What is Sabermetrics? Some Stuff You Should Know

While we’re not the foremost experts on the topic, it’s important that we share with you something that you may or may not be familiar and how we analyze and view baseball.
Neither of us ever knew what exactly Sabermetrics was until we read books like Baseball Between the Numbers. Before that, the lingo and the content on websites like Lookout Landing and USS Mariner seemed like a foreign language. But after reading and learning about the concepts that make up Sabermetrics, our whole view of the game of baseball changed for the better, and we think that yours might as well.

It’s important to at least be acquainted with these terms and concepts as we might reference or use them as reasoning in the future (and the past). So, according to our trusty Wikipedia guide, Sabermetrics is defined as:

the search for objective knowledge about baseball.” Thus, sabermetrics attempts to answer objective questions about baseball, such as “which player on the Red Sox contributed the most to the team’s offense?” or “How many home runs will Ken Griffey, Jr. hit next year?” It cannot deal with the subjective judgments which are also important to the game, such as “Who is your favorite player?

Here are basic concepts you should know, if you want reasoning, click on the corresponding link. I’m going to paraphrase from what I can discern:

Evaluating Pitchers:
W-L record, ERA, and WHIP are flawed statistics because they take into account things that are not within the pitcher’s control, such as defense, and in the case of W-L, offense. A pitcher can’t help it if his defense is terrible or amazing.

Better statistics to use: FIP (available at fangraphs.com) or tRA (available at statcorner.com)
These stats take out things that are outside the pitcher’s control and properly weigh stats like line drive percentage, strikeout percentage, ground ball percentage, home run percentage, and other batted ball or “raw” statistics in order to provide a clear picture of how that pitcher has done regardless of the defense behind him.


Evaluating Offense:
Bad ways to do this: AVG, RBI, Runs Scored, etc. The problem with these stats is, like conventional pitching stats, it takes into account things outside the batter’s control. RBI and Runs Scored are the worst of these because they take other hitters into account. Don’t use them; don’t listen to others who use them.

Before we move onto good methods, we should discuss OBP, SLG, OPS, and VORP. OBP was brought into the baseball forefront by the book Moneyball by Michael Lewis, a fantastic book whose point was badly misunderstood. OBP is the percentage of how often the hitter gets on base. SLG is the average amount of bases the batter accumulates per plate appearance.

OPS is the two stats combined. For a long time this was the best way to evaluate a hitter. However, it is flawed. OPS is flawed because it sees OBP as equal to SLG, which it isn’t (OBP is almost four times as important). Another reason is that all these stats are prone to luck, both good and bad. We could go on for hours about luck’s role in baseball, but suffice to say, it exists, it matters, and it must be accounted for.

VORP is a stat made by Baseball Prospectus a couple of years ago. At the time it was revolutionary. It introduced the important concept of replacement level, which in theory is the performance one would receive from a theoretical AAA guy (more info here). It has been made fun of by stupid people because of its name. It is now outdated. It’s not terrible, but there are better stats to use.

Good stats to use: wOBA, wRAA
These are essentially the same stat in different forms. wOBA properly weighs possible outcomes by a hitter to evaluate hitting (more info here).wOBA is modeled to look a little like OBP. .300 is bad, .340 is average, .400 is really good. wRAA is the same thing but in the form of runs above an average player.

Ten runs = 1 win contributed or lost by the hitter. If a player has a wRAA of 20, he has contributed two wins more at the plate than the average hitter. If he has a wRAA of -14, it means he has contributed 1.4 wins less than the average hitters would have.

Advanced hitting stats are the best stats we have. They are almost perfect and the most stock should be put into them.

Evaluating Defense:
Defense is the one facet of the game that has been truly underrated for a long time. Very few realize its importance outside of Major League front offices, and they just started realizing it a few years ago (and some still don’t get it).

There are three “true outcomes” of a plate appearance which are not affected by the defense. These are the K, the BB, and the HR. Any other outcome is affected by the defense. Defense makes up a very large part of what is commonly recognized as pitching.

Bad ways to evaluate defense: Fielding percentage, errors, range factor, hearing a sportswriter or radio host say a player is fantastic, and worst of all, Golden Gloves
They are totally worthless, built on reputation and lies. Need proof? Rafael Palmeiro won a gold glove in 1999 despite playing 135 games at DH.

There aren’t any surefire ways of evaluating defense. There are some pretty good metrics out there, including UZR (available at fangraphs), PMR (available at baseballmusings.com, but difficult to find) and +/- (available at Fielding Bible for a cost).

The problem is that these stats don’t always agree with each other, and they can vary based on the batted ball statistics they take their info from. The results can often be inconsistent (mainly because players are inconsistent). They also do not take into account positioning and the angle of the ball.

However, that doesn’t mean that they are worthless, and it doesn’t mean that one shouldn’t use them in evaluating talent. How to do it is to look at different years to find a good range for how good a player is.

Instead of saying a player is ten runs above average at defense, we will often say that a player is between 5 and 15 runs above average and split the difference. This provides a more accurate depiction of the player.

Important to note: UZR is the stat we will quote most because it is the easiest to find (right now). UZR is in runs saved or lost compared to average. Remember, ten runs = 1 win contributed or lost. So if a player has a -10 UZR then he is 1 win worse than the average player at his position.

When determining total value compared to players at other positions, the UZR must be modified to include an adjustment for that position. Naturally some positions are harder to play than others and the run value must reflect that.

It is harder to be a league average shortstop than it is to be a league average first baseman. Dave Cameron wrote a good article about positional adjustments, including the adjustments themselves on fangraphs.

Catcher defense is really difficult to evaluate, so take any results there with a grain of salt.

WAR and how to calculate it:

WAR, Wins Above Replacement, is the current end-all stat, combining offensive and defensive value and putting it into an easy to understand form. It can be found without having to calculate it, but if you are so inclined it is quite easy to do, so we will provide a sample calculation.

To get WAR for a position player (pitchers are a little different), simply combine the hitter’s WRAA, defensive value in runs (any advanced metric works but UZR is great because it’s already in runs format), add a positional adjustment, and then add 20 (This makes it so that the player is being compared to replacement level. If you want to compare a player to the average player, then do not add 20). Divide that number by 10 (10 runs = 1 win), and that’s WAR.

WAR is the number of wins that player is worth compared to a replacement level player. A player with a WAR of 3 is worth two more wins to a team over a player with a WAR of 1 with the same amount of playing time. Here is a sample calculation:

Raul Ibanez 2008

WRAA= 16.3

UZR is a little inconsistent with Raul but a generous estimate is that he’s somewhere between -5 and -15 in LF. Let’s split the difference at -10


Positional Adjustment= -7.5

Replacement level= 20


16.3-10-7.5+20= 18.8


18.8/10=1.88 WAR, rounded up to 1.9


Raul was around a 1.9 WAR player last year, 1 runs worse than average.


And that’s WAR. It won’t be the last time you see it on this site. If that was confusing you can look here for more.

Why Bother?

Sabermetrics can be confusing, it requires a basic understanding of math, and sportswriters love to contradict, ignore, and cry against it. So why should we learn more about it? For one, we at the blog can’t analyze baseball and show none of our thought process, so introducing these concepts is a necessary step.

It’s vital to understand that learning more about the game of baseball results in a greater understanding and appreciation of the game. Not that these stats are everything there is to know. Here, we had to rush through everything to introduce key concepts, but there is still so much else to see.

Baseball’s little intricacies are part of what make it different from any other popular sport in America, and are what make the game so appealing. Sabermetrics don’t take away from enjoyment of the game as some have argued, they enhance it.

Quantcast