Pfft, Who Needs A Good K/BB?
Just like he did this past offseason, Billy Beane recently traded one of his top pitchers for a bevy of prospects. Dave and Marc have done a great job profiling this Rich Harden deal and the pool of talent heading to Oakland. Earlier in the season, when discussing the Dan Haren move, I noted how Dana Eveland and Gregory Smith, just one third of the return from Arizona, were performing quite well. At that time, which was after the month of April had been completed, Eveland had a 1.86 K/BB and 1.34 WHIP while Smith had a 1.91 K/BB and 1.09 WHIP.
At that point in time both players were outdoing their FIP via ERA and, while their K/BB totals have taken somewhat substantial hits since my last look, they have continued to “beat” their controllable outcomes. For the season, here are their numbers:
Gregory Smith: 17 GS, 3.62 ERA, 4.24 FIP, 1.57 K/BB, 1.27 WHIP, 74.4% LOB
Dana Eveland: 18 GS, 3.50 ERA, 3.84 FIP, 1.35 K/BB, 1.38 WHIP, 75.3% LOB
Now, over the last thirty days they have both posted great ERAs, but are way out in front of what their FIP would suggest, which is higher primarily due to their ultra-low strikeout to walk ratios. Neither are punchout machines, but they have the 2nd and 4th lowest ratio in the last thirty days. For the season, Eveland has the third lowest in the AL while Smith joins him in the top ten at spot number eight. Here are their numbers in these parameters:
Gregory Smith: 3.34 ERA, 5.15 FIP, 4.85 K/9, 5.76 BB/9, 0.84 K/BB, 81.5% LOB
Dana Eveland: 2.89 ERA, 4.01 FIP, 4.58 K/9, 4.34 BB/9, 1.06 K/BB, 80.9% LOB
Combining for 1.22 WPA wins in this span, Smith and Eveland have found ways to produce arguably better results recently despite the triumvirate of less strikeouts, more walks, and more runners on base. The Athletics rotation currently has a 3.47 ERA and .238 BAA, both tops in the AL. They have struck out the sixth most amount of batters while simultaneously walking the fifth most; this places them 10th out of 14 teams in terms of K/BB ratio. With three pitchers in the top twelve (Smith, Eveland, Joe Blanton) this does not really come as a surprise.
It will be very interesting to see what happens to this rotation sans Rich Harden and Chad Gaudin but, for the sake of having much confidence moving forward, Smith and Eveland should be working towards improving their controllable outcomes; that way their success could be defined by skills and not potentially luck.
Eric is an accountant and statistical analyst from Philadelphia. He also covers the Phillies at Phillies Nation and can be found here on Twitter.
A best-in-baseball 0.723 DER goes a long way to making up for those “missing” strikeouts… Exactly how much of that FIP discrepancy can be accounted for by defense is beyond me.
I imagine the DER could account for around 50% of the FIP-ERA deviation. Given FIP is calculated based on a league average fielding in a league average park, the A’s currently are benefitting from their surroundings.
The problem is that Billy may be building for success in a baseball season but not during the playoffs.
Baseball Prospectus wrote a book, Baseball Between the Numbers, and in Chapter 9.3, they studied what is correlated with playoff success. The three factors that were significantly correlated with playoff success were: 1) K/9 of the staff; 2) closer effectiveness as defined by WXRL; and 3) team defense as defined by their defensive metric. Thus, you want a staff full of strikeouts artists, much like the Giants currently have with Lincecum, Cain, and Sanchez.
Instead, Billy has traded away his two top K/9 starters in Haren and now Harden, while keeping his low K/9 pitchers like Blanton and Duchscherer, and trading for low K/9 pitchers like Eveland and Smith. And his previous rotations were low on K/9 as well with Hudson, Mulder, Zito heading them up.
Relying on low K/9 starters mean you have to rely on your defense for outs, leaving it to chance how well you do in getting the other team out.
Well, since I’ve been knee-deep in old newspapers for research for my next book (on Bucky Walters, pitcher from 1935-48), Reds manager Bill McKechnie did something similar; he would go for some of the best defenders to offset his pitchers average or below average K/BB ratios. The Reds won the world series in 1940.
Billy did mention in Moneyball that he builds for success in the season to get the team into the playoffs and then, once there, it’s essentially a crapshoot where anything can happen.
Eric, please explain how a low K/BB ratio allows a pitcher to sustainingly outperform his FIP to any statistically significant degree, let alone being the primary reason for the discrepancy. Seriously. Your explanations are classic.
Lucas, unless I completely misunderstood Eric’s post, you are saying the same thing as him. His whole point is that they need to/should improve their ratios in order to make their success more legitimately real as opposed to lucky. Looks like a simple misunderstanding.
Yes, the point of this post, and I’m not sure how anyone could miss it really, is that Eveland and Smith have outdone their FIP all year, which has been higher than their ERA due to their low K/BB ratios. FIP is calculated by incorporating the K,BB, and HR, which are the only three controllable outcomes for a pitcher.
Therefore, if your K numbers are low, and BB numbers high–which would result in a poor K/BB ratio, it isn’t very likely to produce a tremendous FIP unless you allow no home runs or something of that sort. Misunderstanding or not, this is all very clear in the article as I just re-read it twice.
There was nothing in this post pertaining to how or why they have outdone it but rather evidence suggesting what they would need to work on for their success to be a result of skill and not luck or defense.
And the point of my post, and I’m not sure how anyone could miss it either considering I stated it as clearly and unambiguously as possible, is that K/BB rates have little if any bearing at all on any discrepancies between FIP and ERA in either direction. Please check the Advanced tab of FanGraph’s leaderboards to confirm this. A low K/BB guy is just as likely to outperform his FIP as he is to underperform it, and he’s more or less just as likely to outperform his FIP as a high K/BB guy is. Yes, you leave more to chance when you don’t strike men out, but FIP already takes all of these controllable skills into account while normalizing the things largely out of the typical pitcher’s control (BABIP, LOB%).
Dana Eveland’s ERA has a slight edge on his FIP because of a .285 BABIP and a 75.3 LOB%. Gregory Smith is kicking FIP’s ass because of a .255 BABIP. It’s that simple. The only misunderstanding was on your part.
Okay, I see your point, and apologize for misunderstanding/taking it the wrong way. FIP measures the HR, BB, and K, leaving the stats out of a pitcher’s control out of its calculation. My point, which I guess didn’t come through as well as I would have hoped, is that both of them have had much higher LOB rates in the last 30 days than in the time prior, while Eveland has had a .300 BABIP and Smith comes in amongst the lowest at .218.
Based on their lower K numbers and higher BB numbers, their FIPs have been 4.01 and 5.15 in this span, however Smith has outdone his much more due to such a low BABIP.
I didn’t mean to imply their K/BB is necessarily correlated with out- or under-performing the FIP with ERA, but rather that, for their FIPs to be lower they will need to increase strikeouts and decrease walks.