Statistical Metrics and what we do with them

Over the past several months, Counting Rocks has introduced numerous new tools to put into our statistical toolbox to attempt to quantify the performances of the players on the field.

A lot of questions have begun popping up recently about how to properly use these metrics, what they mean, and what performances mean what.

Well as we barrel down the stretch, we'll probably have a good number of visitors and new users who will have varying levels of understanding in terms of the more advanced metrics fit in to our analyses, and we want to see to it that we're all up to speed on at least a basic level, and for those new who aren't familiar with them, so that they can have a primer as well.

So let's begin with batting metrics.

For starters, we'll look at the absolute basic metric, a rate stat called Batting Average.

Batting Average is calculated by Hits/At Bats, obviously. It's difficult anymore to try and establish what the levels of "good" and "bad" are, but you know it when you see them. Looking at Batting Average in a vacuum, however, puts the MLB average batting average as .262. Typically, a .300 hitter is considered to be a very good hitter, and below that league average mark is considered to be bad to very bad.

Batting Average can be looked at as a sort of skill stat, as the rest of the traditional batting line is entirely fueled by AVG. Some guys are high average hitters, some are low average hitters, but AVG can be inflated by guys who walk a lot and deflated by guys who don't. It clearly doesn't paint the complete picture, and can be very affected by luck and slumps and normal statistical variance and birds and the sun and karma and what the batter ate for lunch.

But obviously, batting average isn't the end-all statistical metric, as it doesn't account for walks or how far the ball was hit, so we introduced some new ones to further illustrate a batter's performance, ability, and general contributions to his team's run-scoring ventures.

Join after the jump, where we'll talk about OBP, OPS, and beyond!

OBP is the logical follow-up to batting average. It's calculated by Hits+Walks+HBP divided by Plate Appearances.

Reviled by traditionalists as being a direct replacement for batting average by way of equating a walk with a hit, On-Base Percentage is considered to be a way of seeing how efficient a batter is with the outs he is allotted over the course of a season. OBP should be viewed as the percentage of times the batter comes to the plate and does anything that isn't an out - thereby creating a baserunner.

League Average OBP is at .333. An OBP around there will keep you in a lineup, and an OBP of .360+ puts you in the "good player" category, and .400+ puts you in elite company. Once a batter has an OBP below that average mark, there had better be an elite glove or some serious power to go with it.

High OBP is typically accompanied by high AVG, so the "skill" aspect of OBP is a bit harder to discern. We'll touch on that later.

The final slash line statistic is SLG. Slugging is Total Bags divided by At Bats. It basically just says how hard a guy hits the ball, how many bases he gets each time he hits the ball.

The Average SLG is sitting at a neat .419, and of all the metrics, SLG might be the most affected by the park a batter plays in. As Rockies fans, I think we all understand this a bit too well. If you have a guy slugging above .500, he's a good power threat. Above .600, and you're challenging the HR title. On the other end, if you're slugging between your batting average and that league average mark, you'd best be batting 1^st or 9^th (or behind a prolific basestealer), because you're a singles hitter.

So far, this has been a lot of "duh", but for those of us who already follow the advanced stuff, it's a good reminder of where the league averages are.

Finally, Runs and RBI. There really isn't a league average number for RBI, as they're too dependent on the number of games played and such, and they're far too situationally driven to really put an average number on. Colloquially, a 100-run guy or a 100-RBI guy would be considered "good".

Now, this is where it might get controversial, but I really have trouble putting runs and RBI into a "skill" category, as they're a product of a lineup (obviously excluding home runs). For starters, Runs and RBI are counting stats, simply meaning that they start at zero and go up from there. They can't go down, as opposed to a rate stat (AVG, OBP, etc). But we know that a good leadoff hitter with a competent offense behind him will have a good number of runs come season's end, and a good #3-#5 hitter with competent table-setters ahead of them will more than likely have a good total of RBI come October.

Runs and RBI are numbers I like to lump into "performance" stats, rather than skill stats. They tell you how well a player performed within his particular lineup over the season. They tend to not have much predictive value, but they do tell you how well a manager has utilized his best hitters. So not to entirely eschew the value of runs and RBI, they do bring benefit of seeing how well a team optimizes its offense. Although, runs scored by the team as a whole could also tell us similar numbers...

ANNNNNYhow, a lot of the above has been a lot of "duh", but I figured it'd be good to refresh the basics and check out the league averages, maybe put things into a bit of perspective.

Now, we delve into the SABR realms.

The obvious next step we take is to On-Base Plus Slugging, or OPS, for short. It's simply taken by adding OBP and SLG. That simple. The concept of OPS was really brought to the public perception by way of Michael Lewis' best-seller Moneyball, an account of how Billy Beane used this metric to build a successful baseball team that found value in other areas than the traditional measures of value that other big market teams put their money into. Rowbot Charlie77 covered Moneyball in a Fanpost awhile back, you should check it out.

Now, OPS is designed to combine a hitter's ability to get on base with his ability to drive runs along the basepaths, in the sense that you could put a decent #5 hitter and a good leadoff hitter and you'd see a similar OPS (one being more OBP heavy, and the other being more SLG heavy).

The average OPS in MLB this season is .752.  Typically, an .800 OPS is the line of who is a good hitter. Nick Swisher is a fun example of that, as he carries a career .812 OPS, with a decent balance of OBP and OPS. Moving forward, a .900 OPS hitter is typically your middle-of-the-lineup guy, probably one or more of your team's premier hitters. Once you start pushing .950-1.000 OPS, you're dealing with pretty elite batters. You'll usually only have a couple of them in each league every season. Well, nowadays anyhow.

Anything below that ~.750 OPS mark, again, you'd better be a defensive wizard or a leadoff hitter or something.

The issue with OPS is that it's also going to be very driven by batting average, as OBP and SLG also are. This is why we look at the ISO numbers for more of the player's skillsets.

You'll find ISO on most advanced stat websites to signify what I dub as IsoP, or Isolated Power, which is found simply by SLG-AVG. What IsoP tells us is how many extra bags a batter gets beyond the single. Similarly, IsoD is OBP-AVG, which tells us how often a batter walks or gets plunked.

The reason these are important is that they refer more to what the player does outside of his batting average. It's a statistical way of saying "He's developed a power swing" or that a player has improved his patience. Batting averages are very prone to slumps or hot streaks, but typically, a player's IsoD or IsoP will remain constant.

An important thing to note is that there really isn't a "good" or "bad" Iso number, they're merely descriptions of what type of a batter is being examined. Sure, a better IsoP means that the batter is hitting for more power, and a higher IsoD means that the batter takes a lot of walks (or is just an easy target for high-strung pitchers). For example, Ichiro has a career .045 IsoD. All this tells you is that he walks less than the average batter. He also has a career .100 IsoP, which just tells you that he doesn't hit for much power. Ichiro also has a career .333 batting average, so with those Iso numbers, we can pull out a career .333/.378/.433 slash line, good for a .811 OPS. Not spectacular, but this is where you look at his .378 OBP and .433 SLG and decide to bat him leadoff, where the power numbers may not matter as much.

On top of Iso numbers BABIP is an important metric to look at as well, purely because it can illustrate the fluctuations in a batter's performances. Simply put, if his OPS is low, but his IsoD and IsoP are right around normal, and his BABIP is also low, there's a good chance he's just getting unlucky, or basically just falling victim to the lower end of a normal statistical deviation. Average BABIP is .299 this season, just for the record.

So the final step in terms of batting metrics that we use is to get us all the way to wOBA. PF was nice enough to fill us in on what wOBA is in a past CR. Without being too boring, wOBA assigns a run value to every positive thing that you do as a batter, and then divides by plate appearances. It's designed to assign a number to a batter and use it to determine who your team's best hitters are without taking the context of spot in the lineup or if the particular skillset of the batter.

League average wOBA is right around .330 (and was engineered as such by Tom Tango). A "good" batter usually is around a .350 wOBA. .375 and up, you're talking a top notch hitter, probably right in the middle of your lineup. A wOBA of .400 and up, and we're in the elite category. Once you get up to .450 wOBA, you have an MVP candidate. A player sporting between .300-.330, again, falls into that gloveman category. Below .300 wOBA probably means you suck or just aren't ready.

To touch back on the "skillset" idea, take a look at Nick Swisher and Nick Johnson. Both are sporting a .368 wOBA this season, yet Swisher is batting .245/.372/.478, while Johnson's slash line is .296/.419/.408. wOBA tells us that both batters provide a similar value to their lineup, but in very different ways. But if you compared Nick Johnson's .419 OBP to Hanley Ramirez' .419 OBP, there's a .060 dropoff from Hanley to Johnson, because Hanley's power numbers add to his performance far more than Nick Johnson's. It's interesting to note, however, that it's hard to find a batter with similar SLG to Hanley's with a wOBA more than .015-.020 away, because a good OBP tends to go along nicely with a high SLG.

It should be noted that wOBA is deliberately scaled to match OBP.

If you follow pretty closely, you'll notice a correlation with .350 wOBA guys and .800 OPS guys, and .375 wOBA guys with .900 OPS guys. So you might say "why don't we just stick with OPS? Why do we need wOBA?" well, the issue has to do with the weighting of the component stats.

Specifically, OBP is not equally valuable as SLG, point for point. PF pointed a lot of this out in the post that is linked above. wOBA gives more accurate weights to offensive events (basically a double is not twice the value of a single). wOBA also has a specific meaning to it, while OPS is just kind of a number that can only be compared to itself.  Like with batting average, you can multiply a batter's average by his number of ABs, and you'll get the number of hits he should have over that number of ABs. If you multiplied OPS by a number of ABs, you'd get a number that doesn't mean anything. But with wOBA, you can multiply by the number of PAs in question, and you'll have a good idea of how many runs that player was worth.

Also, if you're into more advanced metrics than just AVG, and you like OPS, but think wOBA is a bit too complicated, then that's fine. For the most part, OPS and wOBA correlate pretty directly.

So to give this a quick summary, here's a table:

Make sense? I hope this can help us all in being able to discuss the batting metrics as we move forward toward the postseason.

Please ask any questions you may have, and I hope this article and the comments can answer any questions we have about batting metrics, and how to properly view them.

Next week we'll review pitching metrics, and all the fun and joy they bring with them!