The problem with Wins as an evaluator of starting pitchers is not that it is bad statistic. It is simply a matter of sample size. In a single game, a win or no win is not a good indicator. Why? Small sample size (n=1). However, ERA, for example, is a per inning stat. So in a single game, a pitcher’s ERA will have 5-9 data points (n>>1). Over the course of a full season, stats like ERA+, FIP and tRA have a sample size of 150-220 for each pitcher.
And later on (emphasis all mine):
In fact, in the absence of other stats, Wins is a very good, if not great, indicator of a pitcher’s value. So next time you hear somebody say Wins is a crappy way to evaluate a pitcher, throw a drink in their face and then make them read this post.
To me, this is a lot like saying that in the absence of anesthetic, a piece of wood to bite down on is a good pain management tool. Yeah, I guess that's sort of true, but it's also a completely useless observation in the modern world where anesthetic is always an option. Sort of like how, given the plethora of available information, wins are... ...completely useless. You would and should never prefer them when you have access to other, better statistics. Opting for wins to evaluate a pitcher is like opting for the piece of wood when your leg is being amputated. In the modern world, it's never defensible.
Also, the problem with wins is emphatically not small sample size. Even if pitchers played 1,000,000,000 games every season, wins would still be worth shit because pitchers play with the same offense and bullpen day in and day out. Those pitchers with better offenses and better bullpens will get more wins than those without and there's nothing that a large sample size can do about it. Indeed, larger sample sizes will make clear exactly how large this bias is. Wins are bad because they can do nothing to correct this bias.
Do not use wins. That is all.
**EDIT** J.C. Bradbury gets in on the action here. Worth reading.