I'm not laughing... But it is funny

On August 17th, 1957, during a game against the New York Giants in Philadelphia, Hall of Famer Richie Ashburn hit a foul ball that struck Alice Roth squarely in the face, breaking her nose. The game was then paused as medics came to tend to Roth in the stands.? As they were carrying her away on a stretcher, play was resumed and Ashburn fouled off another on the first pitch thrown to him. This foul subsequently struck Roth as she was being carried off by the medics.

Now, I feel like a horrible person because I laughed when I read this. The first thing I pictured was both the medics and Roth’s husband, a sports writer attending to her, and Alice, trying to understand why Richie Ashburn was trying to kill Alice, and how long he must have had to practice hitting such precise foul balls to carry out his diabolical mission.? This is the only logical explanation, you would think, because the statistics of Richie being able to do that to Alice just out of mere chance have got to be 1:100,000,000 odds.

Of course, it was just a coincidence, but like I said, the odds are staggering. Yet, it did happen, so why does it feel almost statistically impossible? The reason is that by being in that situation, you would be thinking “What are the odds that this could happen to me today?”?

If you went up to Alice on the morning of August 17th, 1957 and told her you wanted to bet $10,000 that she would be hit by a foul ball twice, in a row, during that afternoon’s game, she would have taken that bet, anyone would.? She’s been to many games and not only had that never happened to her, but she had probably never heard of it happening to anyone else either. “Come on twice in a row!” But let’s reframe this incident in a different way. What are the odds that, in the entire history of all sports where a projectile is thrown or kicked, a spectator could be hit twice in a row within the same game?? That seems like something that has surely happened in all of human history. It was just a matter of it happening at a game that was documented in some manner for us to discuss it.

We just described the same event “hitting a spectator twice in one game” in two very different ways, which results in two very different statistical models. But it is the same event. There are three differences between the two characterizations of the event, time frame, a single sport vs all sports, and it being Alice who was hit vs. anyone in all of human history.?

We tend to frame unlikely events that we observe in a very myopic way. This framing makes the odds of the event ever occurring seem very unlikely. This is human, “Why is this happening to ME?”?

Is there a lesson here?? Yes very much so in product development. I see this all the time within product reliability test programs. It is a dangerous perspective to have. Why is it dangerous? Because when we observe failures in test we tend to dismiss them as anomalies or unlikely to happen in real life, the field.? “What are the odds that this combination of factors would occur in the field exactly like that?” We deem it unlikely, the observed issue gets dismissed, and we move along staying on schedule as best we can.

?What the teams aren’t realizing is that there are 10,000 types of failures that can occur in the product with all the different combinations of factors and enough time. Not just the one they observed, i.e.? Alice. Alice only became Alice because it happened to her.? That failure mode became “the issue” just because? it was the one you were lucky enough to get to observe before release to the field. All the rest of the failure modes are waiting out there for the customer to discover. And the probability of them being found, very likely, because it’s a much larger time frame and there are thousands of variabilites waiting to interact.?

Each failure found in test is a gift and should drive the pursuit for understanding the variabilities that caused it. Those thousands of failure modes in waiting will be driven by similar variabilities.? All they need is time to be the next Alice.?

-Adam