Why We Go To WAR: The Use of Modern Stats For Evaluating Offense

Pushing aside our usual analysis of the Indians, past and present, I’d like to explain why we at Burning River Baseball use the numbers we do and where they come from. While I don’t speak for everyone on the site, all our writers use a combination of traditional and modern stats and this is for those who either don’t know what those are or would like to know why. This will be the first of a three part series detailing the use of those newer numbers, starting with offense, then going to pitching and finally the hardest to evaluate of all, defense.

To begin, on a basic level, I like to use traditional stats (like wins, saves, RBI and runs) to describe what a player has done and the newer stats, like BABIP, UZR and FIP to observe how a player is playing at the moment or will in the near future. This doesn’t include WAR, however, which is a great tool for comparing players across different positions and eras. The primary reason for splitting the stats up this way is that counting and “old school” rate stats are unpredictable. From one year to the next, there is very little correlation between most of these numbers including wins, RBI and saves and they can never be used to predict the next season’s outcome. These numbers in particular have as much to do with opportunity as with talent, leading to year-by-year fluctuations, but even traditional rate stats, like ERA and batting average, generally don’t continue from year to year. Thankfully, we now have access to much more accurate predictors, the first of which is the often cited BABIP.

BABIP vs AVG

One of the best ways to normalize batting average is with BABIP. This stat, Batting Average on Balls In Play, removes the plays a hitter has the most effect on, like home runs, strike outs and sacrifice bunts and attempts to find out exactly how lucky a player is when he makes contact. Everything being equal, every player in baseball should have near a .300, or league average, BABIP. While this rate can be increased by hitting more line drives (by default, this means hitting less ground and fly balls), there is no number in baseball (at least as of now) that explains pure luck better than BABIP. When a hitter gets robbed by a diving outfielder and says the breaks never work out, he may be right, but the breaks can be recorded by this rate. Below is a simple chart showing Jason Kipnis’ averages, BABIP, strike out rate and line drive rate from the past three years. As would be expected, the higher his BABIP has been, the higher his average has been and to a lesser amount, the better his line drive rate, the better his BABIP. Used as a predictor, it would seem that Kipnis is legitimately somewhere between a .260 and .270 hitter and it wouldn’t be surprising to see him return to that level through the rest of his career.

While this is useful for predicting future performance to a point, personally, I find it makes no difference at all to the past. It’s the same case for considering opportunities for RBI and runs scored. Once something happens, it is what it is. There is no going back and getting another hit or getting moved higher in the lineup. It is not my place to discredit the efforts of those who played in the past based on what could have been, or to qualify a poor players who didn’t get opportunities because of bad luck. With this in mind, stats like line drive rate and BABIP are reserved for active players and not included in the analysis of historical players, such as those featured in the All-Time Indians series.

WAR

While WAR isn’t an offensive exclusive stat, this seems to be the best place to fit it in. Not just a cool acronym, Wins Above Replacement is the current penultimate stat for comparing players across positions and eras. While some dislike WAR because it is difficult to calculate (unlike even as simple a stat as OPS which is {[(H+BB+HBP)/(AB+BB+BHP+SF)] + [TB/AB]}), it is impossible to doubt it’s accuracy. Using the baseball-reference.com version (WAR varies between sources, another down fall), each of the top 72 players in baseball history are either in the Hall of Fame, on the ballot this season for inclusion, active or suspended (just Pete Rose). While the current “moralism” of the BBWAA may keep some of those players out, there is no question based on numbers alone that each of those top 72 deserves to be in the Hall.

While not consistent year in and year out, WAR is representative of what happened, not what is likely to happen. Because of this, it is a great indicator for who deserves to win the MVP each year and who should be in the Hall of Fame. By including hitting, pitching, defense and base running in one stat, it makes judging players much easier, comparing them with the one stat that truly matters in baseball, wins.

For those who don’t know, these are not actual wins, like are given to a pitcher who throws at least five innings and doesn’t give up the lead, but projected wins that a player’s stats would have earned. In addition, rather than just total wins, they are compared to “replacement level” players of their own time to equalize things across generations. This replacement player is the lowest level of Major Leaguer, a typical AAA call-up that doesn’t actually exist.

This comparison can fall apart some times, such as in 1925 when Yankees first baseman, Wally Pipp had a WAR of 0.0 considering the league average replacement player. Of course, his replacement player was actually Lou Gehrig, who averaged 6.6 WAR per season and had a 2.8 WAR in that rookie season, meaning Pipp’s wins over his actual replacement player would have been around a -3.0. This should be kept in mid with players like Nick Swisher, who had a -1.1 WAR in 2014. This number is made to look worse when considering that those at bats at DH could have been given to a replacement, Lonnie Chisenhall, who posted a 3.1 offensive WAR that was partially negated by a -1.5 defensive WAR. Chisenhall would then have been replaced at third by Mike Aviles, who finished the year with a 0.3 WAR, making him an almost exact replacement level player and better than Swisher overall.

While big decisions (either roster moves, MVP votes or Hall of Fame decisions) shouldn’t be made using WAR alone, it is a great tool to lead you on the right path. Better than average, OBP or even OPS and ERA, WAR is a big picture stat that can at least narrow down the discussion and lead to more intelligent decisions overall.