s a great teammate for the other 24 guys on the Cardinals MLB roster.
The thing is, he isn’t one of the three best options that the Cardinals have among starting pitchers — and an argument could actually be made that he isn’t one of the top 5 options. I’m not simply talking about ERA, or wins and losses, or anything so basic as that. What I’m saying is that a comprehensive look at the luck involved in a lot of Lance Lynn’s victories tells the story of an average player who has been made to look much, much better than he is.
Before any sabermetrics nerds freak out on me and point out that according to FIP he’s one of the best No. 2 starters in the league, allow me to both explain that stat to everyone else, and also explain why when it comes to that stat, I’m not all that trusting.
For those who aren’t aware, FIP is the abbreviation for a stat known as Fielding Independent Pitching. The idea behind this stat is to remove extraneous factors beyond a pitcher’s control when calculating his value to his team. The resulting number is supposed to give you a good idea of the level at which the pitcher should have performed, assuming that all factors beyond a pitcher’s control were set to a league average.
The problem with FIP however, is that it is calculated in such a way that literally no contact is counted unless it results in a HR. Meaning that according to FIP, a pitcher who gives up 10 hard hit doubles is substantially better than a guy who surrenders 2 HR and nothing else, assuming both walk and strikeout a similar number of hitters over roughly the same number of innings.
The exact calculation for FIP is as follows:
FIP = ((13*HR)+(3*(BB+HBP))-(2*K))/IP + constant
In this equation, the constant is defined as a number used to bring the FIP number to an ERA scale, and it is usually around 3.20. Let’s apply that equation to the following hypothetical pitchers:
Pitcher A: pitches 6 innings and surrenders 2 HR for a total of 2 ER. He walks 2 batters, and strikes out 5 more. He gives up 2 hits.
Pitcher B: pitches 5 innings and surrenders 0 HR and 10 doubles for a total of 7 ER. He walks 3 and strikes out 4. He gives up 16 hits.
Pitcher A would have roughly a 6.87 FIP
Pitcher B would have roughly a 3.40 FIP
Again, these numbers are on an ERA scale, meaning lower is obviously better. This means that in this (admittedly unlikely) scenario, the FIP stat says that the pitcher who surrendered 7 ER in 5 innings is twice as good as the pitcher who surrendered 2 ER in 6 innings. Personally, I think that couldn’t possibly be more wrong.
The FIP apologists undoubtedly will look at that last paragraph and start in on the thoughts of xFIP, which is the FIP stat modified to account for flyballs by weighing the average HR/FB rate (the rate of flyballs that become homeruns) into the equation. This ignores two major points, however.
Firstly, this equation still doesn’t account for any ball put in play that isn’t a flyball. For a guy like — oh, I don’t know, Lance Lynn – that means xFIP completely ignores 65.6% of all batted balls. Secondly, and perhaps most importantly, xFIP doesn’t take into account the reason behind a pitcher’s HR/FB rate — namely, the parks in which he pitches, and the either batter or pitcher-friendly nature of said parks. A simple illustration of this point would be to say that xFIP would assume the same HR/FB rate for Adam Wainwright pitching half his games in Busch Stadium as it would for Mike Leake pitching half his games in Great American Ballpark. We all know that doing that is beyond ridiculous.
Some of you may need a real world example of the failings of FIP before you believe me, and I don’t blame you. Remember game 6 of last year’s NLCS — the one in which Michael Wacha was better than he was possibly supposed to be? Remember how Clayton Kershaw got absolutely rocked? FIP doesn’t — it says Kershaw was brilliant.
To refresh your memory, Michael Wacha pitched 7 innings, gave up only 2 hits, and walked just one batter while not allowing a single run. Kershaw, on the other hand, pitched only 4 innings, have up 10 hits, and walked 2 batters while giving up 7 runs. FIP scores it this way:
No, seriously. That’s really what you get using the FIP stat — a blind adherence to FIP as the determinant of a pitcher’s effectiveness says that Clayton Kershaw and Michael Wacha pitched with exactly the same effectiveness in game 6 of the 2013 NLCS. You just can’t make up stuff this asinine, or stuff this outrageous — you just can’t make up anything that stupid.
This is the part of the article where I’m supposed to make a point, right? Otherwise I’m just complaining without a real solution in mind. I’m getting there, trust me. But before I get to that I want to wrap this up with a pretty bow — you know, the way real men are supposed to.
I don’t think FIP is entirely useless — and when it’s used both carefully and taken with a grain of salt, it’s actually a very effective tool in the quest to predict future performance for many pitchers. The addendum to that statement is a simple one — I firmly believe there are better ways to gauge the (luck-absent) effectiveness of a pitcher, and I feel that I can prove it.
I made the statement yesterday on twitter (and was met with a haughty and disdainful response) that I prefer ERA+ when determining the effectiveness of a pitcher, rather than FIP. ERA+ is the shorthand for Adjusted ERA — a stat that takes into account not only the ERA of a pitcher, but whether that pitcher compiled that ERA in largely pitcher or hitter-friendly stadiums. To put it simply, if the league average ERA was exactly 4.00, and both a Cardinals pitcher and a Reds pitcher had exactly a 4.00 ERA, the Cardinals pitcher would have an ERA+ below 100, and the Reds pitcher would have an ERA+ of over 100. For this stat, higher numbers indicate better performance, and the advantage to the hypothetical Reds pitcher is due to the fact that while he had the same ERA as the other pitcher, he amassed that ERA in a park that made doing so harder. Also of note is that an ERA+ of 100 is set to be the league average.
Before I delve the depths of my reasoning behind the rest of the article, let me say in no uncertain terms — I do not believe at all that ERA+ is a perfect statistic, or even that it tells the whole story of a pitcher’s effectiveness. What I do believe is that ERA+ tells more of the story of a pitcher’s effectiveness than FIP — and that proving so is a simple numbers game.
Let’s backtrack just a little, and revisit the percentage of batted balls ignored by xFIP in the case of Lance Lynn in 2013 – 65.6%. Nearly 2 out of every 3 balls put in play against Lynn are completely ignored by xFIP. If you simply use FIP (and many Lynn apologists do just that quite often) the number of batted balls ignored skyrockets even higher to 97.5%. I understand just fine the reasoning behind doing so — I just think that anytime you ignore 97.5% of your data, you might want to open yourself up to the idea that maybe you missed something.
FIP and xFIP proponents are no doubt screaming at their screens right now “but what about BABIP?”. Again, for those of you unfamiliar with advanced metrics, BABIP stands for Batting Average on Balls In Play. In his career, Lynn has a BABIP of .316, and last year, it was .314. For the most part, he is one of the worst in the league in this category, and because of that, FIP guys would have you believe that Lynn is terribly unlucky. They couldn’t be more wrong.
First of all, the best BABIP last year was .240, which means completely ignoring all batted balls in Lynn’s case ignores 76% of all such relevant data. This is a problem for obvious reasons, and it is partially shored up by nice stat knows as tERA, or Total ERA. tERA takes into account all batted ball types, but is otherwise largely the same as FIP — in other words, it’s like FIP, except that it doesn’t ignore 76% of relevant data.
Lynn’s tERA is 3.82 for his career, which is a whopping .01% better than his career ERA. So if you look at just that stat, you could say Lynn is .01% unlucky — but there’s much more to look at. Going back to ERA+ and looking at Lynn’s performance tells us that he is largely lucky to have pitched in pitcher-friendly parks, and that his ERA (and tERA) are likely better than they should be because of this fact. His career ERA+ is 96. What’s more, he had a career low in that department last year, with an ERA+ of 91.
Putting all of this together, we see a career tERA that is sitting on almost exactly the current league average and partially skewed in Lynn’s favor by a great deal of luck because of where he pitches. In other words, we’re looking at a player who (all things considered) is likely average at best, or slightly below average — at least according to advanced metrics that don’t ignore the vast majority of relevant data. But the luck doesn’t end there.
Lynn is also the beneficiary of the some of the best run support that anyone could possibly hope to get. If he was getting just an average amount of run support, and was only pitching with an average amount of luck involved, he would likely be a .500 pitcher with an ERA of around 4.00 — and both of those stats would be entirely indicative of his effectiveness.
If he was ending most years with a record of around 10-10 and an ERA around 4.00, he would (for most teams) be a great No. 5 starter, maybe even an acceptable No. 4 starter. But he’s certainly not No. 3 material. For the Cardinals, his case looks a little different.
Out beloved Redbirds are, by almost any account, blessed with the fortune of having a ridiculous amount of high quality pitching. Because of this, they have a number of option that other teams don’t when it comes to starting pitching — options that should likely exclude Lance Lynn from a starting role with this team. I can understand an argument for Lynn being that 5th starter, and I could even stomach (although begrudgingly) the idea of Lynn as the 4th starter. But there just isn’t any evidence that he’s pitching well enough have started 2 of the first 8 games this season.
Lance Lynn is a decent pitcher. Lance Lynn is a good guy. Lance Lynn is a great teammate.
Lance Lynn is ludicrously lucky.