Baseball Prospectus 1996
" Knowing what has happened is the most important part of knowing what's going to happen. " - Adam Smith
Economists aside, there is some truth to that statement. For about 15 position players in each organization, we have provided a projection of how that player will do in the coming year. I developed the projection system after studying Clay's Davenport Translations and some artificial neural network software. The projection system is called Vladimir, for reasons that have to do with just the right amount of garlic, a bit of melted mozzarella cheese, and rapid delivery service. For now, let's refer to these one-line bits under a player's name as Vlads.
What are Vlads?
Vlads are a projection of what the player's performance will be in the next year, normalized to a neutral park in the National League in 1995. For more information, read the explanation of the Davenport Translations.
Why should I care?
Well, it might help you win a few bucks or bragging rights in your local rotisserie or Strat-O-Matic league. Perhaps it will keep you from spending $3 million of your organization's cash on a bad risk. It depends on who you are. At the very least, you can take them with a grain of salt, laugh at them and publish them to embarrass us at year's end.
How come there's so little information in the projections?
In other words, 'How come they don't project RBI?' In short, because RBI are an antiquated, pathetic, worse than useless measure of an individual player's performance. An individual's RBI total is more revealing of who he hit behind than of some sort of skill the player has. I mean, Joe Carter's been racking up RBI hitting behind people like Brett Butler, Tony Gwynn, Jack Clark, Roberto Alomar and Paul Molitor since the Mesozoic era. He's an above average player, but be realistic - he's not really as good as the press, or most fans, think he is.
The important thing to remember about the Vlads is that they are not projections of actual statistics, but of translated statistics. That is, if they project Dante 'Lottery' Bichette to hit .250/.290/.400 next year, and he puts up something like .290/.340/.500, then it wasn't far off. Vlads are already park- and league-adjusted. A guy might have to hit .390 in the Pacific Coast League to reach a Vlad- projected .300 BA.
As for playing time - don't make the mistake of assuming that Vlad thinks all the player's playing time will come in the big leagues. Far from it. Playing time is the hardest thing to project, and it definitely falls into the wild guess area.
How do the Vlads work? Are they accurate?
The Vlads are three years old, and they're surprisingly accurate. The system is based on an artificial neural network which has been trained by studying the career path of almost anyone who's ever played. If I ran a Vlad for a player, it's in here. Unlike some other more hirsute and slender writers, I'm not going to hide behind my own biases and try to weasel out of responsibility. In 1994, Vlad spoke from on high and told all the world that Andy Van Slyke would return to form in '95, and hit like a beast. I knew this to be wrong, but published it anyway. (I'm still convinced it was a Pentium floating point bug.)
The original Vladimir software was destroyed in a fire on June 27, 1995, along with the original work on this book. It happens. I hope you enjoy the Vlads, and feel free to throw them back in our face come the offseason. Be sure to include a few choice words about our intellects or parentage, and we'll mention you in next year's edition. Thanks for buying the book.