The issue here started when Tom claimed BAVG was a good stat by itself. He used OBP to be the same, with the BAVG different. It that case, then BAVG by itself says that one hitter was a .270 and the other was .310. We could easily use .270 and .310 as their OBP and the higher number is always going to be better. But let's look at the way it he meant it.
Player A = .310 BAVG .335 OBP .375 SLG .710 OPS
Player B = .310 BAVG .395 OBP .505 SLG .900 OPS
Batting average says that both hitters are the same. So by itself, BAVG tells you nothing other than a level of eye/hand coordination it takes to hit .310. However, the difference between a batter's BAVG and OBP can be predictive. Meaning, the greater the difference, the last apt the player will be affected by a hitting slump.
Additionally, as I stated before, counting stats (R, RBI, HR, SB, etc) and certain rate stats (BAVG, ERA, WAR) are just plain terrible. They can rarely be predictive in nature nor can they be could comparison stats. Stat that take into account Linear Weights (i.e. Runs Created, wOBA, etc) are better comparison and predictive stats.
