Baseball is using Big Data to Engineer Pitchers
And Business should Learn these 4 lessons
Baseball is way ahead of business in the use of big data and analytics. I recently read this article on Josh Hader’s fastball and realized how advanced baseball has become in pitching analysis. Baseball is using big data to engineer pitchers and optimize performance. Business should ask itself about how our businesses compare to baseball.
This article is an analysis of Josh Hader’s fastball which is arguably the most effective in baseball. He has the highest swing and miss rate in MLB at 38.4%. Opponents batting average against the pitch is an anemic .132 and an absurdly high 50% of batters strike out against Hader. And yet Hader ranks 66th in average fastball velocity has a below-average spin rate. But baseball is so good an analytics it can figure this out and is the process of using that data to teach other pitchers to be as effective as Hader.
Imagine if we could do that in business. What if we could figure out who our most productive employees are and why? And what if we could use those learnings to make all our employees better. Holy Sh*t! That would product some productivity and some competitive advantage, now wouldn’t it?
Here are four high-level lessons that we in business can learn from the work baseball
- They have defined the optimal outcome (s) that they want
In the pitching example, the optimal notice as the most basic level is a pitch that causes a swing and miss. If a pitcher can throw a pitch and have the batter miss, then that is clearly the best outcome. But they don’t stop there. They measure strikeout rate and batting average per pitch. They measure foul balls vs balls put in play. And they measure exit velocity and launch angle of a batted ball. In short, you want pitchers with high swing and miss rates and/or ones that produce a lot of weak ground balls (Low exit velocity and low launch angles)
In business, we need to focus on better defining and measuring our outcomes — closed sales, resolved service calls, on-time deliveries, etc. And we need more granular metrics and advanced stats.
2. Baseball measures everything
MLB stadiums are outfitted with a Statcast system that uses Doppler radar and other technology to captures almost every data point associated with a pitch and the pitcher’s delivery. We measure the standard stuff like velocity, location, and outcome. But they measure the more advanced and then the super-advanced. — movement, spin rate, spin efficiency, transverse spin gyroscopic spin, release height, release angle, arm axis and much more.
The Statcast systems that are the official metrics of the game are a bit dated, Individual clubs are augmenting with technology with Trackman, Edgertronic cameras and Rapsado that allow even more accurate and granular measurements. In short, baseball is creating an incredibly large and detailed dataset.
3. Baseball has built great models
Having large data sets is not enough. And despite what people think, we can just run AI algorithms against the data set and have answers magically pop out. Pitching experts and big data people have collaborated to build models that predict what makes a perfect pitch and a good pitcher. Turns out that for fastballs, pitches which produce a Magnus effect — pitches which “appear” to rise to produce the most swing and misses. And the models went further to determine that generally high transverse spin rates produce the nest Magnus effect. Lastly, the new baseball data scientist have then built models to determine how best to produce high transverse spin rates. It’s pretty damn incredible. We now know at a very detailed level what pitching motion produces the best pitch.
I don’t think we are very close to that in business. A few of our most recent high tech digital businesses like Amazon, Netflix and Google might be on the path to this, but most traditional business is not even in the first inning of the journey they must travel.
4. Baseball is using the data and the models to teach players to be better
The state of baseball pitching instruction has moved exponentially over the last decade. It used to be a bunch of over-generalized nondetailed instructions that were virtually useless to the pitcher. “Throw strikes”, “Get on top of the ball”, “Don’t open up”, “Finish the pitch”. Anyone who has been to a Little league game has heard coaches yell this from the dugout. But these instructions are useless, they don’t provide the pitcher with information on what to do to fix it.
Baseball is now engineering the perfect pitch and pitcher. Training facilities and bullpen session are now deep examinations of specific body movements and positioning that produce pitches that are hard to hit. Each pitch is measured at a microscopic level and then drills and advanced training machines are being designed to engineer a pitching motion for each pitcher.
And in-game adjustments will be engineered as well. We are not very far away from a pitcher coming off the mound to hear: “Good inning, but the spin efficiency on the fastball rated drooped. Your arm slot showed was 2% lower than your optimal and your release angle was +5% higher than normal, and the pelvic rotation was slower than the previous inning. Let’s just simulate a few motions here to get you back on track.
Most businesses are nowhere near engineering the actions employees must take to produce optimal results. Our employee effectiveness and training programs are often the equivalents of an old man yelling “Throw Strikes”. We need to get on a better path,
Competitiveness and urgency
I have read and experienced enough on the topic of engineering pitchers to be a novice. But I have talked to enough experts to hear that while the datasets and models are now widely available, but that the Dodgers and the Astros have invested the most in exploitation and application of the technology. And where are the Dodgers and Astros? They are at the very top of the league this year and in the past 5 years. It is no coincidence that the Astros and Dodgers each placed three pitchers on this year’s all-star team. They are the leaders in the application of big data to baseball performance and everyone else is playing catch up. I am a Giants fan, so it pains me to say this.
For those of us running businesses, we should think long and hard about this. If you aren’t using big data to engineer performance, your competitor likely is.
