Baseball cards have had stats on the back of them since the 1920s, and Topps made it a regular fixture with their first full set of baseball cards in 1952. For about as long, many baseball fans have paid far too much attention to the wrong kinds of stats. The problem is that it leads to faulty, misguided, and sometimes just wrong assumptions about a player or a team.
While there are several stats for both pitchers and hitters that deserve mention in this category, I’d like to focus on what I consider to be one of the worst offenders, the application of individual W/L records to pitchers. How this stat has been attributed individually to pitchers for so long is hard to understand. Maybe I’m being a touch dramatic about it, but it seems strange that a statistic that is so clearly the product of factors beyond the pitcher’s control is still not only attached to that pitcher individually, but also still generally used to gauge the pitcher’s effectiveness. Many of us still talk about pitchers in terms of their W/L record, and consider 15-20 wins to be the standard for the game’s best. That needs to change.
While the pitcher plays a very large role in the game’s outcome, there are a multitude of smaller factors that add up to a much larger role in the game’s outcome. For instance, even in the NL, the pitcher has extraordinarily little, or none at all, to do with the offensive production of his team. Additionally, when a pitcher leaves a game, he relinquishes it to his bullpen. So, even in a time when he might have pitched extremely well, a pitcher can see a W transformed into a ND without being able to do a thing about it. Those two are the biggest elements, without even mentioning the quality of a team’s defense. Errors aside, a team’s defensive performance can make the difference in the number of runs allowed, and again, that often has very little to do with a pitcher. For instance, a player’s range in the field and/or arm strength can make the difference between a single and a double, a runner advancing, etc.
I’d like to spend time on the more effective stats a bit later, but first let’s take a look at what I consider to be a fairly decent example of this. It highlights one of my favorite former Cubs pitchers (and hopefully future Cub pitcher in the 2015/2016 offseason), Jeff Samardzija.
If you take a look at Samardzija’s 2014 season, there are some interesting things to note. Generally, his individual performance during his time with the Cubs last season was very similar to his individual performance during his time with the Athletics.
While with the Cubs, he made 17 starts and pitched 108 innings. In that span, he gave up 99 hits and 34 earned runs. He struck out 103 batters and posted an ERA of 2.83. After being traded to Oakland on July 5, he made 16 starts, and pitched 111.2 innings. In that span, he gave up 92 hits, 39 earned runs, struck out 99 batters and had an ERA of 3.14. If you were to take a look at the rest of his numbers between the two teams during last year’s season, you would find that his performance stayed pretty consistent. Not only that, but if you examine his overall numbers during his years of playing with the Cubs (though some of this includes time spent shuttling between the bullpen and the rotation, as well as Iowa and Chicago), you will find that he has performed pretty consistently since his debut in 2008. 2009 and 2010 were rough years, for sure, but again, this involves fluctuations to his duties on the mound that were outside of his control.
With all of this considered, I think it is worthy to note that his W/L records are quite different between Chicago and Oakland, although the sample size is admittedly small. I will be curiously following his performance with the White Sox this season, to see if this bears out further. As a Cub in 2014, he was assigned a 2-7 win/loss record, and while in Oakland, a 5-6 record. Generally speaking, a 2-7 record communicates a very different message than a 5-6, not to mention that he moved to the American League, where it is usually considered more challenging for pitchers (i.e., while in Oakland, he gave up more runs, struck out fewer batters in more innings, and had a higher ERA.).
One of the more interesting differences during Samardzija’s 2014 season had nothing to do with his pitching. In his 17 starts with the Cubs, his offense scored 41 runs behind him, and they were shut out by the opposing pitcher on 5 of his starts. In his 16 starts in Oakland, his offense scored 76 runs and they were shut out just once when Samardzija was the pitcher. That alone is probably the largest reason for the difference in his win/loss percentage, and he has virtually nothing to do with it. I say “virtually” here because, yes, while in Chicago he was batting, but even there, he’s not expected to be a contribution to the offense.
In all, a pitcher putting forth roughly the same performance should yield roughly the same W/L, if W/L is a worthy stat for an individual pitcher. The problem is that it’s not. So here’s a brief look at three stats that I think give a much better sense for the job that a pitcher is doing:
Quality Starts (QS): This actually isn’t listed on Baseball Reference, but I think it should be, and possibly in place of the W/L totals. A quality start could be measured fairly easily, and still manage to encompass the spirit of the idea that W/L tries to capture. As a whole, it’s a measure of whether or not a starting pitcher has put his team in the best position possible to win the game. So, in order for a starting pitcher to have what I’d consider a “quality start,” they should pitch at least 6 innings. This means that they are allowing the role playing bullpen pitchers to do their jobs as they are intended. The 7th inning can be handled by one or two specialists, the 8th goes to the setup reliever, and then the 9th to the closer. Along with that the starting pitcher should allow no more than 3 earned runs during those 6 innings, and that includes possible runners left on base for a reliever to take care of after the starter leaves the game. Assuming that the relievers will do what they are expected to do, I think it’s reasonable to expect that even an average offense can manage 3 or 4 runs in a game. For example, Samardzija would have had 6 QS while in Chicago in 2014, and 6 more while in Oakland.
Walks and Hits per Inning Pitched (WHIP): This is actually one of my favorite stats to watch for when looking at a pitcher. It plays off of the idea expressed in the QS stat, but it’s a closer look at how a pitcher does at keeping men off of base. If the offense has the job of getting on base by whatever means possible, then it should be the job of the best pitchers to keep that from happening. So a pitcher that can average fewer than 2 men on base in an inning (as Samardzija has done with relative ease nearly throughout his career, he has a 1.273 average in this stat), he is doing his job in this respect.
Fielding Independent Pitching (FIP): This measures the ability of a pitcher to minimize the worst offenders like HR, HBP, and BB and to cause strikeouts, all measured as independently as possible of the defense that surrounds him. It’s perhaps the best measure out there that shows just how well a pitcher does at doing just his job. In 2014, Samardzija had his best season in this regard, excluding his 2008 performance, because it doesn’t encompass a full season in the majors.
Ultimately, in a time when advanced statistics are increasingly altering the way that we watch games and determine a player’s value, it’s surprising to see that some stats remain a part of the equation at all. Samardzija is just one example, but it’s likely that GMs across the league already know better than to gauge a pitcher based on W/L. If they did, I doubt that a pitcher who goes 2-7 in the first half of 2014 is a part of a trade that involves Addison Russell going to the other team.