This debate has continually grown larger, and has begun to split the baseball community in two. Both sides continue to argue, yet there are many on either side that does not even know the other side's argument. In baseball and in life, mustn't one first understand the arguments of the other side before berating that side's point of view? There is no reason that sabermetrics and pure baseball scouting should be mutually exclusive. And once both sides are understood by the other, deeper and more meaningful debate can occur. For those who consider themselves to be "baseball purists," and who have not tried to learn about such complex numbers, or those who have simply not been exposed to such stats, here is a simple explanation of some basic sabermetric statistics which are used on this blog.
wOBA; For years, batting average, home runs and runs batted in reigned supreme as the main statistics for the analysis of hitters. There were obvious problems with this triple slash line. (Average did not reward walks, there are ways to hit for power other than with home runs and home runs is a counting stat, rewarding those who have had more at bats even if they are not as powerful, and Runs Batted In is too predicated on those around the hitter. A player's RBI's will depend too much on how often the guys ahead of him get on base. So then, OPS, or on base plus slugging percentage was created. One may say, "Oh, this is perfect! Power and on base in one rate stat. But OPS fails because it favors too much power hitters. The highest reward for getting on base (OBP) is 1.000, while the highest for power (Slg.) is 4.000.
So in comes wOBA, or weighted on base average. It is the most inclusive hitting stat available, and it is weighted properly, unlike OPS. In OPS, OBP and Slg. are weighted at a 1:1 ratio, while with wOBA, the ratio is 1.8/1. The basic formula for wOBA is (0.72xNIBB + 0.75xHBP + 0.90×1B + 0.92xRBOE + 1.24×2B + 1.56×3B + 1.95xHR + .25xSB - .5xCS) / PA. Yes, I know that seems crazy, but bear with me. For each possible event, (walks, hit by pitches, singles, doubles, triples, home runs, stolen bases, caught stealings) a value is attached. This value is indicative of the event. For example, a walk is worth .72 and a single .90, while a homerun is worth 1.95, as it is the best possible outcome of the event. Therein lies the advantage of wOBA; it is all inclusive. It rewards hitters for contact, power, on base ability and speed.
So how can I tell whether a player's wOBA is good or bad? Well, wOBA is scaled similarly to OBP. A 'league average' hitter will have a wOBA ~.320. Anything under .300 is bad, and anything approaching .400 is outstanding. .435 is Albert Pujols' career wOBA.
wRC+: wRC+ is a batting statistic very similar to wOBA. In fact, it is basically the same as it judges every part of a batter's performance from discipline to contact to speed on the bases. wRC stands for weighted runs created. It differs from wOBA in scale. wOBA is a decimal number, while wRC+ is a whole number with 100 being average.
So what is the advantage of wRC+? wRC+ is league adjusted, meaning it is adjusted to reflect a player's performance compared to that of his peers. For example, Davey Johnson had a .305 wOBA in 1968. Joe Girardi had a .304 wOBA thirty years later in 1998. Looking at this, one might say that they were the same hitter. However, 1968 was considered the year of the pitcher while 1998 was a year entrenched in the steroid era.
Hitters were better in Girardi's time making his .304 wOBA less impressive than Davey Johnson's .305 mark. This is reflected in wRC+, as Johnson had a wRC+ of 112, which is a good margin above league average, and Girardi had a wRC+ of 85, which is significantly lower than league average. Ramiro Pena, a dreadful hitter, will have a wRC+ of 42, while Albert Pujols has a career wRC+ of 172.
FIP: Pitching is also judged by rudimentary stats; W-L record and ERA. Judging a pitcher based on record is foolish. A pitcher's record is totally dependent on his team, much like RBI's. A pitcher should be judged based only on what he can control, and he cannot control whether his team makes an error, or decides not to score a run. Just look at Johan Santana over the past two years. ERA is not as bad as the other basic stats as it does hold value especially when looked at over a long period of time. However, there are some major drawbacks. For one, it is subject to two things which a pitcher cannot control; luck, and the quality of his defense.
How, you may ask, can luck and defense be removed from the equation? FIP, or Fielding Independent Pitching does the trick. It factors only what pitchers can control. (groundball/flyball rate, K rate, BB rate) Here is the formula; (13*HR + 3*BB - 2*K)/IP + C. What this formula does is provide weighted values for controllable events, homeruns, walks, strikeouts, to figure out what a pitchers ERA would be without luck or fielding. So FIP is on the same scale as ERA, making it easy to know a good FIP (3.00) and a bad FIP. (5.00)
UZR. Fielding is by far the most complex area of baseball statistics and the hardest to quantify. Fielding Percentage did not work because it did not reward range, and outfield assists does not work because it does not account for the rate at which runners are held or thrown out. So, UZR, or Ultimate Zone Rating was created.
I explained UZR in this post on centerfielders in New York. Joe Posnanski defines UZR as “The number of runs above or below average a fielder is in both range runs, outfield arm runs, double play runs and error runs combined.” So if a player has a UZR of 7, his defense has saved seven runs for his team more than an average player. Average is 0, and anything positive is above average, and anything negative is below average. It is a counting stat, so if Player A has 20 UZR and layer B has 15, Player A is not necessarily better as he could have twice the number of innings played. How, you may ask, can one measure this?
The most complicated measurement is range. In short, UZR splits the field into many zones. It averages out a player's proficiency in making plays in the zones around his position for his range factor. The problem with this is that it does not take into account the speed at which the ball was hit, therefore a large sample size is needed so that luck averages itself out. The rest is more simple. With outfield arm runs, it takes into account what happens in certain instances which an outfielder can control with his arm. (I.E. 1st-3rd on a single, tagging up from 2nd or 3rd, outfield assists, etc.) And error runs is simple as it judges how much damage the player causes based on their errors. Error runs is not all that relevant when judging outfielders as most outfielders make only a couple of errors each year.
WAR: Never before has there been a stat as all-inclusive as WAR, or Wins Above Replacement. What WAR is is an estimation on how many wins a player is worth to his team over a replacement level player. So what is a replacement level player? A replacement level player is the average fill in, from the minors or the bench. In other words, if there is an injury to a starter, the average player that would take his spot is called a replacement level player.
So how can one calculate such a stat? To calculate it for a position player, the estimated runs provided on offense and on defense are added together (RAR). Then the runs are adjusted for the league average replacement of that season and for position. Positional adjustments reward players who play premium positions, (i.e catcher, shortstop, second base, centerfield) as those positions on average provide less offense. For example, a .350 wOBA first baseman is less valuable than a .350 wOBA shortstop. RAR, or runs above replacement, uses wOBA and UZR to calculate how many runs a player is worth compared to a replacement level player. One War is approximately equal to ten RAR.
Since I analyzed Posada's career WAR in this post, I will use Posada as an example. Overall, Posada has garnered 230.7 offensive runs, -20.3 defensive runs, 220.8 replacement runs and 94.1 positional runs, for a grand total of 525.3 RAR, or Runs Above Replacement. Statisticians have figured that a team will win one more game per every 10 runs, so one WAR=~ten RAR. This leaves Posada at a hall of fame worthy 51.2 WAR.
Pitchers WAR is easier. It takes the pitchers FIP and innings pitched into account, see how much better or worse it is than a replacement level pitcher, and then calculate the RAR. And again, ~10 RAR=1 WAR.
WAR is a great measure used for a player of any position on their overall value, yet it is not as clear as "he has a higher WAR so he is a better player." It is always recommended that one look at a whole slew of other stats in addition to WAR when judging a player.
So what is a good WAR? A solid player will have 2-3 WAR in a season, a great player will have a 4-5, and a superstar player will approach 6 or more WAR. Albert Pujols will accumulate 7,8, or 9 WAR in his normal season.
That concludes The Subway Connection's explanation on advanced statistics. I hope I was as clear and in depth as I could have been. Also, it will not end with these