I recently attended a session on how data changed professional sports, using the examples of baseball and golf. The room was an international mix of business leaders across industries, and the speaker's framing question set the tone for the whole session:  

“How many decisions  did your company make last quarter on a gut feeling that could have been made  on data?”

I took a lot of notes, and I wanted to pass on what struck me most, in case it is useful to you. None of this is attributed. It is my own synthesis, written to be shared freely.

Baseball: the decision culture changed in 10 years

• Spin rate, exit velocity, and launch angle, the statistics that now dominate every broadcast, were not measurable league-wide until MLB's (Major League Baseball) Statcast system went live in2015. Instagram launched in 2010. The modern statistics era of baseball is younger than an app on our phone.
• Before that, scouting ran on adjectives: "projectable frame, loose arm action." The same pitcher could get ten different grades from ten different scouts.
• What changed was accessibility. Prior measurement systems cost $20,000 to $30,000 and lived in stadiums. Then came a roughly $4,500 box, 11.5 pounds, that sets up in front of the mound in minutes and captures 10+ metrics on every throw. Feedback that used to arrive weeks later in a box score now arrives in the same practice session, and coaches coach players using it in real time.
• Adoption: 3 MLB clubs in 2016, 8 in 2017, 29 of 30 by a recent spring training, alongside nearly 500 college programs and over 400 accounts in Japan across high schools, industrial leagues, and NPB. Scouts now request a player's data file in contract negotiations. "Pitch designer" is a real job title that did not exist a decade ago.
• My favorite part: what this does for young players. Shohei Ohtani threw a 160 km/h fastball at Koshien as an 18-year-old from rural Iwate, a prefecture once considered a baseball backwater. Today the same device that sits in MLB bullpens sets up on a field like his hometown one, capturing the same metrics to the same standard. Students in Japan are already using this to get scouted into overseas university programs. Geography used to limit exposure. Now the numbers digitally travel.

Playing golf without a strategy is like playing darts blindfolded.

The golfers who use the data are also playing better games. Three findings worth knowing:

1. The "safe" club is a feeling, not a fact. Most amateurs hit 3-wood off the tee because it feels safer than driver. Mid-handicap players find the fairway 52% of the time with 3-wood and 49.7% with driver, basically identical across an 80-million-shot dataset, and a 15-handicap is more than twice as likely to take a penalty stroke with the "safe" club. Your dispersion pattern, not your ego, picks the club.
2. "Drive for show, putt for dough" is backwards. Tour pros make about 50% of putts from 8 feet and 40% from 10 feet; nobody is dramatically better at putting than the field. Scoring separation comes from driving and approach play. A golfer who takes this seriously practices the second shot, not the putt.
3. Statistics are a mindset tool, not just a strategy tool. If the best putters alive miss most10-footers, your miss was always part of the distribution. Treat a normal miss as a failure and you start fixing what is not broken: the swing change that was not needed, the new driver that did not need to be purchased.

The framework behind all of this, “strokes gained”, was invented in 2011 at Columbia Business School. Keith Mitchell was ranked outside the top 3,000 junior golfers nationally, used the data to drive his decisions, and climbed to number 3 in the world before turning pro. His mechanics did not change. His decision-making did. His words: "We looked at a hole as a statistical problem instead of 'this feels right.'"

What This Means in Practice

In most industries, the data is already there. HR tracks attrition, promotion rates, ratings, engagement. Sales has pipeline, conversion, client feedback. Marketing has funnels, active users, open rates. Finance has opex, utilization rates, revenue numbers. The open question is whether any of it reaches the moment of decision. Try this:

1. Decide in advance what the data gets to decide. "He's got great instincts. What a nice person." How often have we heard that in a hiring debrief? 5 interviewers, 5 different opinions, and we call the average of those feelings a decision. Before interviews start, agree what evidence would change your mind, and make the questions data-oriented: for a sales hire, not "tell me about your work ethic", but "what is your average number of client calls a week, and what is your conversion?" Same at year-end: define what a "meets expectations" means in observable terms before calibration, or you are not calibrating, you are averaging impressions.
2. Use the data to coach, not just to evaluate. In baseball, the data does not exist to evaluate the pitcher at year-end. It exists so feedback arrives in the same practice session, owned by the player, while there is still time to adjust. Most companies run this in reverse: employees meet their own numbers for the first time as a verdict in a review. The same conversion rate that feels like an accusation in December is coaching in March. Move the data into the regular conversation, and put it in the employee's hands first.
3. Expect the misses.Some percentage of good hires will struggle in the first 3-6 months. Some percentage of strong performers will have a soft quarter. If you expected that going in, a miss is not a crisis, and you are freed from fixing things that are not broken: the hiring process rewritten after one bad hire, the reorg after one rough quarter, the manager written off after one difficult conversation.

A Short Personal Take

I have seen many data-oriented leaders choose to ignore the data when it requires making a particularly tough decision. The dashboard governs every easy call.Then the numbers point at a well-liked colleague, a long-term client, a sponsored project, and the data quietly loses its seat at the table.

I have also seen this decision to ignore done thoughtfully. The data said one thing; we went the other way. The difference was that we admitted it (out loud), said why, and made the choice as a choice. We are allowed to override the data. The discipline is doing it in the open and the thing to watch is the pattern: if we override the numbers for the same person or client 3 times in a year, we have learned something the dashboard could never have told us.

If any of this resonates with how decisions get made in your organization, I would love to compare notes.

Thanks for reading, Diane