I loved Bill James when I first encountered him in the early 80's. He wrote like an angel, he was funny and he saw a common thing in an uncommon way.
But sabermetrics as a genre, as a metier, as the lens through which to observe the game...eh! I realized that, for me, baseball was 'about' Don Larsen striking out Dale Mitchell and Lew Burdette's compact with Satan and Jerry Coleman fumbling Eddie Matthews' chopper and a ground ball inexplicably finding Tony Kubek's throat and Maz taking Ralph Terry deep and then, cosmic justice, Mac's liner finding Bobby's glove (yeah, I was a Yankee fan as a kid...). It was about moments not numbers.
Above all, it was about Hector Lopez. During his Yankee years, 59-66, he never hit more than .284 with modest power, low walks , high strikeouts and an unseemly number of gdps. But let me tell you friends and neighbors, if the game were on the line, the very last person you wanted to see at home plate (assuming you were the opposing pitcher) was Hector Lopez. He was the essence of clutch. He was Marco Scutaro.
Sabermetrics didn't catch the moments. We parted ways. If WAR and OPS were people and I saw them at a party, I'd shake their hand and leave them be.
All of which is a perhaps excessive prologue to my own humble contribution to what I construe as sabermetrical thinking. I offer for your consideration a simple equation:
X/y = z
X = the number of players who can reasonably be expected to significantly improve their performance
Y= the number of players who can reasonably be expected to significantly decline in performance
Z = Predictive Index of Team improvement or PITI
Let us take the 2010 Giants as a our first case in point. Wilson has just whiffed Cruz. What is this team's PITI?
X = Posey + Bumgartner + Schierholz + Sandoval = 4
Y = Huff + Uribe + Torres + Burrell + Sanchez =5
Z = .8
Not all these decisions are clearcut (and, of course, hindsight is 20:20). Arguably, you could take the Huffster out of the denominator; after all, the Giants resigned him for a lot of money so they thought he could break his even-odd year cycle and repeat his exploits. And we didn't know that Freddie's breakdown would accelerate, although the warning signs were certainly there.
Two years later, with benefit of hindsight, we would add Wilson and Lincecum to the denominator and subtract Schierholz from the numerator, lowering the PITI to .42. That team with those players went backwards and had no chance of competing for the throne.
Now let's transport ourselves to Romo undressing Cabrera. What is the PITI at that moment of this team?
X = Belt + Crawford + Sandoval + Sanchez + Pence
Y = Posey* + Scutaro*
This PITI of 2.5 reflects, I would argue, a very cautious prognosis, carefully designed to fly under the hubris radar of the baseball gods. Is it statistically likely that Posey's 2013 will fall short of his MVP 2012? Was it statistically likely that Posey would recover from knee destruction, get stronger as the season progressed, lead the league in hitting, win the MVP and lead his team to the championship? Which is less likely? It could be that we have an incredible player for the ages here who will be truly great for a long time. Discuss among yourselves.
As for Scutaro, I have two words for you: Hector Lopez. This is a nice piece about Blockbuster: http://bleacherreport.com/articles/1473495-marco-scutaro-will-continue-his-success-with-san-francisco-giants-in-2013 It is oriented to his 'make contact' and low strikeout numbers, but Marco's been clutch since he came up with the A's and now he's in the exact right place on the exact right team. He should be fine, with little falloff.
The names on the X line...can anyone argue with them being there? If Hunter Pence hits .260, he will drive in 125 runs. He's a 290 career hitter. And there are only two hamate bones, correct?
Other teams may, conceivably, 'out-improve' the Giants, but there's every reason to expect the Giants to be better.