Stats Glossary: 2016 Edition

As you know, we use a few non-traditional statistics on this site.  I wouldn't consider this a stat-heavy blog, so I'm not going to make this more complicated than it needs to be.  To many of you, this will not be new information and some of you have knowledge way beyond what will be presented here.  Consider this a primer for an understanding of new metrics and also a beginner's reference glossary.  We will leave the main body of this up in our Stats Glossary on the main menu of the blog.

Traditional stats like average, ERA, RBI, SBs, Fielding %, wins, losses, etc. all tell us something about the game, but they don't tell us as much about player the player's individual value because it doesn't discriminate enough from outside factors such as luck, hitting environment, and the quality  of the team/staff/lineup/defense around him. Because of that, they tell us less about the individual performance of a player.  That, in turn, means we don't the same level of predictive value.

This will not be a deep look at new metrics.  I will leave that to more statistically-inclined, but this is an introduction to the metrics you will see on our blog with a brief explanation of what they are and why we think they're more informative.  They're also some of the most basic of the new metrics, so if you are interested, you can use this as a launching point to learn about some of the interesting and more complex metrics.  For example, we will cover FIP, which should at least give you a basic understanding of how similar metrics such as SIERA or dERA (Deserved Run Average).

This is statistics only, so when we talk about things like value (such as we will in WAR), we are talking about quantifiable value.  That is, the kind of value we can measure with objective data.  Thus non-quantifiable qualities such as a players mental makeup or leadership are not factored in here. They certainly add to a player's value and may even impact the value of teammates in a positive way, but because we cannot put that into numbers (at least not yet), so those qualities aren't included for the purposes of this piece.

We'll start with stats that can be used for both pitchers and position players and one that attempts to summarize the overall value of a player in one neat and tidy number.

WAR: Stands for Wins Above Replacement.  We prefer to use Fangraphs' version (or fWAR) because I feel it does a better job of separating individual performance from external factors.  In general, WAR is an all-encompassing number for statistics that a player can control.  For position players that includes hitting, OBP, power, defense, arm, and baserunning.  It is important to note that the position a hitter plays on defense is important.  As we will discuss in the hitting section later, an .800 OPS from a SS adds more individual value than an .800 OPS from a first baseman, so that in turn will affect their WAR.

WAR is a relative value that attempts to measure how much wins that player provides in comparison with a random, everyday joe AAA guy.  That latter player is considered to be "replacement level".   Thus, as far as WAR is concerned, Buster Posey was worth about 6 more wins (5.7 WAR) over the course of a season in comparison with a replacement level player like Taylor Teagarden.

It is important to note too that "wins" aren't in the context of actual game outcomes, but rather a sum of events, each of which has a set value.  For example, a double, single, walk, defensive play, etc. all have a certain value.  Those values add up until they are worth a theoretical run created (or saved) and then those theoretical runs saved/created turn into theoretical "wins".

For the purposes of WAR a game-winning single has no more value than a single when the team is down 10-0.  The reason is that WAR only attempts to assign value to a player as an isolated individual.  Game-winning hits are dependent on external factors such as teammates getting on base, their basereunning ability, perhaps the outfielder's throwing arm, etc.  It even is dependent on what happened earlier in the game.  WAR filters all of that out and only sees the single, walk, defensive play, etc. as an event with stand-alone value.

Those theoretical wins have been shown to correlate with actual team wins.  When the Cubs added Jon Lester, for example, he was a 5 WAR pitcher.  Since he essentially replaced a bottom of the rotation starter (who was essentially replacement level), that 5 WAR value means that Lester was worth 5 more wins in the standings for the Cubs in 2016.  Other upgrades aren't as cut and dried.  For example, the Cubs added Ben Zobrist to play 2B.  He projects as a 3.2 WAR player.  The player he is replacing, Starlin Castro, projects as a 2.2 WAR player.  So if all things remain equal, that 2B upgrade is equivalent to one more win in the standings for the Cubs in 2016.  That is not factoring in the player the Cubs received in return for Castro, pitcher Adam Warren.  Warren projects to 2 wins.  So if we take that sequence of moves as one larger event, then the Cubs upgraded their team by 3 wins (Zobrist + Warren - Castro = approx. 3 wins added).

It is a broad measure of relative value and it is not predictive, so while it is a very useful snapshot, it is not intended to be a number that ends all arguments on a players value.  As an example, David Ross and Jorge Soler rated as  replacement level players last year but we would hardly call them "replaceable" in the grand sense.  Ross' value as a leader was immeasurable and Soler's long term potential is far more important than that low present day value, which was largely due to the negative impact of his defense.  To give you an example of how WAR numbers break down, here's a quick reference chart.  In some cases, we'll have to do dome rounding up and estimating.

Below 0: This is Jonathan Herrera level. It's a substandard MLB player who theoretically should be replaced because when he is on the field, he costs your team run, and, ultimately, wins.  Of course, Herrera was a 25th man who was never going to provide much on the field value anyway, so context is important.  He gave the Cubs value in terms of fitting in on the bench and in the clubhouse, providing solid defense at multiple positions and, of course, the rally bucket.  So the lesson here is that if you can't be a great player on the field, you can extend your career somewhat by providing value (or at least not being a disruption) off of it.  As far as pitchers go, we're usually talking about middle relievers on the AAA shuttle, guys like Brian Schlitter and Zac Rosscup.

0: A replacement level player. As noted above, David Ross and Jorge Soler were examples (see intro above), though neither player could truly be called replaceable outside the strict WAR interpretation of the word.  Perhaps a better example is Taylor Teagarden, a player who doesn't add any on the field quantitative value and from a numbers standpoint, can readily be replaced by a lower salaried AAA player or waiver wire pickup.  Middle Relievers are generally somewhere between 0 and 0.5 WAR.  Examples with 30 or more IPs include Clayton Richard, James Russell, and Edwin Jackson -- and we can see how interchangeable these pitchers were over the course of the 2015 season.

1: A solid reserve level player or good, set-up level bullpen arm. Chris Denorfia was a good example.  Starlin Castro was also just about a 1 WAR player, though he provided more value down the stretch. For pitchers, it's a good bullpen arm.  For reference, on the low side we have 7th inning reliever types.  Jason Motte (0.5) and Justin Grimm (0.6) are good examples.  Closer to the 1 wAR mark we have set-up men such as Pedro Strop (1.1) and good swingmen like Travis Wood (1.3).

2: The minimum of what you want out of your starter--- or an impact role player/high leverage bullpen arm Miguel Montero was a 2 WAR player, providing average production on both offense and defense, though we can say his value to the Cubs came with the nuances (leadership, pitch-framing, game management -- he certainly seemed to have a positive impact on Jake Arrieta, for example).  Kyle Schwarber was a 2 WAR player but that is somewhat misleading because he only played a 1/2 season with the Cubs and was below average on defense.  His offense was among the team's best, but we will get to that later.  When it comes to role players, Javier Baez projects to a 2.3 WAR season and if he can do that filling in at multiple positions, that makes him a very valuable asset off the bench.  When it comes to pitchers we're talking about solid, 1st division bottom of the rotation starters such as Jason Hammel (2.4) on the high end and closers like Hector Rondon (1.6) on the lower end of that range.

3: A good starter.  Castro was often around this level during his Cubs career.  That made him a pretty good player, but not a star.  2015 players who fit under this category are Addison Russell (2.9), whose plus defense gave the Cubs a big boost overall.  Dexter Fowler (3.2) also was in this range with his surprisingly adequate defense and strong top of the order skills.  Chris Coghlan had a strong and vastly underrated season on both offense and defense   While he had an off year in 2015 coming off an injury, Ben Zobrist figures to fall into the 3-3.5 WAR range in 2016.  When we're talking a 3 WAR pitcher, we're talking about a good starting pitcher who best slots in the #3 or # 4 spot.  Kyle Hendricks (3.4) fits here.

4: A very good starter/0ccassional all-star: Cubs didn't have a position player at this level in 2014, but John Lackey (3.6) fit the bill off the mound in 2015.  At the lower end of this range like Lackey, we're talking about a very good mid-rotation starter and a solid #2 at the upper range.

5-6 A great, all-star level impact player.  This is where the fun really starts.  We can start with Anthony Rizzo (5.5) who had a tremendous all-around offensive season.  There is still room for improvement as Rizzo is just entering his prime age season (27 yrs old), had a substandard year defensively, and peripherals which leave wiggle room for better numbers.  On the pitching side we have Jon Lester at (5.0).  Lester is a great example of how a pitcher provides value well beyond traditional measuring sticks such as W-L record, which was 11-12.  The 5 WAR season ranked 8th in the NL and 2nd among "#2" pitchers (behind Zach Greinke).  We think of a 5 WAR pitcher as a top of the rotation arm, which gives the Cubs two frontline pitchers.

6-7: A superstar, MVP contender level player.  We will start this with Jason Heyward (6.0).  Heyward isn't your  classic superstar.  He won't hit 40 HRs or win batting titles, but he is the kind of player that does all the little things that add up to a lot of value.  His defense and baserunning provide tremendous value on top of his well above average offensive numbers.  He'll give you great ABs while getting on base and providing the occasional HR, SB, etc.  Next we have Rookie of the Year Kris Bryant (6.5) who is actually very much like Heyward in that he provides great baserunning and in Bryant's case, surprisingly good defense on top of very good, grind it out offense. Bryant has more power than Heyward and is already the better offensive player overall, so if he can continue to improve that defense, he has a good chance to reach the next level.  Jake Arrieta was the highest rated Cubs at 7.4 which makes him a borderline elite player such as the ones below.

8+: A rare player.  One of the very best in the game and if played at this level consistently, of  his generation.  In 2015, the only players with 8 WAR or above were: Bryce Harper (9.5), Mike Trout (9.0), Josh Donaldson (8.7), and Clayton Kershaw (8.6)

BABIP: Stands for Batting Average of Balls In Play.  To some degree, this stat measures "luck" on batted balls.  If your BABIP is high, it means a lot of batted balls fell for hits.  In other words, more seeing eye groundballs and bloop hits and less line drives right at ‘em.  In Wee Willie Keeler's words, “hitting them where they ain’t”.  A higher number can mean less luck for a pitcher (and more luck for a hitter).  The league average for this statistic tends to hover around .300.  Note that it isn't all luck and BABIP varies from player to player.  Faster players tend to have higher BABIPs because they'll beat out a few extra ground balls.  Quality of contact also factors in. Some players, such as Kyle Schwarber, Kris Bryant, Jorge Soler, and Anhony Rizzo make hard contact consistently and that can certainly have an effect on how many balls get past (or through) the defense.  It is not surprising then, to see both Bryant and Soler among the Cubs top BABIP performers.  An all-fields approach also helps because it is tougher to defend and neutralizes defensive shifts designed to minimize the "hitting them where they ain't" factor.  This is a second advantage Bryant has in terms of high BABIPs.

Hitters Only:

First off, one thing you'll often see me use when describing a player is a "slash" line.  It looks something like this: .305/.367/.489.

  • The first number is batting average,a simple ratio of hits to official ABs.
  • The second number is OBP, a ratio of the times a player gets on base via hit, walk, or HBP versus the  total number of plate appearances
  • The third number is slugging percentage, which is calculated like batting average except that it assigns greater value based on the type of hit (1 for single, 2 for double, 3 for triple, 4 for HRs).

OPS

  • The last two ratios of the slash line added together (OBP and Slugging Pct%) added together is known as OPS, which is a crude but useful measure of a players all-around contribution at the plate.  An OPS of .800 is a general benchmark for what is considered "good".

Note:

The NL average OPS in 2015 was .713, though that varies by position the hitter plays on defense.  The average OPS for a 1B was .793 but more than 100 points lower for catchers (.686) and shortstops (.673).  This is a basic example and explanation of why teams put a premium on up the middle positions.  It is difficult to find players who can play those positions at an MLB quality level while also providing above average offense.  Thus, a SS with an .800 OPS would be considered much more valuable than a 1B with an .800 OPS, assuming both play average defense.  As we alluded to above, the different levels of  offensive expectations based on defensive position is a major reason why WAR gives greater weight to the more difficult defensive positions when measuring overall value.

Here is the projected 2016 Cubs lineup with leaders by position in OPS for 2015 in comparison with the NL average at each position in parenthesis...

  • C:  Miguel Montero --.754 (.686)
  • 1B: Anthony Rizzo --.893 (.793)
  • 2B: Ben Zobrist -- .809 (.700)
  • 3B: Kris Bryant -- .858 (.745)
  • SS: Addison Russell -- .696 (.673)
  • LF: Kyle Schwarber -- .842 (.735)
  • CF: Jason Heyward -- .797 (.734)
  • RF: Jorge Soler -- .723 (.763)

ISO%: Isolated Power.  This subtracts batting average from slugging.  It is a crude measure of extra base power because it's purpose is to remove singles from the equation.

Now, we'll look at 2 more advanced metrics to measure offensive value using the same  players we used in the previous examples

wOBA:  wOBA is my favorite all-encompassing offensive metric that combines both slugging ability and OBP and weighs the two more evenly than OPS (OBP + Slugging).  It's calculated so that it reads like OBP numbers.  So .330 is about your average offensive player in the NL, while .400 would be superstar quality on offense .  If you're below .300,  you're a liability on offense.  Some 2016 Cubs for reference with NL average by  position in parenthesis...

  • C:  Miguel Montero --.328 (.296)
  • 1B: Anthony Rizzo -- .384 (.336)
  • 2B: Ben Zobrist -- .349 (.307)
  • 3B: Kris Bryant -- .371 (.319)
  • SS: Addison Russell -- .304 (.297)
  • LF: Kyle Schwarber -- .364 (.317)
  • CF: Jason Heyward -- .346 (.320)
  • RF: Jorge Soler -- .312 (.327)

Note: Chris Coghlan was at .337, which ranked 4th among Cubs hitters with 100 or more PAs in 2015.

RC+:  Stands for Runs Created and is like wOBA but measures total offensive value on a different scale.  Like IQ, it uses 100 as a baseline for it's average.  In the case of RC+  125 is a very good offensive player, 150 is excellent,and 75 is a liability on offense.  For a top of the scale reference, Bryce Harper put up a 197  RC+ in 2015.  Barry Bonds once put up a 244 RC+.   The advantage here is that it makes it is easiest to use.  You don't have to worry about looking up the average RC+ for any given season as you do with wOBA.  RC+ is always calculated so that 100 is your average.

As far an average RC+ on the Cubs goes, Jorge Soler (96) and Miguel Montero (107) were the closest, but again, Montero's mark gives him more overall value considering his defensive position.  Here are more 2016 Cubs for reference with NL average by  position in parenthesis...

  • C:  Miguel Montero -- 107 (85)
  • 1B: Anthony Rizzo -- 145 (113)
  • 2B: Ben Zobrist -- 123 (93)
  • 3B: Kris Bryant -- 136 (101)
  • SS: Addison Russell -- 90 (85)
  • LF: Kyle Schwarber -- 131 (99)
  • CF: Jason Heyward -- 121 (101)
  • RF: Jorge Soler -- 96 (.107)

Note: Again, Chris Coghlan ranked 4th among Cubs hitters with 100 or more PAs in 2015.  This time with an RC+ of 113.  Jorge Soler was the only player below the NL average at his position for both wOBA and RC+.

DEFENSE

You won't find me using fielding percentage which doesn't emphasize range or arm.  If I do talk about defense, we'll go with UZR/150, which isn't perfect but it does take more factors into account (i.e. errors, range, arm strength).  The "150" extrapolates the number out to 150 games.

Think of it this way:  Who's more valuable to your team? A guy who can makes 100 plays because of his range and arm, but commits 10 more errors or the guy who can make 85 plays but makes no errors?  Basically, the first guy helps you create 90 outs on defense compared to 85 and, theoretically saves you runs over the course of a season.  Like WAR, it is a statistic that is relative to it's environment.  The number in UZR attempts to measure the runs a defender saves relative to a players peers at his particular position,, so if his UZR/150 is 0, then he is a relatively average defender..  The player neither saved you nor cost you runs with his defense when compared to his peers.

Keep in mind that with UZR, there are some weaknesses.  There can be large fluctuations and the larger the sample size the better, so it usually works best if you can go back 2-3 years.  I've also noted what seem to be blind spots, such as defensive positioning.  We noted last year, for example, that CFers who tend to play shallow perform worse in terms of UZR.  Simply changing the positioning seems to significantly change that player's rating, as it did with Dexter Fowler in 2015.  Considering all the changes to defense positioning in baseball (i.e. the exaggerated shifts), UZR will have to continue to adapt to evolving baseball philosophy.

UZR/150

  • 1B: Anthony Rizzo : 3.4
  • 2B: Ben Zobrist: (-13.3)**
  • 3B: Kris Bryant: 4.8
  • SS: Addison Russell -- 8.6
  • LF: Kyle Schwarber -- (-4.1)
  • CF: Jason Heyward -- 75.5 in just 50.1 innings; 22.3 in RF
  • RF: Jorge Soler: (-12.7)

** Zobrist was injured and had an outlier season defensively.  He is at a very strong 8.2 UZR/150 for his career at 2B, including 10.3 in 2014 and 14.7 in 2013.

Other noteworthy ratings in limited action...

  • Javier Baez: 70.1 (!)
  • Austin Jackson: 48.3
  • Chris Denorfia: 27.7
  • Chris Coghlan: 18.9

CATCHING

There is no set established, all-encompassing way to measure a catcher's defensive worth as a whole, but Baseball Prospectus is making an attempt with their just developed Catchella metric.  I won't go into it here, but Jared makes an excellent explanation here -- and the good news is that Miguel Montero is very good.  He ranks 8th in all of baseball.

PITCHERS

FIP

Stands for Fielding Independent Pitching.  This number will look like an ERA but only factors in what a pitcher can control: namely strikeouts, walks, and HR rate.  The attempt is to measure how a pitcher would perform without the influence of the defense, luck, etc. behind him.

It is an important number because, for better or worse, any given pitcher's FIPs tend to change accordingly as their surrounding environment changes.  For example, a pitcher with an FIP lower than his ERA is a good candidate to improve if he goes to a team with better defense.

The average FIP in the NL in 2015 was 3.88.  It was 3.97 for starters and 3.72 for relievers.  Here is how the Cubs projected 2016 staff rates compared to that average...

*Denotes swingmen with 2015 starts

SPs (3.97 FIP)

  • Jake Arrieta: 2.35
  • Jon Lester: 2.92
  • John Lackey: 3.57
  • Kyle Hendricks: 3.36
  • Jason Hammel: 3.68
  • Adam Warren: 3.92 as an SP

RPs (3.72)

  • Travis Wood (2.53)*
  • Hector Rondon (2.68)
  • Adam Warren (2.71 as an RP)
  • Justin Grimm (3.11)
  • Trevor Cahill (3.13)
  • Pedro Strop (3.16)
  • Neil Ramirez (3.21)
  • Clayton Richard (3.26)*
  • Carl Edwards (3.35)
  • Rex Brothers (4.49)

Walk Rate

The average walk rate in the NL in 2015 was 2.95 BBs per 9 IP..  It was 2.74 per 9 IP for starters and 3.34 BBs for per 9 IP for relievers.  Here is how the Cubs staff rates compared to that average...

SPs (2.74/9 iP)

  • Jake Arrieta: 1.89
  • Jon Lester: 2.06
  • Jason Hammel: 2.11
  • Kyle Hendricks: 2.15
  • John Lackey: 2.19
  • Adam Warren: 2.81 as a starter

RPs (3.34/9 IP)

  • Clayton Richard: 1.85
  • Hector Rondon: 1.93
  • Adam Warren: 2.23 as RP
  • Trevor Cahill: 2.65
  • Pedro Strop: 3.84
  • Neil Ramirez: 3.86
  • Travis Wood: 4.19
  • Justin Grimm: 4.71
  • Carl Edwards: 5.79
  • Rex Brothers: 6.97

**Note that the walk rate may also be expressed as a percentage.  For starters, that average is 7.2%.  For relievers, it is 8.7%.

Strikeout Rate

The average K rate in the NL in 2015 was 7.88 Ks per 9 IP .  It was 7.55 per 9 IP for starters and 8.34 Ks per 9 IP for relievers  Here is how the Cubs staff rates compared to that average...

SPs (7.55)

  • Jake Arrieta: 9.28
  • Jon Lester: 9.09
  • Jason Hammel 9.07
  • Kyle Hendricks: 8.35
  • John Lackey: 7.22
  • Adam Warren: 6.28 (as a starter)

RPs (8.34)

  • Justin Grimm: 12.14
  • Trevor Cahill: 11.65
  • Travis Wood: 11.02
  • Pedro Strop: 10.72
  • Neil Ramirez: 9.64
  • Adam Warren (9.42 as an RP)
  • Hector Rondon: 8.87
  • Carl Edwards 7.71
  • Clayton Richard: 5.18
  • Rex Brothers: 4.35

**Again, we may sometimes express the strikeout rate as a percentage.  For starters the average rate is 19.9% and for relievers it is 22.2%.

GB Rate: Whether you prefer a ground ball pitcher depends in part on the ballpark and your defense.  In the case of Wrigley, that preference could change over the course of a season.  Flyball pitchers tend to do well in the spring when the weather is cold and the wind tends to blow in.  Wrigley tends to play like a big ballpark early in the season.  But come summer time you may want a pitcher who can induce hitters to hit the ball on the ground.  Market inefficiencies can come into play as well.  If teams are valuing GB pitchers highly, then a team in a big ballpark such as the Padres may be able to get a good deal on a flyball pitcher.

The average GB rate in the NL in 2015 was 46.3% (virtually the same for SPs and RPs).  Here is how the Cubs staff rates compared to that average...

  • Trevor Cahill: 61.8%
  • Clayton Richard: 59.8%
  • Carl Edwards: 58.3%
  • Rex Brothers: 56.3%
  • Jake Arrieta: 56.2%
  • Hector Rondon: 52.4%
  • Kyle Hendricks: 51.3%
  • Pedro Strop: 51.3%
  • Jon Lester 48.9%

------Avg. 46.3%------------

  • John Lackey 46%
  • Adam Warren: 45.2%
  • Justin Grimm: 45%
  • Neil Ramirez: 38.5%
  • Jason Hammel: 38.3%
  • Travis Wood: 34.5%

So this is mostly what I'll use and I'll consider this a permanent work in progress.  I'll add more if I use them or subtract some if I don't find them useful anymore or something better comes along.  We'll also be changing the examples from year to year.

We are considering adding a few more.  If there is anything we use on the blog that you don't understand, please let us know and we will add it.  What we add may also depend on what Dabs, Myles, and Jared want to use on their recaps and/or analytical articles.  For example, Dabs often uses WPA in his recaps and that is something we can add as well.

 

Advertisement: