OPS

From Wiki Gonzalez

On-Base Percentage
Plus 
Slugging Percentage

OPS is a simple but flawed metric calculated by simply adding OBP and SLG. While it's a fairly accurate measure of offensive value (i.e., it correlates to actual runs scored), it underrates the value of on-base percentage.

For example, a team that always walked (OBP: 1.000, SLG: .000) would have the same OPS as a team who hit a triple every four at-bats (OBP: .250, SLG: .750). The first team scores an infinite amount of runs, while the latter does not. (Note: See TangoTiger's article for examples closer to real-life.)

It has been suggested that this is because OBP is on a scale of 0-1 while SLG is on a scale of 0-4, which is clearly nonsense, since then we'd all know 4*OBP + SLG would be the natural reweighting. GPA, which uses a different weighting, is probably the best of these simple formulas.

For further information, see the two-part series by TangoTiger, "OPS: Be Gone!" (part one, part two) on the inaccuracies of OPS. Aaron Gleeman cited Tango's article when he introduced GPA.

There was also an excellent article by Dominic Rivers, "How Orlando Palmeiro Got Hosed: Exactly How Full of S is OPS?" Like Tango, Rivers also attempted to find a best-fit multiplier for OBP so the modified-OPS figure would more accurately reflect the player's offensive value. However, Rivers used actual year-by-year data and found that the modifer changed depending on league averages.

Rivers's article was discussed on a January entry on David Pinto's blog, Baseball Musings. Unfortunately, the original link to Rivers's article is now stale.

Allen Barra and George Ignatin developed their SLOB (SLG + OBP) system in the mid-70's. What we know today as OPS was invented by Dick Cramer, and popularized by Pete Palmer -- it was first mentioned in Palmer & Thorn's Hidden Game of Baseball under the label of PRO ("Production").