This post is by no means exhaustive. For some it may be too basic, but my hope is to give everyone who is interested a little more information on sabermetrics. For more information go to fangraphs.com, baseball-reference.com, tangotiger.com, or SBNation's beyondtheboxscore.com. There are many other sites out there as well, but those should provide plenty of information and an excellent start if you are interested in learning more. Click on the bold words for links to more information.
Hall: Do you take a look at any statistics when you watch baseball?
Oates: Yes, I take a look at a few.
Hall: Just so you know, I don’t have a problem if somebody enjoys the game without statistics completely. It’s a beautiful game, and great pleasure can be had watching the best in the world excel. But I’m like you, I like to have a little more knowledge when watching a game. For example, I like to know if a pitcher throws a great curveball so I can watch for it in anticipation. On the other hand, I like to know if a hitter doe not hit with much power so if he hits a home run, I can enjoy being surprised. And that really has nothing to do with statistics, but just a greater knowledge of the game we watch. I think statistics can go a long way in helping people understand the game and can help people enjoy watching and talking about baseball. You said you like taking a look at a few statistics when you watch games. What hitter stats do you pay the most attention to?
Oates: Generally batting average, home runs, and RBI, but I know some about on-base percentage, slugging and OPS.
Hall: Home runs are the simplest way to know if somebody is a power hitter, which is something I like to know. Batting average is good, too, but I think there are some flaws in it.
Oates: What do you mean?
Hall: Well, first and most importantly, walks aren’t included. We can both agree that walks are an important part of baseball and help teams score runs, right?
Oates: True, but they aren’t as good as hits.
Hall: I agree. Walks aren’t as good as hits, but let me pose a question to you, if a batter just finished his at bat, and you have no other information would you rather know if he got a hit, or if he got on base?
Oates: I suppose I’d rather know if he was out or not.
Hall: That’s exactly my point, the piece of information we want to know is if the player was out or not.
Oates: There’s a stat for that, right? On-Base Percentage?
Hall: Correct, knowing a player’s On-Base Percentage, or OBP, can help our understanding of what a player does. It calculates how often a player is on base per plate appearance. That doesn’t mean we should throw out batting average, but OBP provides greater context than just batting average. However, OBP can’t tell the full story.
Oates: Because we don’t know if the player hits for power.
Hall: Right, I mean, we can agree that homers, triples, and doubles are worth more than singles and walks.
Oates: Of course, isn’t that what slugging percentage is for?
Hall: Using slugging percentage is pretty popular nowadays. You’ll see it on most broadcasts. It gives a value of 1 for a single, two for a double, three for a triple and four for a home run. Then we divide that number by at bats to get a slugging percentage. Sometimes you will hear people referring to a triple slash line. They are referring to batting average, on base percentage, and slugging percentage with each number separated by a slash.
Oates: Aren’t there only two slashes?
Hall: Yes, but there are three numbers and that’s just what everybody says.
Oates: So understanding the triple slash numbers gives me a pretty good idea of what a player can do. People use OPS to get an overall number to figure out how good a player is, right?
Hall: A lot of people do add up OBP and SLG to get a number they call OPS, on-base plus slugging percentage, so they can combine two different abilities, the ability to get on base and hit for power.
Oates: I assume there’s something wrong with that one, too?
Hall: OPS is a solid statistic. It is a very good shorthand for how good a player is at the plate, but it has its problems.
Oates: Like what?
Hall: The first one is math-related, so nerd alert on this one. Remember when I said that OBP is calculated by times on base divided by plate appearances and slugging is calculated by adding the values of the different types of hits by at bats?
Oates: If you say so.
Hall: Well, OBP divides by plate appearances so it includes times when a hitter came up and walked while slugging divides by at bats, which does not include the plate appearances that end in a walk, so you have two different numbers that you are dividing by, and that just doesn’t seem like a good idea.
Oates: Is that the only problem? Seems pretty weak.
Hall: It’s not the greatest argument. That’s why OPS is a decent stat to use, but the other problem has to do with slugging percentage.
Oates: What’s that?
Hall: Well according to slugging percentage, a double is worth twice as much as a single, a triple is worth three times as much and a homer is worth four times as much. I mean, would you rather have a player go 3 for 4 with 3 singles or hit one triple?
Oates: I think the 3 for 4 with the singles, but I’m not entirely sure.
Hall: Well, what if I told you that somebody much smarter than me went through a ton of historical data and figured out in all different types of situations what different types of hits were worth, and then took a player’s different types of hits and multiplied them by what they are worth scaled to OBP depending on the run environment that year, and then divides that number by plate appearances except for intentional walks. Is that something you might be interested in?
Oates: Beep, beep, beep. Back it up. That’s way too much information all at once.
Hall: Okay. Well, we can agree that different outcomes like singles, walks, home runs all have different values when it comes to the runs we expect to score in any given plate appearance. The value of the hit or walk is determined from linear weights, which is the result of doing all the historical research of differing plays and their outcomes.
Oates: I think I’m with you so far.
Hall: We know that runs are the most important thing for winning, but in any given year sometimes a run is worth less and sometimes it’s worth more.
Oates: You mean like in the steroid era?
Hall: Exactly, during the steroid era, a lot more runs were scored in total which means that the value of a single run isn’t quite as great during that time. For example, if you look at it in a game-by-game perspective, if the score of a game is 1-0, that one run has a whole lot of value, but if the score of a game is 10-9, any run taken at random out of 19 isn’t as important even though the margin of victory is the same.
Hall: So if we know how much a run is worth, and we know how much any type of hit is worth in that context, we can then apply those values to a particular player’s hits, walks, etc. Then we can divide by plate appearances and come up with a single average.
Oates: You said something about scaling the number to OBP, what does that mean?
Hall: When the people who are smarter than I am are figuring out the value of different hits and such and come up with the average, they adjust all the players’ averages so the final number comes out so it looks like OBP?
Oates: Why do that?
Hall: OBP is a number generally people are familiar with so that they don’t have to learn a whole new scale of what a good number or bad number is?
Oates: So what is a good number?
Hall: It depends on the year, but generally speaking average is going to be around .320 give or take five to ten points. .350 is going to be really good and .400 is going to be awesome.
Oates: Does this stat have a name?
Hall: I guess I buried the lede a little bit on that one. It’s called wOBA, for weighted On-Base Average. Tom Tango, who co-wrote an excellent book called The Book all about advanced baseball statistics, came up with the stat.
Oates: Where can I find it? I don’t exactly hear about it too often?
Hall: wOBA is available at fangraphs.com
Oates: That’s interesting but I’m not sure I want to use or rely on a stat that I couldn’t really figure out myself.
Hall: That’s a fair point, but let me ask you this, do you follow football?
Hall: Have you ever heard of something called QB rating?
Hall: What is it?
Oates: It’s this rating they have for quarterbacks where they account for completions and yards and touchdowns.
Hall: What’s a good number to have?
Oates: Anything over 90 is pretty good and over 100 is really good.
Hall: How do you know?
Oates: I look at the league leaders and all the quarterbacks that I think are the best, Manning, Brady, Rodgers, all have those type of numbers.
Hall: If you saw a quarterback up on the list that surprised you, would you discount the statistic as worthless?
Oates: Probably not if all the good quarterbacks were up there.
Hall: Well, what would you think about that quarterback?
Oates: I’d probably think maybe he was a little better than I thought, unless it was Mark Sanchez.
Hall: If I told you the top 5 in wOBA last year were Miguel Cabrera, Ryan Braun, Mike Trout, Buster Posey, and Andrew McCutchen, does that seem like a pretty good list of the top hitters in the league?
Oates: Is Mark Sanchez anywhere near the top?
Oates: Then, yeah that seems like a pretty good list.
Hall: Okay, back to the QB rating, do you know how that’s calculated?
Oates: I know they do something completion percentage and yards and touchdowns, but no I’m not entirely sure.
Hall: But you agree that completion percentage, yards, and touchdowns are pretty important for quarterbacks?
Hall: And for hitters, walks, hits, and homers per plate appearance are all pretty important?
Hall: And if there was a statistic that combined all those and the top hitters every year appeared on the leaderboard for that statistic, it’s probably pretty reliable, right?
Hall: If that’s the case, it seems you are more willing to accept some advanced statistics than you originally were.
Oates: But doesn’t it take a lot of work to know?
Hall: It does take some work, but once you know about it, it takes little more effort than to find out batting average or any other statistic, and you’ll have a lot more knowledge about the performance and value of hitters.
Oates: We never talked about RBI.
Hall: Oh yeah. RBI do an okay job of telling you what happened, but the problem comes when we as fans get into arguments.
Oates: How so?
Hall: Well, if I want to say my player is better than your player because of RBI totals, is it really fair to talk about those totals if one player had say 50 more at-bats with runners in scoring position? What if the only reason they got more RBI was because they had more opportunities?
Oates: Well then can’t we look at a batting average with runners in scoring position?
Hall: If you want to look at it over the course of a 20 year career, it might be meaningful, but in one season or even a collection of seasons, you don’t get to look at that many at-bats. Wouldn’t you rather know how good a player is overall as a hitter in all situations? You wouldn’t want to back a hitter in an argument who didn’t try as hard when runners weren’t on base? Shouldn’t they be trying hard every at-bat?
Oates: I guess, but it seems like some players just do better in clutch situations.
Hall: Generally speaking, the clutch players you are talking about are simply the better players overall. They are more likely to do something positive in all situations which means they will probably do well in clutch situations. If it seems like somebody is super clutch or a huge choke-artist, it is more likely that we just don’t have enough at bats in those situations. They also could be facing better pitching in those situations. It is likely that the statistics will even out over time.
Oates: One thing I really don’t like about sabermetric statistics on television is the statistics they show like average in day games or against a certain pitcher or when batting at some spot in the order when at home on Thursdays. I can't go for that. No can do.
Hall: That’s not really sabermetrics. Most of the things you mentioned above are statistics, but most people in baseball who advocate for sabermetrics don’t advocate for those statistics. They can be interesting, sort of like finding out two players have the same birthday or went to the same high school, but because we usually see so few at bats in particular situations, those types of statistics don’t help set expectations for what is likely to happen. Just because a statistic is obscure, that doesn’t mean that it is really involved with sabermetrics and the use of advanced statistics in baseball.
Oates: Oh…okay, I get it. Let me ask you, has any conversation in real life ever gone this easy for you?
Hall: Oh God no, but as you can see, as far as sabermetrics go, you're not really out of touch.
Oates: I'm out of time. See you soon.
If anyone has any other questions/criticisms, please post them below and I will do my best to answer. There are a lot of other great stats out there that I chose not to go into like OPS+, wRAA, and wRC+ because I didn't want to get bogged down when the main purpose was the "why" to use these statistics. If this has been remotely helpful to anyone, I will follow up with a post on pitching and another post on WAR which will include fielding.