By In Stuff

Statcast Stories

If you have been here before, you know about my preocupation with baseball statistics. I love ’em. I’m fascinated by ’em. I spend way too many hours every day thinking about them.

But it’s also true, as Lina Lamont once said, I ca-a-an’t stand ’em

Yes, it drives me to distration when people rattle off small-sample statistics like “Batter X is 0-for-3 against Pitcher Y” or unhelpful stats like “Fred is hitting .330 with runners in scoring position this month,” or misleading stats like “He now has more postseason wins than Sandy Koufax.”

Bill James has explained it as the “power of language” — some statistics have the power of language and some do not. I tend to think of it in storytelling terms. The statistics I love are ones that, at least for me, open up the world just a little bit, take me into the history of the game, uncork a side of baseball I’d not seen before, challenge my beliefs and conviction. The best statistics do not necessarily give easy answers. They tell fascinating stories.

Statcast is a wonderful MLB tool that measures (in accurate terms previously unimaginable) player movements and baseball velocity and the geometry of the game. It is a Niagara Falls of data — route efficiency, exit  velocity, launch angle, hang time, velocity AND perceived velocity, on and on and on.

As always, the data comes first. And the storytelling follows slowly but surely.

Well, it takes time to figure out what these statistics mean … and more time to figure out how they can add to our understanding and enjoyment of the game. I’m a huge tennis fan and so I watch match after match with absolute horror as the announcer give me the IBM Stat of the Day. It is always something stupefying like, “If Kei Nishikori wins 47% of his second serve points, he has a 72% chance of winning the set.”

I have absolutely no doubt that there is some significant and intricuate math behind this, but even to me — a hardened tennis fan — it means absolutely NOTHING. It tells me NOTHING. It is like the Spinal Tap jazz odyssey — it takes me nowhere. Even if I could do the various calculations to figure out what it means (which is probably no more than “Nishikori needs to win points on his second serve to win”) it would leave me utterly cold.

I’m not saying I’m immune from such number dancing — even a cursory glance at my archive shows that I go down the statistical rabbit hole more than most do. I’m saying that with all this extraordinary technology now and all these new ways to collect data, we are entering a wonderful and perilous new time.  We can tell fantastic stories that would have been absolutely impossible to tell berfore. And we also can drown.

I’ll give you two examples from Game 1 of the Cubs-Dodgers series because I thought the contrast was striking.

On TV during the game, they showed us the Statcast data on a nice play made by, I believe, Dexter Fowler. On the screen they showed that on the play, Fowler had a 96% route efficiency, which could be interesting if I knew what it meant.

And it showed that he reached a max-speed of 18 mph which isn’t interesting at all.

The route efficiency, if I understand it right, shows how straight a line the outfielder made from his starting point to where the ball was eventually caught. There’s promise in that one, but we need some context. Is 96% good? It sounds good. But, as the old line goes, if an airline 96% efficiency in planes landing safely, that would be a horrendous percentage. I’m sure people steeped in Statcast already have a good feel for what route efficiency means but I think it will take some time and history for that to translate into story telling. I know Tom Tango works hard on this.

The 18 mph max velocity thing — it’s just not impressive or revealing. It reminds me of another tennis stat, one where they show how much a player runs during a grueling point. That number is ALWAYS surprisingly low. It’s always something like 113 feet. I mean: big deal. He ran 37 yards. Whoopee. The sheer physical beauty of watching Novak Djokovic run down four impossible balls in a row is so awesome that any number showing the distance run is bound to be an anticlimax. It’s like showing Usain Bolt run the 100 and, in the awed aftermath, saying that he’s about 40% as fast a jaguar.

So, all in all, that Statcast hit did very little for me.

But a little bit later on, there was another Statcast stat — I don’t know if they shared it on TV. The Dodgers’ pinch-hitter Andre Ethier hit what appeared, off the bat, to be a pop-up. Seasoned announcers like Joe Buck rarely get fooled by a ball off the bat. They know to follow the outfielders’ motion. But the outfielders were fooled by this one too. It caught the Chicago wind stream and just kept sailing and sailing until it was over the ivy at Wrigley. It was a fairly startling home run even to the naked eye.

Now, Buck and John Smoltz did make the point that the ball looked like a pop-up off the bat and it just carried. So the point was made.

But Statcast made the point in a much more interesting way. It told a story.

The exit velocity of the hit was 98.3 mph. The launch angle was 43 degrees.

That combination of speed and angle adds up to what is known in the business as an OUT. Every time. When you hit a ball at that speed, at that angle, it is a popup. It is a lazy pop-up.

And because Statcast is so precise, you can cross-reference and prove the point. Baseballs were hit at that speed and angle 40 times. Thirty-nine of those times, the ball was an out.

So you say — well, wait, that means one time it was a hit. That’s true. It was a double.

Here was that play:

So, uh, yeah. That’s an out too. Yes, we could tell that the Ethier homer was unusual. But Statcast shows us that it was a bloody freak of nature. If the Cubs had lost that game, that home run could have become legendary, the immaculate homer. As it turned out, the homer ended up being incidental because Dodgers manager Dave Roberts decided to anger the Intentional Walk Gods, and lo they rained down harsh (but fair) justice upon his team.


But the point remains. I’m pumped up — Tango has promised to help me learn the secrets of Statcast. There are stories in the them thar digits.







Print Friendly

7 Responses to Statcast Stories

  1. Cantankerous says:

    I feel largely the same way about StatCast. They reference it constantly on MLB Tonight with absolutely no context.

    Though similar to Eithier’s home run, they showed the StatCast of Javier Baez’s long fly out from last night and I don’t remember the angle/exit velocity, but the batting average on the combination this season was .900.

    Really, all they need to do is start providing that sort of context and it can be such an interesting, if not always useful tool.

  2. Marc Schneider says:

    Although I generally buy into sabermetrics, my eyes glaze over at some of the stats. I realize, I guess, the significance of exit velocity but I don’t really care. Some of the stats do provide interesting information but others bore the shit out of me. It’s just a baseball game; I don’t feel like I need to stretch my intellect every time I watch a game. And some in the sabermetric community have such contempt for anyone that doesn’t invest totally in the advanced stats. Many of the articles in The Hardball Times, for example, are complete Greek for anyone who isn’t a statistician. I would like to see more/better integration of advanced metrics into baseball broadcasts but I also wish that sabermetricians (and the networks) would stop dumping these often useless stats on us just to show us how cool they are.

    • invitro says:

      I might have to check out The Hardball Times, because most of what you say is news to me. For example, when you say “…isn’t a statistician”, that implies to me that their articles have p-values, t tests, that kind of thing, and I’ve almost never seen those things in sabermetric articles. I.e., Bill James never uses them. And sabermetrics means to me not advanced metrics, but a scientific approach to baseball, an evidence-based approach, to use a catch phrase. You can get a long way toward sabermetrics just by asking “but does the evidence show that is really the case?”

  3. Paul says:

    Example: TBS just praised Aaron Sanchez for having a 2.90 ERA after the jays score, touting him as a stopper. His ERA was 3.00 in all situations… just not very useful information.

  4. Ken says:

    In the same vein, I feel like we are still waiting on the benefits of Pitch FX. It would be interesting to see, for instance, who paints the outside corners the best, or what is a good break for a slider. Other than velocity, we still don’t see much pitch data that’s not result-oriented.

  5. ERam says:

    Joe, I know you hate small-sample stats, but aren’t you ignoring the potential psychological impact of those stats? If a hitter is 0 for 5 against a pitcher, maybe that doesn’t tell us much about what will happen the next time they face each other, but as a viewer, that stat is interesting to me, because I’m sure both players are also aware of it. I also assume they can get caught up in “meaningless” stats just like the rest of us humans. I doubt the pitcher says to himself “Well, 0 for 5 is a small sample, so I’ll just push that data out of my head”. I’m pretty sure he has a boost of confidence knowing that he hasn’t allowed a hit from the guy. Perhaps my example is the exception: small-sample stats are only interesting when they reference an 0 for X situation.

    • invitro says:

      ‘I doubt the pitcher says to himself “Well, 0 for 5 is a small sample, so I’ll just push that data out of my head”.’ — I dunno, I think that’s exactly what many of them say when the pitching coach tells them that. More to the point though, even if there is a psychological impact, research has shown that it has no impact on the baseball game. So ignoring it is a good thing.

Leave a Reply

Your email address will not be published. Required fields are marked *