BP Daily Podcast

I wonder why Dan asks his question. Only because I assume he has to know the answer. I'd love to hear what you think, Dan.

Unlikely events are likely to happen but it's just difficult to peg exactly what that event may be. It goes for life and it goes for baseball. At the beginning of the day, what were the chances that I would stumble upon Michael Doleac's Wikipedia page? Pretty low. Yet, everyday we all find ourselves doing things that would have seemed unfathomable at another point in time.

Baseball works the same way but with tighter results due to the possibilities of a ball game being smaller than the possibilities of a 24-hour day. Like Sam and Ben said, the chances that Chris Davis hits 74 homers this year is pretty low. That being said, surely something as obscure will happen this season. Computer projections work with big data - they cater toward the larger middle and ignore the tails. To be 100 percent successful you need to account for random fluctuation - something based on luck alone. It's not efficient.

A scout isn't a computer and focuses more on nailing a single projection. Scouts don't have to deal with the massive sample sizes of simulations. He's (or she) is projecting a single player's career. And the nature of this prediction is more rash when one considers both the human element and the tighter confidence interval around a player's career versus an individual season.

Ultimately, the answer to the question is analysis of big data vs. analysis of not big data.

These podcasts are just about the only thought-provoking baseball material I can get my hands on these days. As always, keep it up.