Triple-counting singles obviously doesn't make the stat better, it makes it worse. You'd do much better adding extra credit for the walks instead since they're only counted once and are more than half as valuable as singles. But you don't know what weights he's using. If it's anything close to 1:1:1, obviously making singles THREE times as valuable as walks is borderline nonsensical. But on the other hand, traditionalists are very into batting average. Most likely this is a product of the fact that it's easy to calculate and used to show up in newspapers along with HR and RBI, sometimes SB. So it was ingrained that it's important. The truth is some very good offensive prediction models totally ignore average; you can get a very good picture of how valuable a player is offensively from only OBP, SLG, SB, and CS. But you can't fault a guy for trying to develop a model that his grandfather might be able to appreciate. I think putting some extra weight on average provides a useful stat for things like MVP prediction and perceived value, even if it actually reduces the correlation of the stat to true offensive value (that's not to say that guys with high averages are inherently not valuable; obviously a guy like Cabrera is helped by this, and he is an offensive stud; what I do mean is that if you compare, say, 2008 Adam Dunn who batted .236 with an .898 OPS to 2008 Ichiro who batted .310 but only OPSed .747, triple counting the singles really overrates Ichiro relative to Dunn. If you use the Bill James ratio of something like 1.7*OBP + SLG it actually moves Dunn further ahead of Ichiro, so a stat that instead brings Ichiro back closer is clearly missing the boat. On the other hand, at least it gives him some credit for his ridiculous 43/47 stolen bases/attempts.)