It was a “Dewey Defeats Truman” moment.
Shortly after the polls closed on primary election night in New York, CNN made a bold prediction. Its exit polling showed Hillary Clinton and Bernie Sanders locked in a tight Democratic primary race. Clinton’s win would be close, Wolf Blitzer said: 52 percent to 48 percent.
Less than an hour later, that prediction was as laughable as the famous November 3, 1948 Chicago Tribune headline announcing that voters had elected Thomas E. Dewey President of the United States.
Statistically, the CNN call was far worse. In the end, Truman beat Dewey 49 to 45 percent. Clinton won New York — 58 to 42 percent.
When the News is News
One interesting aspect of the CNN mistake is how quickly it disappeared from public sight. That’s because all major media outlets use exit polling to predict results as soon as they can. First-predictors are the first to attract viewers. There’s no incentive for any of them to throw mud on a process that they all use as a marketing gimmick.
Another aspect is the paucity of discussion over what went wrong at CNN. I don’t know the answer, but this article isn’t about that. It’s about the real lesson of the episode: The use of statistics can be a perilous exercise.
Law Schools
Data are important. It’s certainly wise to look at past results in weighing future decisions. But it’s also important to cut through the noise — and separate valid data from hype.
For example, if less than one-third of a particular law school’s recent graduates are finding full-time long-term jobs requiring a JD, prospective students are wise to consider carefully whether to attend that school. But it becomes more difficult when some law professor argues that the average value of a legal degree over the lifetime of all graduates is, say, a million dollars.
It’s even more challenging when law deans and professors repeat the trope as if it were sacrosanct with a universal application every new JD degree-holder from every school. And it sure doesn’t help when schools with dismal full-time long-term JD employment outcomes tout, “Now is the Time to Fulfill Your Dream of Becoming a Lawyer.”
Law Firms
Likewise, based on their unaudited assessments, leaders of big law firms confess that only about half of their lateral hires over the past five years have been breakeven at best. And that not-so-successful rate has been declining.
Law firms are prudent to consider carefully that data before pursuing aggressive lateral hiring as a growth strategy. But it becomes more difficult when managing partners seek to preside over expanding empires. And it doesn’t help when law firm management consultants keep overselling the strategy as the only means of survival.
Data should drive decisions. But the CNN misfire is a cautionary tale about the limits of statistical analysis. Sometimes numbers don’t tell the whole story. Sometimes they point people in the wrong direction. And sometimes they’re just plain wrong.
The Belly of the Beast 
Past performance is not predictive of future results, as lawyers must surely know. If using attorney statistics is intended to focus the firm’s recruiting on the candidates most likely to contribute to the firm’s bottom line via new client revenues, then the calculus misses critical inputs. The real problem is that the data used does not measure attorney/client relationship dynamics or client motivations and incentives. In other words, stats being used are attempting to predict lateral success in a new platform without inputs for how the clients benefit from the move. As such, I suspect the lack of client focus in law firm’s lateral analysis produces a distorted forecast. Certainly, billing data, behavioral scores and other such data is useful in some regard and over the long run should be included in the data. But without an objective focus on the client relationships and monitoring the markers that can predict movement, firms will continue to grasp at the promise of growth and not actually manage it.