Getting Crunched by the Numbers

In our age of data analytics, with all of our powerful methodologies and computational power, it sometimes feels that we merely need to turn on the analytical engines and sit back and wait for the discoveries. It all boils down to number crunching. Or does it?

Often there is more to a story than portrayed by the data. The data can mislead us. Unleashing our data analytics tools to gobble up the data can set us up to get crunched by the numbers.

That’s why we need to use our intuition and perspective taking to get a sense of what’s behind the numbers. We need to know the intentions of the decision makers in order to make sense of their performance.

Here are three examples:

Example 1: Baseball.

In game 1 of the National League Division Series, October 3, 2019, the Los Angeles Dodgers started a 25-year old, Walker Buehler, a left-handed pitcher. He had a very good record this year, 14-4. However, his last few outings were terrible. In his previous 16 innings he gave up 8 runs and walked 9 batters. For those who don’t follow baseball, these are very mediocre data — not the reassuring performance you’d want in a pitcher to kick off your post-season. Several commentators questioned whether Buehler was up to the job. But the data were misleading, and Buehler did an outstanding job in that first game.

When asked about it afterwards he explained that the Dodgers had clinched their division leadership very early and he decided to use his final appearances to tweak his pitches and try different things. He didn’t particularly care about his record for the year. Instead, he wanted to hone his skills for the playoffs. So those last 16 innings didn’t mean anything. If you didn’t know his intentions and just tracked the data, you’d have drawn the wrong conclusion.

Example 2: Rowing.

My younger daughter Rebecca rowed in college. During the winter, crew teams didn’t compete in the open water in New York, but instead held competitions indoors on rowing machines.

In one such competition, Rebecca’s team matched up against another college and Rebecca quickly spotted the star of the other team, a young woman who was much taller than Rebecca (and height matters a lot in rowing). I described this incident in my book Streetlights and Shadows (2009), but the gist is that Rebecca realized she’d have a tough time beating her opponent so she altered her strategy. In a 2000-meter race, the typical strategy is to do a final all-out kick for the last 200 meters. Rebecca expected that her taller opponent could out-do her in that final kick, so Rebecca decided to start her kick with 500 meters remaining, and to preserve enough energy to pull that off.

The race started and the opponent jumped out to a large lead. Rebecca thought that if the opponent could keep that pace up for the entire race there was little chance of catching her, and it wasn’t worth trying. But Rebecca had a hunch that the opponent was going to slow down after 500 meters, and so Rebecca just kept to a comfortable pace. Sure enough, the opponent did slow down and by the half-way point, 1,000 meters, Rebecca had closed a lot of the gap. At this point, Rebecca envisioned the rest of the race — at the pace they were rowing she’d pull even by 1,500 meters and then start her kick, taking a commanding lead and holding it. She was pretty sure she was going to win the race.

Rebecca’s teammates were oblivious to Rebecca’s strategy and intentions, and at the half-way point they were shouting encouragement to her not to give up and that she had a good chance for second place. Rebecca remembers being amused to think that she was the only person in the gym who appreciated that she had the race pretty well sewn up. No one knew her intentions to start her kick with 500 meters to go. No one knew that she had preserved her strength to carry out that plan. So, without knowing Rebecca’s intentions, you’d have no way to anticipate what came next — the easy win Rebecca had arranged. And, yes, Rebecca was correct that her opponent had a stronger kick than she did, and if Rebecca had followed the typical strategy of waiting until the last 200 meters to kick, she’d have lost the race.

Example 3: Helicopter navigation.

Many years ago, my colleagues and I did a research project on Army helicopter teams and how they coordinated during a mission. We had a chance to observe a simulator training exercise. There would be ten 2-person crews navigating a difficult course, maneuvering around simulated hills and dodging simulated missiles. I was stationed in the simulator control room where I could watch the action on a set of monitors. My two colleagues embedded in the ready room so that they could interview the members of each helicopter crew as the crew prepared for its turn. I vividly remember one crew — it was composed of two women, and the men in the simulator control room had very low expectations for them. Sure enough, after the first few legs of the mission the female crew seemed to get lost. Instead of turning west after the first set of hills, they kept going straight and only realized their mistake when they got to a second set of hills and turned west after passing these hills. The comments I heard in the simulator control room were not kind.

The observers were surprised when the female crew made up the time they lost and finished with a strong overall performance.

And then it was my turn to be surprised. I mentioned this crew to my colleagues, and how it had gotten lost, and they told me, “Oh no. They never got lost. That was their plan. They looked at the map showing the simulated enemy anti-air batteries and decided that by going around the second set of hills they’d be shielded and could move really fast, rather than ducking missiles all the way.

Because I didn’t know their intentions (nor did the others in the simulator control room), it was so easy to draw the wrong conclusion.

All of these examples illustrate how critical it is to understand the intentions of the decision makers. Otherwise, just relying on the data, we can draw the wrong conclusions. We can get crunched by the numbers.