7/19/2012

Win Dependent: Why Starting Pitcher's Win Totals Are Overrated



Yes, you play to win the game. But it's not always under the starter's control
Herm Edwards may hunt me down for saying this, but I think wins are overrated.

Actually, wins themselves are not overrated, but when they count into a pitcher's or a quarterback's personal record, I believe it becomes an overrated stat, especially in the short term.

For today, let's keep the focus on baseball.

For the casual baseball fan, a pitcher's win-loss record is how they judge a pitcher to be good or bad. If you're a smarter fan, you realize there are many factors that go into whether a pitcher gets a win or loss. Here are some of them:


  1. Run Support - In order to get wins, before a pitcher can get one, he needs runs--usually about 4-5+ does the trick. However, if a pitcher is on a team whose offense is on one of the best offenses in the league (i.e. the Yankees), he will have a better chance of wins over the course of a season, even if the pitcher isn't considered an elite one.  Chien-Ming Wang comes to mind. During a two-year stretch recently with the Yankees, Wang went 38-13 despite having a slightly above average ERA during that time (about 3.66). This inflated record was largely due to the strength of the Yankees lineup, which is almost always one of the league's best.

    Meanwhile, a guy like Felix Hernandez struggles to win many games on a yearly basis despite being one of the best pitchers in the league. Out of his seven full seasons, he has only won more than 14 games once, despite regularly being one of the top 10 AL pitchers in ERA. At least Cy Young voters have given him credit, voting him as the award winner in 2010 despite a 13-12 record (he struck out 232 in a league high 249+ innings while also winning the ERA title at 2.27). If he were able to net close to similar numbers on a team like the Yankees, he would likely have a chance to win about 20 games a year.
  2. Bullpen - The strength of a team's bullpen can also go a long way in determining games that count for,  against, and don't count altogether against a pitcher's record. A guy can throw an 8-inning gem in which he is winning 2-1 when he leaves, but he doesn't get a win because his reliever allows a run in the 9th. A team with a shitty bullpen can distort how effective a starter has been if we're looking strictly at wins and losses.
  3. Strength of Opponent - With this, I'm not just referencing the lineup that the guy has to face, but also the pitching matchup. It is out of a pitcher's control as to how good or bad the opposing starter will be pitching-wise on a given day. One game, you may get a win despite allowing 6 runs because your opponent's starter allowed 8. Another, you may lose while only allowing 1 because you face a staff ace that shuts your team's offense down. While many would say that this stuff balances its stuff out, I don't believe that to be the case for everyone. Not all pitchers experience the same balance, as some may always have the benefit of a strong offense (where scoring 4-5+ is the norm) whereas a weaker offense may have problem scoring more than 3 runs consistently, putting considerable pressure on the starter to earn "his" win.
  4. Strength of the Starter - It may seem a given that a pitcher's performance plays a role in whether he gets a win or not, but it does seem weird to say that his performance is probably towards the middle of importance in terms of whether he gets a victory. The examples provided above show that the pitcher's actual performance doesn't always dictate whether he will get the win.

    Obviously, the starter has to go 5+ innings to qualify for a win, but beyond that, it's up in the air as to whether you can look at a guy's pitching numbers and say whether he won the game or not.

    Below are two separate 4 game stretches by pitchers this year. One of these pitchers went 1-0 with 3 no-decisions, another went 0-1 with 3 no-decisions:

    Innings Hits Runs Earned BBs Ks
    6 5 5 5 3 3
    8 3 0 0 1 10
    7 11 5 5 1 7
    5 8 5 5 2 4
    Totals 26 27 15 15 7 24
    WHIP: 1.31 ERA: 5.19
    Innings Hits Runs Earned BBs Ks
    6 2 1 1 2 4
    7 5 4 4 0 7
    10 7 0 0 0 7
    6 5 2 2 1 6
    Totals 29 19 7 7 3 24
    WHIP: 0.76 ERA: 2.17

    Example 1 is R.A. Dickey's last four starts before today - he won the game he deserved to win, but didn't lose any of the 3 games he deserved to lose. This is because his offense bailed him out in those bad starts.

    Example 2 is Cliff Lee's first four starts of the year, a stretch of starts that includes a rare 10-inning performance in which he picked up a no-decision despite allowing no runs. You could argue that in this stretch, he should have been 2-1 at least.

    Granted, these are both small samples, but do indicate how much outside influence is required in getting wins and avoiding losses.
When you take everything I've mentioned above, you can see why looking at wins is probably not the best way to judge a pitcher. Yet, it seems that the media judge how good or bad a pitcher is based on the pitcher's record. After all, when you hear a pitcher's stats announced by a baseball analyst in a game cast, his win-loss record is always the first thing they mention.

I'd say that long-term, pitching wins are more of an indication of how good a pitcher is than on a short-term basis. There isn't a 300 win pitcher in baseball history who I'd say is an average or even a slightly above-average pitcher. Over the long haul, you have to pitch well to rack up a ton of wins. These guys will almost always have seasonal ERAs better than the league average with lower amounts of base-runners allowed.

If you happen to be on a team that provides you with a ton of runs, perhaps you can be like Wang and get close to 20 wins once or twice. But don't be that fool that falls for one-year wonder and suddenly things a guy is a good pitcher.

What makes a good pitcher is not wins. Look at the numbers and ratios that are under his control (for the most part), like strikeouts and walks per 9 innings, to judge how effective and good/great a guy is.

While we are on the subject of wins...I wish there was a way to edit how we credit winning pitchers. Say for example, a guy throws 8 shutout innings, but his bullpen allows the game to become tied in the ninth. However, in the bottom part of the inning, the team gets a walk-off win. In this scenario, whatever reliever was pitching last in the ninth gets the win, even if it's the guy that blew the initial lead by allowing one or more runs. All the while, the guy who allowed NO runs gets no win.

I say in cases like that, either give the starting pitcher the win, or don't credit any pitcher with the win. It makes no sense to reward a pitcher with a win if he's the reason they didn't initially win the game in the top of the ninth.

Anywho, I'm done. Looks like Herm Edwards hasn't found me yet