From BR Bullpen
The Bases Fallacy is a common flaw related to a number of baseball statistics, including Tom Boswell's Total Average, Barry Codell's Base Out Percentage, and countless others. The distinguishing characteristic of this fallacy is the process of treating all bases gained by a batter equally. A basic formula suffering from this problem would be:
Bases per PA = (TB + BB + SB - CS) / (AB + BB)
At first glance, this appears to be a quality metric for offensive performance -- it's a ratio of bases, which players are trying to accrue, to plate appearances, the opportunities in which players try to accumulate those bases.
Unfortunately, players are not trying to accumulate bases. The point of baseball is to score runs, not gather bases. Total bases does not properly weight the run values of each type of hit relative to each other, and the practice of adding walks + steals (neither of which are worth as much as a single) to TB only compounds the inaccuracy. Because of this, numerous studies have found that Total Average and its ilk do not even match basic Runs Created in terms of correlation to run-scoring.
We recognize that formulas which fall victim to the bases fallacy seem logical at first. Total bases can feel more appealing than linear weights because they involve simple, whole numbers as weights, but linear weights are naturally more accurate because they were developed using empirical evidence. As Colin Wyers once wrote at The Hardball Times:
"The main reason I bring [Total Average] up is the same reason that you studied World War II in high school: Those who cannot remember the past are condemned to repeat it. Somewhere, right now, on a message board or website, there is a young man who is proposing this as the Next Big Thing that will change the way we look at baseball players.
Please, do not be that young man."
 Further Reading
- US Patriot - "Common Fallacies"