What do we do?

May 13, 2008


While that sounds entirely too simple, it probably is. However, it’s the best solution.

We know that everyone can find a set of numbers to fit their argument. The beauty of baseball statistics is that almost any statistical category can be used for any argument. However, statistics like VORP are attempting to make it so that journalists have no excuse to make stupid arguments.

Baseball has always been a game of traditions, so it’s only natural that the statistical revolution is treated sometimes like a black person at a KKK convention. The smart people know that racism is wrong, but there’s a group of it who is contrary and is pretty vocal with their opinions. (No, I am not trying to equate the two)

Cheating has always been a part of baseball. Spitballs, pine tar, the hidden ball trick; you can go on and on about the traditions of cheating. Steroids were no different. Bud Selig did nothing to disprove that baseball doesn’t like cheating–because cheating is engraved in the game–by letting the players in the Mitchell Report off the hook. And to take it a step further, look at the owners colluding against Barry Bonds. According to reports, they are prepared to pay whatever fine comes against them for collusion. Cheating will always be a part of baseball, even though the smart people know it’s wrong and that if baseball wanted to be taken seriously, they’d actually attempt to punish cheaters.

It’s the same way with sabermetrics. It’s an unconventional, to many, way to look at baseball. However, with the amount of resources and numbers at our disposal, we have a pretty good idea that the players who are performing on the field are the ones with the best numbers and VORP.

The baseball writers who have grown up with sabermetrics in a majority of their lives are helping bring statistical analysis to the forefront. Yes, there will always be a sector that doesn’t agree with it, but for the most part, the anti-sabermetric views are coming from the veterans of the sportswriting world.

While there are doubtlessly sexier options than waiting, it seems like waiting is the best option. As some nontraditional statistics have worked their way into the mainstream

But what I think doesn’t dawn on most people is that to embrace statistical analysis, one doesn’t have to completely discount the “scout” theory.

“High school numbers really don’t mean much, so it’s the scouts with the best eyes who help teams succeed there. Those same scouts, though, tend to be so against statistics because of an irrational fear that they could be replaced. Are scouts as imperative as they used to be, or as prevalent? No, and with pretty good reason: teams have gotten smarter and understand that scouts tend to be a lot more fallible than a century of history interpreting numbers in a certain manner,” Jeff Passan of Yahoo! Sports said.

There will always be a place for scouts in baseball, it’s just that statistical analysis is now being used to actually help those scouts do their job, yet it’s being viewed as an intruder onto sacred ground.

If more people realize that statistical analysis is here to stay and is a–very important–tool, then I think much of the hate will dissipate. And anyway, that hate that is being spouted, many is from people who admit that they know nothing about it in the first place. Does that make any sense?

Slowly, non-traditional stats are creeping into the mainstream, and as we go, it’s inevitable that more will make their way into common baseball conversation. While it’s unlikely that statistics like VORP will ever appear next to the box scores in the newspaper, it’s reasonable to expect that sabermetrics will become more known. But hopefully with that inevitable growth, comes an embracing of sorts.

Oh yeah, we should Fire Joe Morgan too.


PECOTA and Brian Bannister

May 4, 2008

Along with VORP, Baseball Prospectus has derived a projection system called PECOTA.

“Stands for Player Empirical Comparison and Optimization Test Algorithm. PECOTA is BP’s proprietary system that projects player performance based on comparison with thousands of historical player-seasons. Analyzes similarities with past player-seasons based not only on rate statistics, but also height, weight, age, and many other factors.”

The projections are available on BP’s website and in their annual book before each season.

On the whole, PECOTA projections are pretty accurate, and have correctly forecasted steep dropoffs and upswings for players.

However, there always will be exceptions to the rule. Take Royals pitcher Brian Bannister.

“Their theory is that pitchers can control only three things — strikeouts, walks and home runs allowed. Once a ball is put in play, it will result in a batting average of .303. Since Bannister’s BABIP (batting average against balls hit in play) was the third-lowest in baseball last season at .264, and he had 77 strikeouts and 44 walks in 165 innings, the belief was that Bannister would digress.”

Digress to the point of a 5.19 ERA  and 1.52 WHIP with a .304 BABIP.

In the summary paragraph below BP’s projection, BP wrote that Bannister’s intelligence would allow him to keep from being a one year wonder. Despite having stuff that stat geeks aren’t in love with, Bannister has a mind that a stathead dreams about.

“One thing that I work a lot with, and that is not factored into common statistical analysis, is what counts a pitcher pitches in most often – regardless of what type of “stuff” he has. Most stats only measure results, not the situations in which those results occurred. In the common box score, an RBI is an RBI, but it doesn’t show the count, number of outs, and number of runners on base when it occurred. For me, the area where pitchers have the most opportunity to improve or be better than average is in their count leverage.

Let me give the fans and young pitchers out there one example of a way that I try to improve my performance, this time with regards to BABIP.

Question to myself: Does a hitter have the same BABIP in a 2-1 count that he does in an 0-2, 1-2, or 2-2 count? How does his batting average and OBP/SLG/OPS differ when he has two strikes on him vs zero or one strike?

These are the type of questions that I will come up with and employ in my starts to see if I can improve my outings. For example, here are my career numbers in the counts mentioned above:

2-1: .380 (19/50)
1-2: .196 (20/102)
2-2: .171 (18/105)
0-2: .057 (3/53)

It is obvious that hitters, even at the Major League level, do not perform as well when the count is in the pitcher’s favor, and vice-versa. This is because with two strikes, a hitter HAS to swing at a pitch in the strike zone or he is out, and he must also make a split-second decision on whether a borderline pitch is a strike or not, reducing his ability to put a good swing on the ball.”

So is Bannister an exception to the rule, or a step ahead of even the sabermetricians?

“There are always exceptions to anything. Brian appears to be one in this case. The general rule doesn’t always cover the specific case,” Dutton said.

Statistical analysis is far from a finished product, especially in regards to fielding and pitching analysis, and if Bannister continues to have success, it only furthers that notion.

Scout vs. Stat and Old Guard vs. New Guard

May 4, 2008

Moneyball–the bestselling book by Michael Lewis in 2003–brought the scout vs. stat conflict to the forefront. Billy Beane, the A’s GM, was a proponent of the stat philosophy, using numbers and statistics to exploit inefficiencies in the market created by the scout view, which relied more on the observation of physical tools that a player had.

The schism that the book exposed grew even greater after the book’s release. Those who didn’t like Beane or his philosophies were quick to dismiss the book as propaganda and bullshit. Others thought that the book was a veritable gospel, the blueprint for which franchises should be built around. (It should be noted that I don’t really think that either viewpoint about the book is accurate. To me, the book was more of a portrait of Billy Beane and his philosophies, not a how-to book on how to win using sabermetric statistics)

And as the divide grew, the reluctance to use the tactics of the opposing philosophy grew as well. Baseball people were almost forced to choose sides, and for what reason? We all know that the truth probably lies somewhere in the middle, and with the growing popularity of sabermetrics, the middle ground is finally being discovered.

“I think the line between stats and scouting is starting to blur a little bit, which I think is a good thing,” Derek Carty of Hardball Times said.

Bob Dutton of the Kansas City Star added: ” All stats, including saber stats, are a wonderful tool and offer important insight. But they’re only part of the picture despite what some advocates think. The human element is important, too. Those who don’t have access to players and club officials tend to minimize this. To me, that’s foolish.”

While there has been a growing acceptance, there also have been writers who refuse to admit the significance of statistics, despite the blurring of the lines between scout and stat.

“From my limited – stress limited – experience with some of these Old Guard types, I’ve noticed that they aren’t really open to new ideas.  They are content to do what they have been doing, even when newer ideas are proven to be better.  I think this is the problem some of the more statty guys (I realize this isn’t a word) have with them.  The stat guys are looking for the truth about baseball, regardless of where it comes from, ” Carty said.

“If truth can be found outside of raw numbers, then that is perfectly fine by them.  Sites like the Hardball Times and Baseball Intellect and Saber-Scouting, for instance, have no problem using a blend of stats and scouting.”

Basically, those refusing to change are the old guard.

“I think in very general terms, the ones who have a problem with statheads are the old guard, the ones who just don’t understand it, have gotten along for decades covering the game just fine without, don’t need it, no thanks, I don’t understand it so it sucks. It’s probably just like anything else. How many older people do you know who text message? I don’t know any. I know older people who make fun of me for text messaging, but I think that’s because they don’t understand how to do it, they can communicate just fine with their phones without it, don’t need it, no thanks, I don’t understand it so it sucks,” Sam Mellinger of the Kansas City Star said.

And sometimes, those people who say it sucks, have no idea why it sucks in the first place, except for the fact that it’s unfamiliar.

“Resistance to change is certainly part of it. Traditions run deep in sports, and in none deeper than baseball. I think journalists are — consciously or not — predisposed to uphold those traditions. In this case, we’re talking about the ridiculous traditions that say that batting average is the most important offensive stat, and wins and ERA are how you know how good your pitchers are.

But really, I think the main problem is ignorance. I think most journalists who take the time to familiarize themselves with sabermetrics realize that there’s at least some value in them. The most vehement anti-sabermetric arguments are made by the very same journalists who pride themselves on not having spent any time trying to find out just what exactly sabermetrics is, or what some of the new statistics are. (See Murray Chass, among others.), ” dak from firejoemorgan.com said.

And while thanks to the new guard and sites like Baseball Prospectus and Hardball Times, stats like OBP are creeping their way into the mainstream, that assimilation is going to take some time. While the acceptance of sabermetric statistics will never be 100%, there should eventually come a day that batting average is rendered irrelevant.

“Five, ten years ago I would have said: no way. But there are already signs that the MSM has begun to accept some of the basics of sabermetrics. Players’ OBP has finally found its way onto scoreboards at ballparks, and onto local and national telecasts. Twenty years ago, that would have seemed absurd. I think it’s a small step in the right direction, and I think we’ll see more steps like that in the future. At least I hope so.

Even ESPN employs sabermetricians: Keith Law, Rob Neyer. If nothing else, there’s a growing market for that kind of stuff. And since baseball (and the covering of baseball) is, after all, a business, I think we’re only bound to see more sabermetric principals find their way into the MSM as time goes on,” dak said.


May 3, 2008

Yes, VORP sounds like a Star Trek phrase. I don’t think that anyone is arguing that.

Value Over Replacement Player attempts to do just what it name leads you to believe. Find out just how much better a player is at his position over the major league average. Seems pretty logical and easy, right?

However, probably in some part due to the funny sound that it’s anacronym makes, VORP has come under fire by those who don’t like, or don’t want to like sabermetrics. Jon Heyman, a senior baseball writer for SI.com, even went so far as to create the term “VORPies” when declaring that Jimmy Rollins was the best choice for National League MVP.

Yes, you read that right. VORPies.

“Even so, I wasn’t shocked that stats people have taken issue with Rollins winning the MVP award. There are numbers crunchers out there — including a firejoemorgan.com author who wrote a guest piece in Sports Illustrated last week — who believe baseball writers rank somewhere between morons and idiots for voting Rollins as MVP over David Wright, who had a higher VORP. The stat people seem to believe VORP — a Baseball Prospectus statistic that stands for Value Over Replacement Player — defines a player, but why haven’t many of them championed last year’s VORP leader (Hanley Ramirez) as MVP instead?”

Rollins: 30 HR 94 RBI 41 SB .341 OBP .514 SLG

Wright: 30 HR 107 RBI 34 SB .416 OBP .574 SLG

Hmmm… Maybe those dreaded VORPies are on to something. Wright was tied or better in every category except stolen bases.

But wait, you say, Heyman said that Hanley Ramirez had the highest VORP in the league. Yeah, he did. But what Heyman ignored–he’s a smart man, so I give him the benefit of the doubt–was that little line in the Baseball Prospectus definition of VORP: “VORP scores do not consider the quality of a player’s defense.”

While defensive metrics (more on that later) still have a bit of a ways to go, we’re going to use FRAR for our comparison.

FRAR: “Fielding runs above replacement. A fielding statistic, where a replacement player is meant to be approximately equal to the lowest-ranking player at that position, fielding wise, in the majors. Average players at different positions have different FRAR values, which depend on the defensive value of the position; an average shortstop has more FRAR than an average left fielder.”

So Ramirez and Rollins will have an inherit advantage.

Wright posted a score of 31 in 2007, saving 31 more runs than a replacement level 3B. Hanley Ramirez was -2, which means that he gave up 2 more runs than a replacement level shortstop. Not good.

Jimmy Rollins was at 27, still less than David Wright.

Many people blamed the Mets’ late season collapse on Wright, which is completely inaccurate. In September and October, Wright posted an OBP of .432 with 6 HR and 20 RBI. The Mets’ struggles were far from Wright’s fault.

By the way, the MVP award was voted on by the baseball writers.

What are sabermetrics?

May 2, 2008

Sabermetrics is the analysis of information based on statistics, the goal of which is to objectively determine who is the best or worst in a multitude of categories. According to Bill James, the forefather of sabemetrics—and sadly, a Kansas graduate—sabermetrics is “the search for objective knowledge about baseball.”

That search usually involves computers, as the formulas for many sabermetric statistics are too complicated and lengthy to do by hand. This use of computers has frustrated “the old guard” of baseball writers and observers, who feel that they can determine themselves who and what is good and what isn’t good, numbers be damned.

The most commonly used sabermetric statistic is OBP, which is the percentage of times a batter gets on base, including base hits.

OBP takes the place, and is much better than batting average, which is probably the most commonly used statistic. Batting average only takes into account the amount of times a hitter gets on base due to a hit, and doesn’t include walks or hit by pitches. A walk is just as good as a single more than 90% of the time, yet batting average doesn’t even take walks into consideration.

Adam Dunn, a Cincinnati Reds slugger, has a career batting average of .248, and is criticized by many for striking out too much. While Dunn does have a prolific k-rate, he is infinitely more valuable than his batting average would lead you to believe.

Dunn has a career OBP of .381, which means he takes a walk or gets hit by a pitch in more than 13% of his plate appearances, a walk rate that puts him among the elite.

The next entries will explain how many media members are uneducated and ignorant about sabermetrics and how the media could better utilize them to change common perceptions that they have created. While sabermetrics are becoming more common and part of the common baseball vernacular, they still have a long way to go, thanks in no large part to the media members who despise them. By further educating the media, we can also further educate the general public, and when has a more educated general public ever been a bad thing?


May 1, 2008

BABIP: Batting Average on Balls in Play. Measure of batted balls that fall safely into play. Formula is (H-HR)/(AB-K-HR+SF)

BR: Base Runs, which quantifies the number of runs contributed by a batter. The fundamental formula is (baserunners*scoring rate) + home runs

DER: Defense Efficiency Ratio. The percent of times a batted ball is turned into an out by the team’s fielders not including home runs. The exact formula is (Batters faced-H-K-BB-HBP-0.6*E)/(BFP-HR-K-BB-HBP). Similar to BABIP, but from the defensive team’s perspective.

ERA+: ERA measured against the league average and adjusted for ballpark factors. An ERA+ over 100 is better than average, less than 100 is below average.

FIP: Fielding Independent Pitching. A measure of all those things for which a pitcher is responsible. The formula is HR*13+(BB+HBP)*3-K*2)/IP, plus a league specific factor, usually around 3.2, to round out the number to an equivalent ERA number. FIP is used to help tell how well a pitcher pitched regardless of how his fielders fielded.

ISO: Isolated power, measuring the “true power” of a batter. SLG-BA

LD%: Line Drive Percentage. Baseball info solutions tracks the trajectory of each batted ball and categorizes it as a ground ball, fly ball or line drive. LD% is the percent of batted balls that are line drives. Not necessarily the hardest hit balls but they fall in for hits about 75% of the time.

OBP: Proportion of plate appearances in which a batter reached base successfully, including hits, walks and hit by pitches.

OPS: On Base Plus Slugging Percentage

OPS+: OPS measured against the league average and adjusted for ballpark factors. OPS+ over 100 is better than average; under 100 is below average

Pythagorean Formula: A formula for converting a team’s run differential into a projected win-loss record. The formula is RS^2/(RS^2+RA^2). Teams’ actual win loss records tend to mirror their pythagorean records, and variances can usually be attributed to luck.

RC: Runs Created. Invented by Bill James, RC is a measure of the number of runs a batter truly contributed to his team’s offense. Basic formula is OBP*TB, but can be measured to include hitting well with runners in scoring position and adjusting for ballpark impact.