Trouble with the curve: Is 'Moneyball' fading?

1

BaseballThere’s been quite a lot of chatter lately about my Boston Red Sox and their recent shift ‘away’ from using analytics or ‘sabermetrics,’ as data science is often referred to in baseball (Jeff Passan, one of my favorite baseball writers, chimes in here – Forbes also commented that the Sox are ‘sending Moneyball to the showers’).

I feel this is being a bit exaggerated. The Red Sox still employ the father of sabermetrics, Bill James, and have an impressive analytics team. But talking about this issue can provide insights into sensible organizational analytics approaches. It can also provide lessons for dealing with the inevitable backlash caused by Big Data hype and unrealistic expectations for how quickly and how much analytics would transform business processes. There are two angles we need to examine.

First of all, I agree that the Red Sox need to find the right balance between sound analytics and strong scouting programs. Every organization needs a strategy based on analytic insights balanced with strong domain knowledge -- that knack for knowing what breeds success. Some might call this ‘gut feeling’ and that wouldn’t be a bad thing. Sometimes risk taking or ‘bucking the trend’ is precisely what's necessary for taking performance to the next level.

That said, when gut feeling pushes a decision outside of what the numbers are telling you, analytics should be used to make you acutely aware of how much you're bucking the trend. This allows for a quantified approach to risk that can be balanced with other, perhaps more conservative decisions. So, even when business acumen is telling you to go against the trend, analytics can help there as well.

Let’s take a recent controversial player signing by the Red Sox as an example. Pablo Sandoval was a core piece of the San Francisco Giants recent World Series successes. When he became a free agent after the 2014 season, Red Sox analysts looked at Pablo Sandoval’s history and saw a mid-range defensive 3rd baseman who got a lot of hits. On paper, it looked like a no-brainer that the Red Sox should flex their financial muscle and go out and grab the star.

However, scouts look at Sandoval and see a guy who swings too much and is overweight. Scouts will call these signals out as ‘inevitable regression’ indicators (yes, scouts know something about statistics as well). Does the scout trump the analysis?

Well, I propose a healthy balance of both analytics and scouting;  both are core elements to baseball success. If you ask me, Sandoval’s signing was based on an outdated over-adherence to pure-play numbers evaluation, resulting in a classic overpay, but even that could be an ok strategy if it's in line with the rest of team spending. In short, the right hand (analysts) needs to communicate with the left hand (scouts).

Second, despite the Moneyball hype, we’re not even close to everything we can do with baseball analytics. Big Data, the Internet of Things and sports sensor companies all mean that organizations will be able to collect more and more player and team data to get to the core of what drives performance.

On top of that, ‘subjective,’ unstructured data such as scout reports, medical records, doctor’s notes and coach reports will all start to be brought into analyses. Analysts and scouts will need to work together to reap the benefits of this data windfall. Scouts will have ‘big data’ environments with approachable analytics where they can look for pockets of potential in the growing data stores and pass those insights back to the analysts.

Success in cultivating this feedback loop will be the next area where teams will tilt the competitive balance, much like the Billy Beane and the Oakland A’s were successful doing in the late 90’s, early 2000's. After a few years, everyone caught on with what the A's were doing, all teams hired analysts and it was no longer a competitive edge. It was simply a cost of doing business (If you haven’t read Competing on Analytics by Tom Davenport, that’s his basic premise…that organizations will use analytics to find temporary areas of competitive advantage until others catch on - then that particular brand of analytics simply becomes a cost of doing business).

In baseball, as in most businesses, business acumen can be successfully combined with data analytics to drive competitive innovation. On the one side, 'approachable' technologies such as SAS Visual Analytics and Visual Statistics will allow business-saavy experts to get their feet wet with advanced analytics.

This not only means the business users may find great insights, but that they can easily share those insights AND be less intimidated by the results indicated by their data scientist counterparts. On the other side, competitive edge will continue to develop by using new data and new inclusive algorithms to even better quantify on-field and team construction success factors.

If you want to watch a good flick about the flip side of the Moneyball effect, check out 'Trouble with the Curve.' It may not solve your profit/loss questions, but it will reassure you that baseball players, and thus people, cannot be entirely reduced to numbers...yet!

If you just can't get enough of sports analytics, or want to explore more of the similarities with your analytics intiatives, check out All Analytics Radio on Thursday, March 10, at 2 p.m.  Tom Davenport will be presenting: Sports Analytics: Driving the Business Forward.

 

Share

About Author

Andrew Pease

Principal Business Solutions Manager

After 14 years in various roles at SAS, Andrew is currently responsible for advanced analytics in the Center of Excellence. Andrew helps financial institutions, major retailers, pharmaceuticals, manufacturers, utilities and public sector to understand and use powerful analytic techniques such as decision management, predictive modelling, time-series forecasting, optimization, and text mining.

1 Comment

  1. Pingback: Your roundup of breaking analytics news - Industrial IoT/Industrie 4.0 Viewpoints

Leave A Reply

Back to Top