Hi there, Beyond the Boxscore readers.
Recently, I've embarked on a journey to normalize the different WAR stats (found at Baseball-Reference, FanGraphs, and Baseball Prospectus) to a common, equivalent measure. Though the stats attempt to measure the same thing (value over replacement-level players), they don't always align in terms of numerical value. fWAR tends to run "high", WARP tends to run "low." That's a massive oversimplification, but it's close enough for now.
My goal is two-fold here:
(1) If we can align the three major WAR metrics to a similar baseline, we can more accurately define the differences (deltas) between player scores. This way we can see where the biggest outliers are, from one system to the next. Though (as our Adam Darowski has pointed out), the example of pitching WAR from Baseball-Reference and FanGraphs measure VERY DIFFERENT THINGS -- this might be a more useful exercise for position players.
(2) A WAR index (tentatively titled WARi) -- a number that sums, then averages the three scores -- could certainly be useful to baseball fans who are not as immersed in the saber community as the rest of us are. Instead of using one of the three metrics, we're creating a snapshot of the saber community at large. I like to compare it to the BCS system. Very few college football fans care about the math that goes into the development of a particular computer score -- but the rankings themselves can be very telling, and may be more relevant to those who are not neck-deep in analysis.
My question here is -- what do you think the value of this exercise is? Pointless? Useful?
What we need now is input from the community. In our comments section below, feel free to weigh in on the discussion.