Run Differential Says Things Aren’t So Bad

Screenshot 2016-05-25 at 10.20.45 AM

Following last night’s game, the Dodgers are coming off of a three-game winning streak. Despite this success (let’s celebrate the small stuff,) you will not be shocked when I say that this team is underperforming. The offense is impotent and the bullpen is a mishmash of wet noodles for arms and Kenley Jansen. The rotation –  not complete garbage as others may make it seem – is having struggles of its own. Without Clayton Kershaw, the rotation is the not-so-proud-owners of a 4.34 ERA. Throw Kershaw back into the rotation and the pitching staff’s ERA drops to 3.55. When all of this is combined, we end up watching a Dodgers team playing to the tune of a 24-23 record, 4.5 games out of first place.

A section of the Dodgers Twitterverse will have you believe that if the Dodgers are not in first place, all is doom and gloom. Of course, this same group of Twitter will also have you believe the worst is just around the corner even when the Dodgers are in first place. The point of this post is not to convince you the team is free of its flaws; that is far from true. Instead, I am going to talk about run differential and how the NL West should  look according to runs scored and runs allowed. Hopefully, that will alleviate some stress felt when looking at the standings.

A Primer on Run Differential

Simply put, run differential is the difference between runs scored and runs allowed. Theoretically, a run neutral team – one that scores exactly as many runs as it allows – is a team with a .500 record. If we take a 162-game season, a team that scores the same number of runs it allows should finish the season with 81 wins and 81 losses. If a team scores more runs than it allows, it should also have more wins than losses. In fact, every ten more runs scored in relation to runs allowed translates to one extra win. This is not an exact science. Rather, run differential is an indication of whether a team is underperforming or overperforming. Naturally we want to see the Dodgers outperform its run differential. In fact, we have seen this happen the last three years.

Season Runs Scored Runs Allowed Run Diff Actual W/L Pythagorean W/L
2013 649 582 +67 92-70 88-74
2014 718 617 +101 94-68 91-71
2015 667 595 +72 92-70 88-74

Looking at the table, you can see that the Dodgers outperformed their expected win-loss record the last three seasons. Were the Dodgers of years past bad teams? Definitely not. Think of it this way, if you expected to get a B in statistics but instead got an A, you were never a terrible student to begin with. Instead, you simply did better probably for a variety of reasons.

What Run Differential Says About 2016

Now let us look at the standings so far. The Dodgers are 4.5 games out of first place. Is that the expected outcome? Looking only at run differential, the expected outcome has the Dodgers in first place.

Team (Current Standings) Run Differential Acutal W/L Pythagorean W/L (Expected Standings)
Giants +13 29-19 25-23 (2nd)
Dodgers +22 24-23 26-21 (1st)
Rockies -11 21-26 23-24 (3rd)
Diamondbacks -15 21-26 22-25 (4th)
Padres -38 19-28 20-27 (5th)

Granted, what is expected does not change what has already happened but it should provide some hope for a season that has been frustrating, to say the least. Yes, the Giants are a team that should be feared, after all, they won the World Series after gaining the second Wild Card spot with only 88 wins. If the Giants continue to outperform their Pythagorean expectations, it will likely be because they dabble in devil magic and Hunter Pence is Lucifer incarnate. And I guess because they do have a solid team. Should the winning streak end and the Dodgers hit another skid, it will be difficult to regain ground lost against San Francisco. But the Dodgers will get back on course. Though it has been tough to watch them, they are not playing as bad as the rest of the NL West. Those teams are playing below expectations with a negative run differential. That’s like losing a footrace because you’re slow and you have no legs. At least for the Dodgers, they have legs, they’ve just tripped over several hurdles.

Even if you look at this and say, “Dude, this is shit,” you cannot change the fact that there is a ton of baseball to play – over 100 games still! Being 4.5 games back in May is different than being 4.5 games back at the end of September. The Dodgers have the ability and the time to make up the difference in the standings. Whether they do so remains to be answered.


ERA and why 0.00 ERA over three wins does not matter.

The next few posts will be sabermetric focused and provide some insight into why some stats are great for the layman, but are highly misleading. First up, ERA and what are some better ways to analyze a pitcher’s talent.

From time to time, especially when talking about my favorite player, Clayton Kershaw, I’ll gush about or bemoan a pitcher’s ERA. It just happens. Deep down I know ERA is a bad stat, but when it’s the one pitching stat that will forever be seared into baseball lore, it’s hard not to discuss Kershaw’s historically low ERA. And while Earned Run Average provides a glimpse into a pitcher’s effectiveness or ineffectiveness, it is through a heavily blurred lens. To explain why ERA is flawed, we have to examine what an earned run is.

An earned run is any run for which the pitcher is held accountable. For example, if Kershaw allowed Batter A to reach base on a hit and that Batter A came around to score, it is an earned run against Kershaw. Makes sense, right? It seems intuitive. But in a different scenario, say Kershaw allowed Batter A to reach base and then he was pulled from the game and was replaced by Brandon League. If League allowed Batter B to get a hit and Batter A scored, the earned run still goes against Kershaw because he allowed Batter A to get on base. You might be thinking, “But it shouldn’t count against League because it’s not his fault Batter A is on base.” And you’re right. But League was the reason why Batter A scored. The other problem with ERA is that some pitchers with a high ERA are backed by a lousy defense. Balls that get by slow or inept infielders, or drop in because the outfields are lazy schlops go against the pitcher who is normally lights-out. On a more analytic level, ERA does not account for league differences – American League pitchers are more likely to have higher ERAs since they face a designated hitter – nor does it consider differences in parks, the umpire, the scorer, and just random luck in general. So ERA doesn’t limit the amount of variables that can affect a pitcher in a single game. 

If you’re one who would just like to look at a stat and not have to put much thought into it, K/BB is a great ratio to measure a pitchers efficiency. It simply looks at a pitcher’s strikeout rate over walk rate. Kershaw has a 8.6/1.9 K/BB. So for every one walk, he has about 4.6 strikeouts. Brandon League has a 4.66/2.56 K/BB, or 1.82 strikeouts for every walk. Another quick stat to look at is WHIP, which is Walks + Hits per Innings Pitched. This stat will tell you how many walks and hits a pitcher allows per inning. A WHIP rate of 1.32 is about average, with 1.00 being excellent and 1.60 being awful. This season, Kershaw is sitting on a nice 0.86 WHIP rate. League is at 1.42

If you simply cannot let go of ERA, then I suggest you take a look at FIP. FIP stands for Fielding Independent Pitching. FIP looks at strikeouts, walks, home runs, hit-by-pitch pitches.  Yes, some variables that affect FIP are the umpire, the park, and some luck, but it does a great job at examining over which events a pitcher has control. FIP is scaled to ERA, so a person interested in this stat can easily understand the stat at first glance. A variant of FIP is xFIP and it is a slightly more accurate calculation of FIP – it doesn’t assume that home runs are the fault of the pitcher, so instead it looks at fly-ball to ground-ball ratio instead of home runs allowed. To have an idea of what is a good FIP or xFIP and what is awful: an average FIP/xFIP is 4.00, with excellent being at 2.90 and awful being 5.00. Clayton Kershaw has a 2.47/2.97 FIP/xFIP this season. League has a 4.78/4.17 FIP/xFIP.

Now, if you’re suspect of ERA, but K/BB, WHIP, FIP, and xFIP haven’t convinced you, perhaps tERA is a better stat. Again, like FIP and xFIP, it is scaled to ERA so looking at it is intuitive for any baseball fan. tERA does not differentiate between earned and unearned runs. How a ball is hit can determine how a ball is fielded or not fielded. Since how a ball is hit is determined by the pitcher, we can extend the logic and conclude that how a ball is fielded is also determined by the pitcher. A groundball is expected to be an out more so than a line drive is. A fly ball is more likely to lead to runs scored than a ground ball is. With that said, a fly-ball pitcher is more likely to have a higher expected run to expected out ratio. With that ratio, tERA can be determined and we can gauge how hard a pitcher is being hit and what exactly a pitcher is doing. For reference, an average tERA is 4.20, with 3.20 being excellent and 5.50 being awful. This season, Clayton Kershaw is at a 2.78 tERA. League, poor poor League is at a 4.86 tERA.

So there you have it. The next time you hear that Brandon League has found his groove with three wins and a 0.00 ERA over those three wins, take a look at his peripheral stats and let them guide you to a more accurate discussion.

TL;DR: Kershaw good, League bad.