Burning Questions: What Is Your Favorite Baseball Statistic?
The numberFire baseball staff will be posting a weekly feature called Burning Questions. The idea is simple: we pose a general question to the numberFire baseball staff, getting contributors to provide an answer and an explanation on the particular subject.
This gives you, the reader, a chance to hear opinions from many different experts, who, believe it or not, don't always agree on everything. What we do have in common is a knowledge of and love for the game, and we want you to be a part of the conversation. Feel free to pose an answer to this or a future Burning Question on Twitter, or tell us why you agree or disagree with one or more of our answers. These features are designed to start the conversation, not to offer a comprehensive solution, and often there is not a clear correct answer.
And now, our answers to this week's burning question: What is your favorite baseball statistic?
Average Fastball Velocity
Bradley Wilson's thoughts:
One of the best parts of looking back at Moneyball is the recognition that, 12 years ago, the A's front office got ahead of the pack with some analytics technique that seems sort of primitive today. It's completely true that the basic tenant â€“ small-market teams need to exploit inefficiencies in the market to reliably compete with wealthier clubs â€“ remains completely true. And players whose main value was their ability to reach base represented an untapped market inefficiency (which has since balanced itself).
The primitive part comes from the front office's rejection of some of the other conventional ideas about how to draft, develop, and measure the productivity of players. It's been proven since then, for example, that athletic players with foot speed develop at a higher rate, play better defense, and age better than the kinds of players that the A's were targeting (the Jeremy Browns of the world). Scouts already knew that, but the A's didn't see evidence â€“ at least, not enough evidence to convince them to spend the extra money trying to lock up athletes whose on-field contribution may be primarily defense, for which a single universal or quality metric remains elusive. They could see Jeremy Brown's college OBP, and know exactly how much that helped his team.
Since then, scouting and sabermetrics have begun to merge. We've begun to quantify what conventional wisdom has always held up. This leads to stats that measure not a player's production, but his process.
These stats give a person who doesn't get a chance to see a player every day some idea of what a scout might see: not just how good a player was at certain things, but what the player does on the field to gain the advantage. How many line drives does he hit? How well can he hit a curveball? How often does he swing and miss?
My favorite process is evaluating how hard a pitcher throws. Average fastball velocity is a process stat that tells you a litany of things that no amount of examining a pitcher's FIP can do. You can't tell how good a pitcher is â€“ you need FIP, ERA+, or whatever to do that. But you can measure trends in a player's career and try to identify an injury or a potential tank year (hello, C.C. Sabathia). With his downward trend in fastball velocity, how long 'til Justin Verlander is an average pitcher?
You can also measure how likely a young guy is to succeed â€“ the guy who throws 96 is a lot more likely to maintain that sparkly minor-league K rate than a guy who throws 89. It can be an indication of repeatability when the sample size is too small to get a handle on whether a bullpen arm, for example, is a good investment â€“ you only need one radar gun reading of 98 to know that a guy can throw 98. You need about 500 innings of baseball to get a reliable reading on whether a player's FIP lines up with his real ability. Lots of quality bullpen arms never reach that number in their whole career.
There are plenty of guys who are great and throw 88-90, so you can't just look at fastball velocity; it only makes sense in conjunction with other stats. But no other stat tells you what kind of pitcher somebody is better than the one that tells you how hard they throw. What could be purer?
Jim Sannes' thoughts:
Scoring runs in baseball is important, right? Well, some teams may act like they're not (I'm looking at you, Jeffrey Loria), but I'm going to pretend that they are for the next few paragraphs.
As a Twins fan, it may be a surprise that I do enjoy runs. I also enjoy correlation coefficients. So, I decided to combine the two to see which rate stat (between OBP, SLUG and wOBA) had the highest correlation coefficient with runs scored over the last three years.
The winner (in a close contest) was wOBA. Its correlation coefficient with runs scored was .9477. Slugging came in second at .9209, while OBP was third at .8748. (For fun, I did the same thing with batting average, which clocked in at a poor .7844).
Because of this, wOBA has earned my undying love, and is my favorite stat. Yes, it's easy to win my affection. How else would you explain my being a fan of the Twins and the Jets?
Daniel Lindsey's thoughts:
Home runs, batting average, and runs scored are probably three of the most commonly used statistics in baseball, but they don't tell the whole story. The more I write, the more information I want to have so I can tell a better part of the tale. That makes BABIP my favorite statistic.
BABIP stands for Batting Average on Balls In Play. While a player's average just shows how often a batter doesn't get an out, BABIP will take into account fly outs and ground outs, not accounting for strikeouts and walks. BABIP isn't necessarily used to determine a player's success, but it can give an indication of how a player might do on hit balls over a period of time.
Traditionally, folks have dubbed a .300 BABIP (roughly) to be an expected. As a result, players above the mark were said to have more luck, and were likely to soon regress. However, things like player speed and line-drive rates can change this expectation, meaning a high BABIP can be sustainable for a player who can simply hit the ball well. Like Mike Trout.
BABIP can still help us project regression though. If a player has a similar batted ball profile from one year to the next, but his batting average rose, it could be a result of his general luck. Michael Cuddyer, as an example, saw a giant jump in numbers last year, but he also had the third-highest BABIP in the bigs. Considering this, there's little chance he hits for the same type of numbers he did in 2013. That's when BABIP can come in handy.
Dan Weigel's thoughts:
What is it good for? Absolutely nothing!
Well, maybe not. War did give us this jam and also is a pretty useful statistic for evaluating baseball players.
Iâ€™ll say it now, WAR is not a perfect statistic and there are disagreements on the best way to calculate it. But the advantages are tremendous. WAR is a statistic that considers every contribution (or lack thereof) on a baseball field by any given player over any given timeframe on a scale easy to understand and use for other calculations, such as proper price tags for free agents. Instead of evaluating hitters by AVG/HR/RBI and pitchers by W-L/ERA/K, WAR gives us a more thorough evaluation by considering and properly weighting each category relative to its effect on winning baseball games.
Moving beyond this, WAR is useful in comparing players of different positions and/or among different ERAs. Was Andrew McCutchen more valuable than Clayton Kershaw over a given time frame? WAR can answer that question. Did Barry Bonds have a better career than Babe Ruth? WAR can answer that as well.
The statistic is context dependent, meaning that it's generated by comparing a player to the league average to account for the fluctuations in league averages over time. This makes Pedro Martinezâ€™s 2001 season, where he posted a 1.74 ERA while the next best qualifier, Roger Clemens, posted a 3.70, much more valuable than Kershaw in 2013 or Bob Gibson in 1968.
The two most prominent versions of WAR are found on the websites for FanGraphs and Baseball Reference. Averaging these versions is typically seen as the most effective way to quantify WAR, though letting a version stand alone is still quite useful.
Ladd Davies' thoughts:
The first statistic most people pay attention to when looking at a pitcherâ€™s stat line is ERA. Why not? It gives you the average runs a given pitcher allows for every nine innings pitched. Though at first glance it can seem like a decent indication of a pitcherâ€™s success, it can be quite misleading.
A pitcher may have a pretty ERA that sits below 3.00, but getting there could be quite nerve-racking. As a Red Sox fan, I know this all too well. Watching Daisuke Matsuzaka in Boston was quite the exercise in patience. He had a talent for getting two runners on with no outs, and would somehow get a strikeout and a lucky double play. No damage done. But this would eventually come back to bite him by way of a five-run inning.
On the other hand, middle relievers can have a huge earned run averages because of one bad inning, and they'll have a tough time lowering it due to their limited innings on the mound.
This kind of inconsistency can be very frustrating for fantasy baseball owners especially - there had to be a way to indicate this with some numbers. Baseball is all about numbers, right?
A greater barometer for pitching consistency exists, and itâ€™s called WHIP. One of fantasy baseballâ€™s pioneers, Dan Okrent, invented WHIP in 1979 with the idea to create a stat that shows the number of runners a pitcher allows on base each inning he pitches. He did this by adding the total number of walks and hits then dividing by total innings pitched. As a result, we can get a much better idea of pitching consistency without as much luck being involved.
Max Scherzer was the only starting pitcher to average below a single runner allowed per inning (.097) last year, for instance, and he now has a Cy Young award to show for it. Coincidence? Of course not - it's the power of WHIP.