Ask any soldier how long a battle plan lasts, and he’ll tell you that everything changes as soon as the first shot is fired. Considering this rather obvious condition of combat, the idea of “precision war” screams of oxymoronic hubris. Of course, that didn’t stop Donald Rumsfeld from believing that our military had become so data-driven that Iraq could be fought and won with such exactitude that we would hardly lose a soldier or hurt a civilian. Obviously, this turned out to be preposterous arrogance and folly of the highest order.
War is messy, and even the weakest opponent has a say in its outcome. No amount of data can anticipate the seemingly tiny ground level decisions that can turn the tide of a battle or even the entire war. No algorithm can stop a soldier from being ambushed by insurgents or startled in the dark of night, causing him to mistakenly shoot an innocent. The variables in war are impossible to measure, and only a fool would think that it could be sterilized and made precise by the calculations of computers sitting far away from the front lines.
Entrepreneurship, like war, is also messy. And those who have engaged in the fight know that it takes persistence, adaptability, quick-thinking, and a fair amount of luck in order to win. But increasingly, venture capitalists are following Donald Rumsfeld’s lead and basing their decisions not on an entrepreneur’s ability, but on algorithms much like a FICO score.
For example, recently launched Correlation Ventures prides itself on the fact that its investments are determined by its predictive software which includes such variables as how many VCs sit on the board. This strikes me as somewhat blind to put that much faith in the predictions of the data.
Much like Donald Rumsfeld was led astray by his unwavering belief in a data driven war, and disregarded the value and importance of our military leadership, basing an investment on questionable data inputs seems to ignore the importance of the skill and character of the entrepreneur. Apparently, entrepreneurial talent matters little in this new world of venture capital as long as the computer says it’s a good bet.
Those who claim that algorithm based venture investing is like “Moneyball” are either woefully naive or just covering up their own lack of due diligence. Compared to launching a new company in an uncertain market, a baseball game has a very limited set of statistics to measure.
This is why predicting the winner of a ball game is much easier than predicting the trajectory of the stock market. Even an amateur can collect a few stats and pick baseball game winners at a much better rate than .500, but aside from Warren Buffett, beating the stock market on a consistent basis is a total crapshoot. There are simply far too many variables in the business environment and drawing comparisons between the two is as ludicrous as saying a bookie is the same thing as a good investment manager.
Of course an algorithm can be easily rigged to show consistently accurate results if it’s measuring macro inputs, but this would be like Rumsfeld asking a computer if the United States would beat Saddam’s army. Of course we would, that part was never in doubt. Such a question is akin to asking if you should invest in a company whose cancer treatment has shown 99 percent positive results in lab tests. An algorithmic decision would always be correct because the outcome is obvious.
I suppose algorithms are at least an improvement over the current trend of pattern-matching which seems to be predicated on finding college dropouts wearing hoodies, but ultimately both methods are poor substitutes for due diligence into the product and the abilities of the entrepreneur.
Donald Rumsfeld learned the hard way that war is never precise, and all the data and number crunching are no replacement for decisions made on the ground in the heat of battle. Rumsfeld’s data didn’t tell him how resilient the insurgency would be, how deep the religious divisions cut, how many civilians would die, and what personal toll the war would take on our troops returning from multiple tours of duty.
His precision war now looks laughably imprecise and his arrogance absurd. Venture capitalists basing their decisions on data and algorithms are being similarly overconfident, and yet the industry continues to move in that direction. Strange how VCs, most of whom are very smart people, have become as misguided as Donald Rumsfeld.
[Illustration by Hallie Bateman]