 Most of the time when you watch a ballgame on TV, the first two stats the announcers will use about a pitcher are his ERA and his win-loss record. These two numbers are supposed to give us some idea of the quality of the pitcher, but do they really? ERA isn't perfect, but the more or less reliable of the two stats to me, it seems is win-loss record. Let's compare two pitchers. The first guy pitched in 31 games in 2012, he had a 3.05 ERA. Pretty good stuff. The second pitcher had similar stats, pitched in 30 games, had a 3.16 ERA. Also pretty good. So we would probably assume that these two guys have a similar win-loss record. Actually not so much. Pitcher A went 17 and 6, which is pretty darn good. Pitcher B, on the other hand, only won 6 games. And no, I'm not making this up. These two players were actually in the same starting rotation in 2012. This is Cole Hamels and Cliff Lee. What's the deal here? Let's break it down further. Hamels pitched 215 innings, Lee 211. Hamels gave up 73 ERAs, Lee 74, again almost identical. Hamels walked 52 batters. Lee walked just 28, he actually led the National League in walks per inning. Strikeouts were 216 to 207. And if you really want to go that far, both had a war of 4.2. Very similar performances all around, but what happened? There were obviously plenty of factors at work here, but one of the most obvious is that the Phillies scored an average of 1.3 more runs when Hamels was on the mound than when Lee was. I think most people would agree to say that it would be just absurd to suggest that they were playing harder when one guy was on the mound than the other. So really that just boils down to luck, which probably isn't what we want to be basing our stats on. If things can really get that out of whack though, then why do we even care about pitcher wins? Let's go back a few years to the guy who has won more games than a single season than any other. That would be Charles Radborne, or Old Haas as he is commonly called. He won 59 games in 1884. Keep in mind that today if a pitcher wins 20 wins in a season, they're considered awesome. Old Haas tripled that, and the way he did that was by receiving a decision in almost every game he pitched in that year. In 1884, Radborne pitched in 75 games, and he was the winning or the losing pitcher in 71 of them. Again, compare that to Lee, who was one of the best pitchers in the league last year, and received 15 decisions in 30 games. In Radborne's season, actually at less games. Perhaps more revealing is the innings pitched. Old Haas threw 678 innings, which did lead the league in 1884, although not by too wide of a margin. In comparison, Justin Berlander led the league in innings pitched in 2012, with 238. So obviously things have changed here, and the game has evolved a great deal. Let's go back to those decisions. Old Haas was 59 and 12 with only 4 no decisions, so what this 59 and 12 record is essentially telling us is his team's record on the days that he pitched, and it makes absolute sense to keep track of this stat when pitchers are going to pitch most, if not the whole game, 95% of the time. But what does Lee's win total really tell us? Radborne had a no decision in 5% of his games. Lee is getting a no decision in 50% of his games. So wins are really just telling us exactly what they're defined as. They're games in which Lee threw at least 5 innings that his team scored enough runs to get the lead before he left, and that his bullpen did not blow the lead after he left the game. Now that might tell us something about Lee's quality, but I'll leave it up to you to decide how much.