This is the “weather tomorrow will be like the weather today” theory extended to entire seasons. Suckitude begets suckitude. It worked spectacularly well for predicting the M’s in the 80s, and seems to be in play once again. At this point I’m looking for the M’s having the highest payroll per win (which is going to be hard with the Royals paying so much money to Meche). But hey, whatever it takes to bring the Hargrove/Bavasi era to an end (and hopefully ashcan some people higher up the chain too).

(BTW, Evan, your story is apocryphal, or at least overly specific. Until the late 20th century “the weather tomorrow will be like the weather today” was, when averaged over time, the most accurate prediction possible not just in the arctic but for every location on the planet. The Atmoshperic Sciences department at the UW has run a weather predicting contest for decades, and the old professor who won it regularly used that system [defeating countless grad students with countless other systems]. It was only in the late 1980s that computer models got consistently better than that for 24 hours out; since then they’ve been extending their reach both further into the future and into finer detail on the ground, which is important for places like Puget Sound where geography can have such a huge effect on the microclimate. But if you don’t have a computer model or accurate data over a large volume of the earth’s atmosphere, “tomorrow will be like today” is still the best guess.)

Unless you think we already make things up and have no credibility, in which case I’d wonder why you were here.

Derek, can you just make up something and post it as news? Just to get things stirred up around here. Something like “Bavasi trades Jose Lopez (or Adam Jones, or heck Felix!) for ________ (insert old, broken down expensive player here)” ought to get those server lights blinking.

The college I went to used to print a “jack issue” every once in a while full of somewhat believable lies. It was always good for an initial shock and some laughs as you watched your peers glance at the headlines and do a doubletake.

Or are you wrong about the last part thus negating your comment entirely? Oh the logic!

Ah, so there’s the interesting question – do people look at projection systems and side with the ones they agree with,

That would be kind of sad wouldn’t it?

do people weight PECOTA more than three-year averages?

My projections are based heavily on a story I heard about natural gas production in the high arctic.

So, this oil company develops some extremely sophisticated software to predict the local weather, because the weather can get really awful and they’d like to know about that in advance (and no one else is really predicting the weather with any accuracy up there because no one else ever needs to know). So, this software turns out only to be right about 70% of the time, but the tech guys learn that there’s an operator up there who predicts the weather with 85% accuracy, and has been for years. So they head up there to ask him how he does it.

And he says, “The weather up here doesn’t change much, so I always predict it will do tomorrow exactly what it’s doing today. When the weather does change, I’m wrong, but then I’m right again a bunch of days in a row as the new system settles in.”

That’s basically what I’m doing. I expect the offense overall to be a bit better than 2006, so I’m adding a bit to each player’s AB total (because more runs means more plate appearances), and then reproducing rate stats as closely as I can, with slight adjustments based on whether I expect players to improve or decline. Johjima’s walk rate, for example, was extremely low in 2006, so I bumped that up a bit, but otherwise I expect him to be the same hitter he was last year.

I am not so sophisticated that I then reduced Johjima’s AB total by the number of extra walks I predicted over the pro-rated amount. Perhaps I should have.

Ah, so there’s the interesting question – do people look at projection systems and side with the ones they agree with, or do they reconsider their position if they find their opinion disagreed with?

That is the key to how the Wisdom of Crowds works (when it works). If people were simply blindly taking an average of the projections, then we could make projections that were just as good as the Crowd’s, by doing the same thing.

But each individual in the Crowd is not simply taking an average, they’re looking at the information available, including the expert projections, and then making the Key Crucial Decision: go with what those projections say (i.e. changing their mind) or decide that the expert projections are off-base and either tweak them or use a projection that’s entirely their own.

Some individuals will be swayed hugely by the projections out there, some will not be, and the hope is that overall they are making that decision at least somewhat wisely based on their own knowledge, intuition, information, etc. Wisely enough so that their tweakings/non-tweakings hopefully add additional information to the overall average.