Dr. Giganto

Another cool thing was when they talk about replacing Damon and Giambi.

They figure that Damon's offensive numbers were actually average at best, and his major contribution to the team was in his center field defense. But, how do you quantify defense? All the widely used fielding statistics have major flaws. Fielding percentage, range factor, etc. All terribly inaccurate.

So, they hired a statistics company to figure out the probability of a run scoring at all possible game situations (leadoff batter, or bases loaded 1 out, or runners on first and third 2 out, etc.). So, they figured out the value of each thing you can do offensively by what it added or subtracted to the probability.

So, for instance, they figured out that the average number of runs that score in an inning is .55, but the average number of runs that score after you have a runner on second with nobody out is 1.10 runs. Thus, the value of a leadoff double is .55 runs (1.10-.55=.55)

Next, they tracked down the location and speed of all balls that were put into play for the last 10 years, and what the result of the play was (out, single, double, etc.). For instance, let's say a ball hit at a certain location at a certain speed is a double 95 percent of the time, a single 3 percent of the time, and an out 2 percent of the time. If Johnny Damon makes a leadoff putout in that situation, he's saved approximately .53 runs (95% of .55, plus 3% of however many runs a leadoff single is worth).

Then, they took all the catches that Johnny Damon made, and figured out how many runs he would save as compared to his replacement, Terrence Long. They figured out that Damon would save about 15 runs for the entire season. It doesn't seem like a big difference, but it kinda is.

There is a concept called Pythagorean W-L. Statistical analysis has shown that a team's winning perecentage can be reasonably approximated using the number of runs they've scored and how many they've allowed. A team with a lower actual W-L record has been unlucky, and a team with a higher actual W-L record has been lucky.

Oaklan'd Pythagorean W-L that year was 104-58 (they actually finished 102-60, so they were a little unlucky). By adding those 15 runs that Damon saved, the Pythagorean W-L goes down to 102-60, which means that when you add up all the little bits and pieces of runs that Damon saved, his defense and his defense won two more games than they would have won if Terrence Long (not a terrible fielder) had been playing CF

It doesn't sound like a lot, but also remember that you have to add the defensive contributions from all other players, as well as hitting and pitching contributions (so, you have to add Giambi, Tejada, and Damon's bats, as well as the pitching wins contributed from Zito, Hudson, and Mulder, which make up the bulk of the 104 wins). Plus, two wins can easily make the difference between making the playoffs and playing golf in October.

Then, they thought about offense. They realized they could never replace Giambi's bat, but then they realized that they were replaciong three players: Giambi, Damon, and Olmaedo Saenz. Those three combined to have an OBP of .364, so they needed Terrence Long and two additional hitters to have an average OBP of .364. So, they signed David Justice, who was old but still could draw walks, and Jeremy Giambi, who was a much worse hitter than his brother, but had a decent eye.

As a result, they actually had a BETTER record the following year, at 103-59 (their Pythagorean W-L was much lower...96-66.) and made the playoffs once again, without Giambi and Damon.

A little nerdy, but I find it fascinating