Why don’t wins count anymore?

In Kansas City, baseball fans are celebrating. In St. Louis, they’re fuming.

It’s usually the other way around. Right now, Royals fans are celebrating Zack Greinke’s highly deserved Cy Young Award. In St. Louis, fans are complaining that Chris Carpenter and Adam Wainwright, expected to finish 1-2 in the voting, got "snubbed" and lost to San Francisco’s Tim Lincecum, who won a total of 15 games.

Greinke, for what it’s worth, won 16.The Cy Young Award usually is "the pitcher with the most wins" award. And that makes a little sense–Cy Young won 511 games in his career, the most all-time. And that itself shows the problem with wins.

Cy Young is the winningest pitcher of all time, but he’s not the best. Walter Johnson won 417 games pitching mostly for last-place Washington Senators teams. Put him on the teams Young pitched for, and he would have won more than 511 games. Win 110 games over the course of your career and you’re considered a pretty good pitcher. Johnson pitched 110 shutouts.

I learned playing Micro League Baseball in the mid 1980s that wins are an overrated statistic. Cy Young was an outstanding pitcher, but Walter Johnson and Lefty Grove could beat him most of the time. Advanced baseball statistics barely existed in the mid 1980s and my Commodore 64 sure didn’t know anything about them, but I quickly started paying attention to WHIP–walks plus hits per innings pitched.

In their best seasons, Johnson and Grove permitted fewer than one baserunner per inning. And they permitted fewer baserunners than Young. Fewer baserunners means fewer chances to score, which means a better chance of winning.

Greinke and Lince*censored*won on the strength of their advanced statistics. Carpenter and Wainwright were very good this year. But they gave up more baserunners per inning than Greinke and Lince*censored*did, and other advanced statistics also indicated that Greinke and Lince*censored*were the better pitchers last year.

In the case of Greinke, the Royals lost six games in which he gave up one run or fewer. Yes, you read that right. Six times, Greinke took the ball, pitched seven or eight innings and gave up one run, or zero runs, and the Royals still lost.

So it’s easy to imagine a scenario where Greinke would have won many more games. Had Greinke pitched on a team that could consistently score more than two runs, had the Royals had more than one reliable relief pitcher to back him up, and had he had more than one above-average fielder playing the field behind him, for example.

Greinke realized he only had one guy behind him who knew how to catch the ball, so he would intentionally pitch in such a way as to make them more likely to hit a fly ball to wherever David DeJesus was playing, usually left field.

Lince*censored*suffered from less bad luck than Greinke did, but still won his 15 games while pitching for a weaker team than the Cardinals.

According to Baseball-Reference.com, pitching for a team with average offense, Lince*censored*and Greinke each would have won 18 games. Under the same normalized conditions, Carpenter would have won 15, and Wainwright would have won 17.

Both pitchers had good years, and admittedly they played for a team that had problems. But Tim Lince*censored*pitched for a team with even bigger problems.

I see the words "which pitcher gave their team the best chance to win every fifth day" thrown around by St. Louis fans a lot. The answer, when you normalize the statistics, is Lincecum.

Or, to look at it another way: Carpenter’s and Wainwright’s win totals showcase just how good Albert Pujols is.

The case for Tim Lince*censored*was less clear than the case for Greinke, and that was why the vote ended up being so close.

But it’s obvious to me that the voters got it right in both cases. And that’s good.

Twenty five years ago, it wasn’t as easy to go much deeper than conventional statistics like wins, losses, and ERA. Today it’s simple, so there’s minimal excuse to pay attention to them.