Why I was wrong about Jeff Samardzija

One of the great things about the internet (besides this) is its ability to preserve statements forever. And by great, I really mean the worst thing about the internet ever. It was about a year ago that I had the following exchange with Cubs Den blogger John Arguello regarding Chicago pitcher Jeff Samardzija:

Arguello went on to make his case (which was reasonable) and I backtracked slightly, but ended with the following statement:

In hindsight, this makes me look quite bad. We know now that Samardzija easily cleared the “competent bullpen arm” bar I set last March, throwing 174.2 quality innings for a mostly hapless Cubs team, with an ERA of 3.81, a FIP of 3.55, and an xFIP of 3.38. His strikeouts per nine innings (9.27) and walks per nine innings (2.89) were similarly strong. The Notre Dame product even ended the year as the Cubs’ ace (because every other candidate was injured or traded, but still). Just this week he turned down a Cubs’ offer of “well above” $30MM to sign him through his next three years of arbitration and first two years of free agency. In April he will make the first Opening Day start of his career. And so on.

The dude had a solid year, okay?

So, what happened? And how did I miss it? I’ve been thinking a lot about this lately, because as much as I hate being wrong, I value the ability to learn from such missteps more. Being wrong is not the enemy. Being wrong happens, especially in a venture like this, where so much luck is involved, our information on these players is far from complete, and even Hall of Fame hitters get out as much as they reach base.

So why did I whiff?

Very simply, I took my eye off the ball. I stopped collecting data because I had already decided what Jeff Samardzija was and what Jeff Samardzija wasn’t. He was the football guy from Notre Dame, the one the Cubs threw too much money at. The one with the stupid (okay, awesome) hair. He was the pitcher with the good fastball, an inconsistent slider, and poor command. He was the guy who laughably saw himself as a starter heading into camp last spring when most thought he was competing for a bullpen job. He certainly was not a starting pitcher. Not hardly.

A cursory check of Samardzija’s statistics told me all that I knew I already knew. The problem being, as is often the case, things were more complicated than that. Too often in sports we forget that each athlete is a human being, with a unique set of circumstances behind their personal evolution. Yes, we’re also barraged with stories that mean nothing and have no effect on that player’s performance going forward, but sometimes that stuff isn’t bogus. Sometimes the 27-year-old prospect really is still developing after having split his interests between two sports on college. Sometimes coaches make a tweak that dramatically alters a player’s career arc. Sometimes the numbers are trying to tell us things that we are too busy talking over to hear.

I won’t go into too much detail here, because Samardzija has been covered well enough here, here, and here. What concerns me now is the data from 2011 that I missed because I glanced at Samardzija’s 5.11 BB/9, saw an unsustainably low BABIP of .253, confirmed my bias, and moved on with my life. What I missed, were the details. Specifically, just how much Samardzija’s control improved during his time as a reliever in 2011. Yes, the final tally wasn’t pretty, but it belied strong and steady progress throughout the season.

A Hardball Times Update
Goodbye for now.
K/9 BB/9 K/BB K% BB%
2011 Mar/Apr 9.98 9.39 1.06 24.30% 22.90%
2011 May 11.05 5.52 2 26.10% 13.00%
2011 Jun 8.22 4.7 1.75 21.50% 12.30%
2011 Jul 7.07 3.86 1.83 17.20% 9.40%
2011 Aug 9.56 3.38 2.83 27.90% 9.80%
2011 Sept/Oct 7.11 3.5 2 19.60% 9.80%
2011 1st Half 9.06 5.9 1.53 22.40% 14.70%
2011 2nd Half 8.67 3.96 2.19 23.70% 10.80%

Samardzija dropped his walk rate virtually every month. His ability to do that while maintaining his strong strikeout numbers make it all the more impressive. There were other signs of his progress from 2011, as well, including jumps in his ability to elicit more swings on pitches outside the strike zone, and more swinging strikes in general. It’s clear that, by the end of 2011, the Jeff Samardzija coming out of the pen was not the same one who was doing so the previous April. He already was more than a “competent bullpen arm” and was well on his way to last year’s breakout. My eyes were just too closed to notice.

That said, he made even more progress last season:

K/9 BB/9 K/BB K% BB%
2012 Mar/Apr 9.38 3 3.13 23.80% 7.60%
2012 May 9 2.48 3.64 25.50% 7.00%
2012 Jun 7.71 5.79 1.33 17.20% 12.90%
2012 Jul 9.82 3.55 2.77 27.70% 10.00%
2012 Aug 9.86 1.64 6 27.50% 4.60%
2012 Sept/Oct 9.56 1.13 8.5 27.40% 3.20%
2012 1st Half 8.88 3.29 2.7 23.20% 8.60%
2012 2nd Half 9.82 2.33 4.21 27.50% 6.50%

There’s one obvious outlier there (June) and it’s fair to wonder if that was related to the curveball he was screwing around with. When he dropped it, his walk rates were downright elite.

Where do we go from here? The cat is well out of the bag on this particular player, and I was smart enough last year to realize the folly of my ways before it was too late, acquiring Samardzija for virtually nothing in my dynasty league. And now, having reflected on my past error, maybe (just maybe) I’ll think before I scoff next time.


4 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Fatbot
11 years ago

I think you’re being too hard on yourself. I luckily picked him up before his breakout, but not because of some secret formula or skill. The stats you present make it look obvious, but for every Shark there’s probably at least 20 guys with the exact same 2nd half improvement that fail. If not, then 1st half/2nd half comparisons would be more popular than WAR, FIP, etc.

Bottom line, it wasn’t a “mistake”, it was just luck. Which is why the stats remain descriptors, not predictors.

Steve-O
11 years ago

I said the same thing three years ago to co-worker. He believed in the shark, I thought he was a minnow.

Jack Weiland
11 years ago

@Fatbot … I think you’re underestimating the amount of snark I intended to send John’s way. But yes, I agree to some extent. I think my mistake was not being open to the potential that a once highly regarded prospect (in some circles, at least) had made such significant progress. That’s what bothers me more than anything. And I do think it’s fair to read into the improvement in control in 2011, although it’s very possible the knowledge of 2012 is skewing that information for me. At any rate, the point is: keep an open mind, and research, research, research.

@Steve-O … you were right for about a year and a half. That ain’t bad.

Brad Johnson
11 years ago

It’s easy to get stuck in the null hypothesis mindset. For Samardzija, the null would have had something to do with him retaining his poor control and inconsistency. For most pitchers like Samardzija, that’s what happens.

We need to recognize that we can be 95% certain of something will happen and the other thing could still happen. Samardzija was indistinguishable from any number of other pitching busts. On the flip side, we also need to recognize that we form confidence intervals without referencing some of the most important information.

I think the lesson here is to reserve the snark for all but the most egregious cases. For example, I had a co-worker last year (Giants fan) who insisted that Freddy Sanchez was a good hitter.