Before reading this post, I would like everyone to take a second (or ten minutes) and watch this video:
That is my inspiration for writing this.
As many of you have probably recognized, I am a supporter and user of sabermetrics when evaluating players. As soon as I saw that they were out there, I took the time to study and understand them. And after doing so I immediately grasped them and wanted to know more. I saw their value, and how much more accurate they are than more common stats.
However, I know there are a lot of people out there who do not feel the same way, and that is fine. I can understand people wanting to stick to what they know for whatever reason. I do not necessarily agree, but that doesn’t matter. Everyone has their opinion. But as seen in the video above, it is enormously clear that some people, like Hawk Harrelson, just do not understand the stats they are refuting. It is very hard to accurately dispute something without understanding it. What’s that old saying? The best way to win an argument is to understand the opposition’s position or something? Well Hawk clearly does not find that idea particularly important, as seen when he says “You’ve got your OBPS….” He does not know what he is arguing about. He, and many others, argue against the idea of sabermetrics rather than the stats themselves.
And therein lies the problem. It seems like the overwhelming population of people who are dead-set against advanced statistics do not care to understand them in any way. I would be willing to bet that the majority of those people do not know any advanced stats other than WAR, and maybe BABIP. How can you say you do not like sabermetrics when you only know two of them, neither being the “best” advanced metric out there? You just can’t. People like Hawk and Harold Reynolds default into ad hominem nonsense-arguments that are almost too basic to argue.
Now, this is not the case for some. There are those who have taken some time to study the stats and just do not find them useful for whatever reason. But these people are an overwhelming minority, as most people that take the time to understand sabermetrics so appreciate them, and see their value. Because numbers don’t lie. When someone sees that wOBA has a 97% correlation to actual run creation compared to somewhere around 73% for batting average, it is hard to ignore.
But this post is not designed as a big sabermetric debate. You know my preferences, I just wanted to add a little background to them. The real purpose of this post is a quick rundown of some of my favorite sabermetrics, and a description of each. I will also try to connect them to the Mariners in some way to keep them relevant to Sodo Mojo.
wOBA- This is probably my favorite advanced statistic of them all. As I already mentioned, it has an extremely high correlation to actual run creation, which is the goal of any stat. That correlation can vary from year to year, but I often see it as 97%. The formula forweighted on-base average is as follows:
wOBA = (0.691×uBB + 0.722×HBP + 0.884×1B + 1.257×2B + 1.593×3B + 2.058×HR) / (AB + BB – IBB + SF + HBP)
As you can see, the stat uses linear weights to properly value each hit. The problem with OPS is that it assumes a double is worth two, single worth one, etc., and that isn’t true. OPS is not a bad stat by any means, but it isn’t quite as accurate as wOBA. As a general rule, stats that use linear weights or something else derived from actual events are likely going to be more accurate because they aren’t working on assumptions. They are using real data to be as accurate as possible.
This is how Fangraphs grades out wOBA:
Kyle Seager is the current wOBA leader on the Mariners, with a .383 mark. According to the table above, that is between great and excellent. So go Kyle Seager. As for the team though, it isn’t quite as exciting. The M’s team wOBA is at .299, good for 25th in the league. Hey, at least it isn’t “awful.”
wRC+- wRC+ is another go to stat of mine, often used in tandem with wOBA. Some argue that wRC+ is the better stat, and it some ways it is. It is essentially wOBA, but adjusted for park effects, and put on a different scale. 100 is the league average, so anything about that is what you would ideally want. And it is very easy to work with too, as a wRC+ of 110 means that player creating 10% more runs than the average player. And inversely, a 90 wRC+ means the hitter created 10% less runs than average. Between that and the park adjustments, I don’t know why I prefer wOBA. I just do. But nonetheless, they are the two best offensive stats we have at this time. And I hate to say something like this, but you could use these two stats to compare two players offensively and be fine. You really wouldn’t need anything else. But, more information is always better.
Here is the formula : (((wOBA – lgwOBA) / wOBAScale) + (lgR/PA)) * PA
The league wOBA (lgwOBA), wOBA scale and league runs per plate appearance will all vary from year to year, so there is another clue that we have a good formula on our hands. It is able to evolve and adjust from year to year as the value of a run changes.
Seager is also the team leader in wRC+ at 150, meaning he creates 50% more runs than the average player. Man, I love Seager. In terms of the team however, more disappointment. They are at 92 wRC+, which is 21st in the league. On the bright side, they aren’t last like they were last year… in every category.
ISO- ISO stands for isolated power, and what it does is…Well the name kind of says it all. It attempts to measure a player’s extra base power. You can kind of think of it as slugging percentage without the singles, as it is based on a players extra bases rather than total bases.
The forumla, and the stat in general, is extremely simple: ISO= Extra bases/ At-Bats
See how simple but awesome the stat is? You can also calculate it by subtracting batting average from slugging percentage, but you won’t always get the same answer, and it isn’t quite as accurate. If you want to measure a player’s straight up raw power, ISO should be your go-to stat. However, you have to be very careful with sample size as it does not stabilize quickly at all. So you may see a .400 ISO a month into the year, but know it won’t stay that high.
As you can see in the table, (again from Fangraphs) .145 is considered average. But, the M’s average is a below that mark at .131, 22nd in the league. This time, Kyle Seager does not lead the team. Instead Franklin Gutierrez is at the top with a huge (and unsustainable) .278, due to the fact that 7 of his 14 hits are for extra bases.
FIP- Fielding independent pitching has been my go to pitching stat, as I have bought into its message. It essentially suggests that there are only three outcomes that are in the pitcher’s control: strikeouts, walks and home runs. Hits are dependent on defense, which is something the pitcher cannot control. People sometimes criticize FIP because it “undervalues” guys like Jered Weaver or Matt Cain. Guys who aren’t strikeout guys and pitch to contact instead. But I would argue that strikeout pitchers are in fact more valuable because there are less balls put in play that could turn into hits. A strikeout is an out 99.9% of the time. Balls in play fall, on average, about 30% of the time. I’ll take almost certainty over 70% probably.
FIP is also generally thought best as a ERA predictor rather than a measure of what actually happened. And that is partially true. But that doesn’t mean it can’t be used both ways. It is often used along with ERA to kind of dig deeper into the ERA and tell us if it is what it “should be.”
This is the current formula as listed by Fangraphs: FIP = ((13*HR)+(3*(BB+HBP))-(2*K))/IP + constant. Again we see some weights to get the true impact of each outcome.
Above is another table, which tells us that 4.00 is average, very similar to ERA, as it is designed to be. For us Mariner fans, FIP can be a little disappointing at times since “our” pitchers tend to have deflated ERAs from pitching in Safeco. For example, fan favorite Jason Vargas looked to some like a #3/2 pitcher because of his high inning total and ERA. However, his FIP over the last three years are 3.95, 4.09 and 4.69 respectively, suggesting he wasn’t quite as good as some said.
And then, if you look at his xFIP — which estimates what a player’s HR rate should have been based on their fly-ball rate — it gets even worse. His last three xFIPs are 4.60, 4.45 and 4.45, mainly due to the fact that Safeco Field help suppress the fact that he was a big fly-ball pitcher and was susceptible to the long ball.
The current FIP leader on the M’s is, surprise surprise, Felix Hernandez at 2.34, which breaks the scale in the table above. However, a little more disappointment as Hisashi Iwakuma, who has looked very good thus far with a 1.99 ERA, has a FIP of 3.78, which suggests he isn’t quite as great as his ERA says he is.
So those are just a few sabermetrics that I feel are very useful in evaluating players, but often go unnoticed. It seems everyone wants to cling to WAR as thee sabermetric, and criticize that people use it as a be-all stat. In my experience, the people who look at is as a be-all end-all are the ones opposed to it, because they lack an understanding of the stats purpose.
The fact is, WAR is just one of many advanced metrics available to us (and there are plenty that I put ahead of WAR). Information is everywhere, most of it better than what we had before. Whether you choose to use it is up to you. Just know that things are changing, and if you attempt to argue with traditional stats against someone using sabermetrics, you will be out-gunned. That is just the nature of analysis and information.
What are your guys’ thoughts on sabermetrics? Are you more like Hawk and Harold, or Brian Kenny and Myself?