We Don’t Know Jack
Sports fans love to prove their knowledge. They simply need to show that they know more than other fans, their friends and their family. Fantasy owners take that concept even further by putting money where their mouth is. Whether projecting breakouts, predicting busts or drafting minor leaguers they expect to become the next superstar, there are countless opportunities for fantasy owners to show they know their stuff. Unfortunately, for as much as we think we might know about the game and explaining player performance, there are examples every season that convince me that in fact, we don’t know jack.
When looking at breakouts and busts, we look deep, analyze every FanGraphs metric, watch the games with our expert scouting eyes, and try our hardest to explain why what we are seeing is happening. But that still falls short.
Let’s start with perhaps the season’s biggest disappointment, Adam Dunn. As we are well aware, Dunn has posted an embarrassing .162/.292/.308 line and .277 wOBA after years of extreme consistency.
So why has Adam Dunn sucked?
Well to start, his K% is a career worst 43.2%. Okay, well why? I don’t know.
His IFFB% of 16% is above league average and at its worst mark since 2003. Okay, well what’s the reason for this? I don’t know.
He has just 2 hits in 63 at-bats (a hilarious .032 average) vs lefties. Why the sudden massive ineptitude vs lefties? I don’t know.
His power has gone missing, as his ISO and HR/FB ratios are at career worsts. Why has his power gone MIA? I don’t know.
He cannot hit the fastball, with a negative pitch type value for the first time in his career. Fantastic, but why did he suddenly lose his ability to crush fastballs? I don’t know.
We could speculate on explanations until we’re blue in the face, from his appendectomy, to the switch in leagues and learning new pitchers to becoming a full-time DH. But the bottom line is we just don’t know.
And how about Hanley Ramirez? He has been another high profile bust this year and we could run the same exercise getting absolutely nowhere.
What is behind Hanley Ramirez‘ bust of a season?
He is hitting the lowest percentage of line drives of his career, leading to a career low BABIP. Great, but why? I don’t know.
He has posted the highest GB% of his career. That will certainly limit his power output, but why has it happened? I don’t know.
His ISO is at a career worst and below the league average. I could see that, but what’s behind this power outage? I don’t know.
According to pitch type values, he can’t hit fastballs, sliders or change-ups. Why the sudden struggles though? I don’t know.
Has Hanley’s back bothered him all season, which could explain his poor first half? Possibly, but unless he comes out and states this, then in keeping with the theme of this post, we just don’t know.
We FanGraphs/RotoGraphs authors have a tough task sometimes. When trying to analyze a surprise performance, we end up going halfway with our explanation, and I am guilty of this as well. We could explain a hitter’s struggles exactly the way I listed some reasons for both Dunn and Hanley above, but we still have no idea why. And the why is the key. Truly answering these “why?” questions could help us to determine if these performance levels will continue or a return to previous averages should be expected. Maybe HitF/X will push us in the right path, but until then, we will be left scratching our heads when these surprise performances occur. Whether these performances continue at these rates or the player reverts back to what we originally expected before the season is now just a roll of the dice.
Mike Podhorzer is the founder of ProjectingX IQ, an advanced fantasy baseball analytics platform that transforms projection data and in-season performance signals into actionable intelligence. He is the 2015 Fantasy Sports Writers Association Baseball Writer of the Year and three-time Tout Wars champion. He is the author of the eBook Projecting X 2.0: How to Forecast Baseball Player Performance, which teaches you how to project players yourself. Follow Mike on X@MikePodhorzer and contact him via email.
My pet theory on Dunn is that his appendectomy led to either a subtle change in mechanics or else reduced his bat speed. A greater incidence of infield flies and more strike outs supports the slower bat hypothesis as both often correlate with declining bat speed.
It could just be age or a combination of injury and age. Dunn wouldn’t be the first player to thrive through his 20’s and disappear in his early 30’s. And his skill set doesn’t leave a lot of room for error.
I think Brad’s first theory is worth exploring.
http://joncolosimo.com/2011/07/adam-dunn-where-did-the-bat-speed-go/
Nice article. I agree that it is amazing how much we still don’t know, despite all the information we get.
Has there been any analysis on whether switching leagues has a real impact? It’s certainly a sensible theory.
I think with hanley its been his effort into his game that has been the problem and with McKeon as the new manager he might be tryin to bust his ass. Remember Hanley has had this problem in the past and is one of the reasons why the red sox gave up on him.
I think its easy for us to forget sometimes that these players are humans and not machines. We do not have a lot of information on their personal lives and we cannot track those stats so I think a lot times that is where the “i don’t know” answer may lie.
I think an obvious and huge element that isn’t considered on Fangraphs is Hard Hit Ball %. Baseball HQ tracks this but it is Video based data – that is, they do not measure the actual speed at which the ball comes off the bat, but rather opt for subjective analysis of whether a specific hit was mashed, hit just solidly, or hit softly.
I see no reason why, theoretically, we can’t measure how hard a ball is it – a metric can be developed that looks at how hard a player hits a GB / LD / FB vs the rest of the competition and scores it based on where he falls in the pool. Or you can simply look at the MPH of a ball hit.
HH ball % is likely as important to sustaining a high babip as LD % is. It makes sense. Guys who mash the ball get more hits. Period. Guys who mash the ball the hardest get the most hits.
Guys like Dunn and Hanley who are just flat out sucking and making very poor contact on the ball are not squaring the ball up well. Yea, there are “why” questions behind that as well, but knowing how hard they are hitting the ball would help explain a lot of the why questions above.
It’s not just that we don’t know some of these things – “I don’t know” may in fact be the correct answer. In other words, let’s not forget the role of random variation. At this point in the season we start to feel like sample sizes are becoming large for individuals, but let’s not forget that we are also dealing with a large total number of players… and in a large sample of players there will always be SOME major outliers on both sides of the mean. When we look at those outliers individually we feel like there must be some reason for their over/under-performance. But when viewed in the context of all players in baseball we may just be looking at the tails of the distribution.
At least this is what I keep telling myself with Zack Greinke…
This is a pretty good take. I was going to say there’s something almost quantum at work, in addition to the complexity and the range of factors both measurable and unmeasurable.
The problem with predictions as I see it is that you usually get a mix of qualitative and quantitative analysis assembled into a narrative which is basically arbitrary and bias filled. A typical analysis goes something like this: “this player has a low BABIP and a line drive rate that indicates he’s been hitting in bad luck. Plus he’s in a contract year and recently moved up to the 5th spot in the line up where he should benefit from seeing more pitches to hit.” But there’s no evidence that this particular combination of factors is actually predictive of anything. On the surface it appears reasonable and makes sense, but it’s narrative sense. If said player then experiences a resurgence we validate the methodology, but if it fails to happen we question whether or not we used the right factors and will often find the opposite supporting evidence in retrospect.
I think that we know that our predictions, when done well and with rigor, are still not actually predictions but probability ranges. And way too often, especially in the case of fantasy, we forget the importance of sample size relative to these probability ranges. That’s why I’ve always said the fantasy analysis is far too focused on tout-style predictions about players and pays too little attention to tactical gaming and broader strategic considerations.
In short. Correlation does not imply causation. Story of science.
I mean, I wasn’t expecting this much of a decline, but I avoided him like the plague going into this season. His walk rate plummeted last season, and his O-Swing skyrocketed. It seems like people kind of ignoring that when projecting 50 homeruns in the more friendly environment.
I certainly think that Dunn can rebound in August. Maybe hit the gym a little more and defintely work on his mechanics on the off days. There are two things that standout to me in his stats, and thats his zone% which is at 39% and his o-swing% which stands at close to 58% (used to be 62%). I think there is slowly some progression here in improvement. I wouldn’t be surprised if he turns it around very soon.