Sabermetrics: A look at the statistics of innovation
May 18, 2012
By Dick Morley
The word sabermetrics describes the objective analysis of baseball. Evaluating a player and the team’s chances of winning a game are dependent upon many combined objectives. These include, but are not limited to, batting average, on-base statistics and runs created.
To win a game, the players must get on base. This is not the same as the batting average. And, by the way, the analysis is not the same for winning the game and winning the series. See the movie Moneyball for Hollywood’s take on sabermetrics.
During my college days, three major gaming scams were suggested by students: How to win at roulette, blackjack and poker. The movie made about blackjack (21) is fairly accurate. A couple of friends and I invested in the poker guys. We figured that three mathematicians with IQs of infinity couldn’t lose, and locally they did not. We were excited, but needed a big stake to make real money. Finally, we had enough to try Las Vegas — a big mistake. Our prideful smart mathematicians came back with empty pockets. They did not play poker in Las Vegas; they played the players. Vegas players could ascertain characteristics of your card hand by looking at indirect measurements of unconscious movements by the players. How fast you reached for a card somehow was evidence of your hand value. The Vegas players knew what was in our player’s hands by reading his “tells” — twitches, motions and other actions.
Angel investors use a form of sabermetrics. These statistics of behaviour and analysis have been used by the Breakfast Club, the group of angel investors that I’m a part of, for many years to select winning investments. A winning investment is similar to the “on-base analysis” of a baseball player. Knowing these numbers can give us a way to predict when the winning hand (or company) will be played.
Worldwide statistics for venture capitalists are about eight percent. This means that only eight percent of their investments pay off. They do not know which investment will pay off, but they can be sure that the ratio will stick. Some investment groups have a gateway statistic of 44 percent.
Angel gateway statistics are usually better than venture capitalists. Angels make small investments and are knowledgeable in the area of investment. For our Breakfast Club, 30 percent of our 100-plus investments were winners. This suggests that we are statistically assured that one winner will rise out of a group of five to 10 investments.
The chart accompanying this column shows the likelihood of a win with just 10, 20 and 30 percent on-base statistics for each investment partnership. These are generally viewed as significant innovations that make a company grow.
If you invest in what you think is a single good idea and no others, it is likely that that choice will fail. If, however, you invest in five or 10 chosen by the Breakfast Club — with a 30 percent success rate — it is extremely likely that you’ll find a winner rising out of that group. Is it important to look at pruning the group, hard and fast. Don’t try to rescue what you think is promising, but find out what not to invest in. The statistics will take care of you.
Recently, we set up investments in the process of innovation, not the products of innovation, for larger companies. A typical plan would be to allocate and manage, not by the project, but by the cost. Imagine, if you will, setting up five investments at $75,000 apiece on opportunities selected by the Breakfast Club. The expenditure is a fixed expenditure with variable time. Even the management of this effort is different. Most projects have a fixed time and “unlimited” expenditures. The object is to prune each of the five investments, and the one that refuses to be killed is the winner. But there’s only one. Each of the birthing efforts has a very small SWAT team managed by abandonment. The team has a fixed monthly budget to prove to the world that they can win; but only one can win.
This concept of budget management of small SWAT teams with fixed expenditure rates seems to work. Your columnist is the gatekeeper representing the Breakfast Club door. The ideas can come from anywhere. This is a way for the large company to innovate; not by buildings and analysis, but by using sabermetrics. The big company will win if you invest in the process, not the individual idea.
Baseball scouts find winners in high school players by looking at attributes. Being on base is more important than your batting average. Be the casino, not the gambler, and invest in the process, not the individual innovation.
This column originally appeared in the May 2012 issue of Manufacturing AUTOMATION.