I am not completely new to advanced statistics, but am nowhere near a master of sabrmetrics. I've done some research into them, and I understand them very well once I am familiar with the process. However, there are some metrics that I am unfamiliar with.
I was recently thinking about how prominent OPS and OPS+ have become while evaluating player's performances. I agree that SLG/OBP are much better than pure average and the counting stats when trying to determine value. However, it seems a little flawed to simply add SLG and OBP for a few reasons. They have different numerators, both count some of the same things, but also count some things differently (2B as 2 and 1 respectively, 3B as 3 and 1, etc).
Therefore, from a purely logical viewpoint, wouldn't it make more sense to add extra bases into the OBP equation? I understand there are metrics which have linear weights for singles, doubles, triples, homers, walks, and more. But I don't see any simple metrics that do this.
So let's call this TOP (Total On-base Percentage), with the equation for this stat being [(TB + BB + HBP)/(PA)]. I ran these figures for 14 of the 140 players who qualified for the batting title in 2013 - the 5 best OPS, the 5 worst OPS, and the 4 in the middle. I will list their TOP and their OPS.
The TOP's correlate with the OPS's pretty well on a relative basis, and this is expected because the numbers are reached through similar equations. The TOP of each of these players was somewhere between 57-67% of their respective OPS's. Also, there were a couple instances where out of 2 players, 1 had a higher OPS while the other had a higher TOP. The fact there were a couple of these disparages is a good thing, because it means that the 2 metrics have similar results (again, in relative terms), but they do not measure the exact same thing.
I don't have the time to evaluate every player, but I would expect the league average TOP in 2013 to be ~.460-.465. So, while Pablo Sandoval had an OPS+ of 119 last year, his TOP would be barely above league average. (Worth noting, Sandoval is almost exactly the median for OPS in qualified players, but well above league average in OPS+ because most of the players who didn't rack up enough PA's to qualify for that list were below average hitters). Now Miggy, Crush, Trout, etc. should still top the list of my new metric, and the usual suspects should round out the bottom of the list. But, as you can see with Sandoval, players who rank in the middle of OPS charts might rank significantly different in my TOP list.
There were several reasons for me doing this post. First of all, I would like to know if there is a current metric out there which does what my TOP stat does (in which case, I wasted a good amount of time haha). But whether or not I am the first person to do this (probably not), I want to raise the question - Should we use TOP to evaluate player performance, in a single-season or career context, or is OPS still better?
I currently don't have the time/mental capacity to evaluate this, maybe someone who reads this does/know someone who does. Maybe OPS correlates better to runs. Or maybe TOP is not reliable enough year-to-year. TOP makes more sense to me logically, but maybe my logic is flawed. Hopefully someone can confirm my thinking. But even if you disagree and have evidence (or better logic) to prove me wrong, I would like to hear that too.