The Battle of Dueling Baseball Philosophies
By Mike Silva ~ August 10th, 2009. Filed under: Mike Silva.
On Friday James K did a great follow up piece on Amazin Avenue regarding my comments about WAR or Wins Above Replacement. First, I was flattered that NYBD has come far enough where I would be the topic of conversation on a site like Amazin Avenue. In my humble opinion, AA is right there with the top Mets blogs on the net. With that being said I felt I needed to clarify my position, provide perspective to the “far stat dependent” audience, and give my thoughts on WAR.
My issues with WAR are not because of an aversion to statistics. I rely more on traditional baseball methods to determine a players value (runs, hits, slugging, OBP, batting average, WHIP, ERA, K to BB ratio, errors, etc.). However, to quote the great Vin Scully, “Statistics are used much like a drunk uses a lamp post: for support, not illumination.” I need to see a player in order to get the entire package. I like to know what people in the game think of him as a person, in the clubhouse, and if he has a reputation of performing in big spots. Also, how will he handle failure? Those are not “talk radio” concepts, but tried- and- true aspects of the game of baseball.
WAR is a perfect example of assessing a player in a vacuum. It’s a very dangerous method of evaluation. I spoke to a baseball person on Friday and he told me, to his knowledge, this stuff isn’t even used by teams in arbitration. That should tell you something right there. James even sent me this Sports Illustrated article back in April which cites teams are still very leery of many modern statistical theories. St. Louis was one of the teams. I don’t think it hurt them in 2006.
My first problem with WAR is crediting a “win to an individual player”. All sports are won and lost as a team. You can’t tell me Ryan Church is worth two more wins over Jeff Francoeur, because we just don’t know how games would play out if Church stayed in New York and Francoeur in Atlanta.
Next, WAR cites the “average minor league player” as its benchmark. Who is the typical average minor leaguer? Anderson Hernandez? Nelson Figueroa? Sergio Mitre? Josh Towers? Fernando Martinez? Shelley Duncan? Do you Get the point? Average to you and me can mean all different things. We are probably in the same ballpark, but I value a Nelson Figueroa much more than Josh Towers. I also think the gap between Figueroa and Mike Pelfrey is smaller than, lets say, Josh Towers. The fact that defense, a subjective measure to being with, is weighted also gives me pause. The lack of respect given to the position of first base tells you how flawed the stat really is. Ask the greatest infield ever if John Olerud was overrated at first versus Todd Zeile the following year. What about Keith Hernandez and his impact for the late eighties Mets. Do you think the Yankees are better off with Teixeira’s defense at first? I believe excellent defensive first baseman makes everyone better by saving throws and cutting down base hits.
The biggest value WAR may have is assessing the opportunity cost from a financial perspective when filling out a roster. Would Anderson Hernandez do just as good a job as Alex Cora next year for 1.5 million dollars less? It’s ok to take a look at it, but one would be foolish to use it as a guiding concept.
There is nothing wrong with using modern statistics like WAR. A good general manager should have a department that focuses on sabermetrics. To not consider all angles of baseball research is unacceptable in this day and age. I believe the sabermetric crowd deserves a seat at the table, but a very small say in the final decision. None of those numbers account for makeup, character, and fortitude. Julio Franco may not have been a valuable asset to the 2006 Mets on paper, but early on he helped Carlos Beltran overcome his aversion to the harsh New York fans. He might have been worth the contract just for that. Alex Cora has been cited by Jerry Manuel as one of the clubhouse leaders. Let him walk for A-Hern and you may lose in a way numbers can’t measure. These are just two of many examples I can cite.
In the end I believe this is all great debate. Everyone has a point and a valuable piece of information that helps achieve the ultimate goal: a winning ballclub. It’s almost become like American politics where both sides draw a line firmly in the sand and won’t consider some valid points by the other. It amazes me how the old school baseball philosophies completely reject any aspect of modern statistical analysis. It’s equally, if not more, laughable that fans run some fancy numbers and dismiss what those who played and managed the game say. A good company will not let finance, accounting, or sales run their total operation. Each has dueling objectives and philosophies. They are also biased to their principles. A good executive will have a diverse staff and be smart enough to assess all facts and make a decision. Even WAR, which some take as gospel, is predicated on a manmade criteria that can be challenged.
The point of my show and site is to have all aspects of discussion open to the New York baseball fan. My goal is a balanced approach to baseball coverage. Do I have my personal biases and philosophies? Sure. I am going to probably connect better with old school methods than modern sabermetrics, but it doesn’t mean I am not giving them credence. Give James and AA credit as their WAR analysis came within one game of predicting the Mets 2008 record. I like to put everything in the proper perspective. That’s my nature and I believe the right way to manage any thought process. Perhaps my biggest learning from this is that NYBD needs to have more dialogue about the topic. Healthy debate will continue to grow new media and achieve a large and diverse audience. Who knows, maybe NYBD one day will go down as the great unifier of dueling philosophies. Now that truly would be ironic.
August 10th, 2009 at 10:20 am
Mike, it appears you totally missed the point of that SI piece. Major league teams are not leery of the defensive metrics, they are increasingly using them because they are reliable. It’s the players (like Torii Hunter) who are leery. But who cares what the players think. Just read this from the piece, which quotes a baseball insider, something you seem to put a lot of weight behind:
“There are still teams stuck in the Dark Ages,” says one American League general manager, “but the secret’s getting out. Defensive metrics have almost caught up to the offensive side. Some people would say they didn’t think they’d see this day. But the revolution’s here.”
August 10th, 2009 at 11:47 am
The piece is very clear that the saber crowd is trying to get on the payroll and nobody is sold on doing that just quite yet. Maybe one day we’ll all become robotics and live life through calculating probability. Baseball isn’t a simulation game, defense improves when a player works hard at improving his weakness’ and increase the practice reps. You guys can stay in the lab to try and create the perfect ballplayer, but don’t expect any sort of paycheck coming for that. We have enough agendas in this world leave it out of sports please.
August 10th, 2009 at 1:15 pm
James
I think it would be interesting for a team to come out and try these models for one season. If you could have that type of benchmark perhaps MLB will mainstream the idea. Using it behind closed doors, where it’s hard to ascertain the level of engagement.
Baseball is very closed minded and modern metrics need more of a benchmark of success than Pythagorean records. Obviously, UZR and other defensive metrics are starting to catch on. I remember the Mets citing a zone rating of some sort back in 04′ when they signed Mike Cameron. Obviously, it didn’t take into account how age caught up with him because I don’t remember being impressed with him in centerfield. He actually played a better RF, IMO, than CF during his time in NY. Wonder if metrics confirms that.
August 10th, 2009 at 1:46 pm
actually WAR cites the average replacement player, aka bench guy. You know, the guy who spends all his time on the bench, and plays when someone goes down, and hits .220/.300/.350 (Alex Cora)
What baseball person did you speak to?How can one baseball person make something true?
August 10th, 2009 at 1:48 pm
Also “character, clubhouse leadership…” Great. Why don’t we get rid of Carlos Beltran (no clubhouse leadership) and replace him with Darin Erstad. Excellent. We’ll gain 10 games in the standing because of him
August 10th, 2009 at 2:02 pm
Now WAR is a bench guy and not the “average AAA player”. This theory is swiss cheese.
August 10th, 2009 at 2:36 pm
listen Chris Silva, WAR is both. But why don’t I quote from http://www.fangraphs.com/blogs/index.php/win-values-explained-part-four
“we prefer to value players above replacement, as that gives us a fixed baseline of $400,000 in salary – the league minimum.” and later in the article: “These guys have fallen into that Four-A category, where they show more ability than your average Triple-A veteran but not enough to hold down a major league job. They’re usually available every winter as minor league free agents, via the Rule 5 draft, or as cheap trade acquisitions where a team can acquire one of these players without giving up any real talent in return.”
do you get it now, or is it too big for you brain?
August 10th, 2009 at 2:55 pm
None of this is too big for my brain. The math is math. Its the theory behind it I don’t buy. The fact that this topic needs some much intrepretation proves its an agenda more than anything. I find it to be very speculative and leaves too much room for intrepretation. Bottom line, these numbers can be made to say whatever they want. Stats like batting average, RBI’s are very clear cut. Number of hits divided by number of at bats. Simple. I applaud the commitment to the idea, but its sounding more cultish than factual.
August 10th, 2009 at 3:02 pm
yes batting average is amazingly clear cut. The fact that someone hits .300 and doesn’t walk or hit for power makes them better then someone who hits .260 walks 100 times and hits 40 homers. Yes Batting average is not speculative at all. You know the .300 hitter is better than the .260 hitter.
August 10th, 2009 at 3:31 pm
I didn’t infer that at all. The difference between .260 and .300 is about 18-20 hits. It’s a barometer to measure the level of consistency of a hitter. Each player is valued based on the complete package that is brought to their position and team. Stats are a part and their overall impact sets the difference.
August 10th, 2009 at 4:21 pm
Chris -
WAR is a formula, just like Batting Average.
Just because the WAR formula is more difficult to understand, does not meant it is invalid.
August 10th, 2009 at 5:16 pm
Dan,
This is not that difficult of a formula to understand. My issue with it is that you guys are weighting its value far too much. You can put people and the game into as many computer simulations as you want. It’s still played on the field and the performance out there is the measure of it. You’ll never have a successful team by using formulas to come up with the perfect statiscally mix to win. Are we really nieve enough to think an organization uses theses formulas and make decisions on which combination produces the highest number of wins? The only tanglible defense I’ve heard from the community is “its too difficult to understand.” This is said but everybody who receive some sort of resistance to the idea. Proves the illagimaticy of the concept. You ever hear the phrase “Don’t believe the hype?” This is exactly what this is a whole bunch of hype. If you say something long enough and loud enough you’ll get people to listen no matter how crazy the idea.
August 10th, 2009 at 5:55 pm
Uh, Chris
see Beane, Billy, and Athletics, Oakland. Using statistical analysis, they won 4 division titles and a wild card in a 10 year period. They went to the playoffs in that period only second to only to the Yankees. They might not have won a world series (Boston did, using the same analysis, they had Bill James) but they did a lot better than my favorite team, and one you guys write about, the Mets.
August 10th, 2009 at 7:34 pm
There was never an arguement that statiscal analysis is no factor at all. I’ve said it is a factor but shouldn’t be the main source of decision making. I agree with Beane’s philopsphy on college players vs high school players. College players spend 3 to 4 years playing a mini version of “minor league ball.” Also, I think its more beneficial because players learn how to deal with being away from home and not always being the best player on the field. I think a lot of high school players fail because of those specific factors. However, its hard to use a high school or college players stats as a major indicator. The aluminum bats play a big part in the inflated numbers and their schedules aren’t as big so not as long of a season. Also, I think Beane’s stays away from high school players for the exact reason I mentioned above that many don’t develop because they can’t handle the difficulty of the next level. I looked at the players that came up through their system and the number of guys who made it vs didn’t seemed the same, possibley a few players higher. But, nothing that alarming that stands out drastically above the rest. In 2002, the A’s had a good team but seemed to have trouble playing in Yankees Stadium. Plus they had attrocious middle inning relief and were a bat short with Giambi’s exit to the Yankees. They had a great year because Anaheim (World Series Winner that year) won 99 games and Seattle won 93 games and missed the playoffs. They did play in a tough division. From the roster, I don’t see anybody that out to me that was a major nobody and turned to a star. Tejada suspect due to the steriod admission. Most of the players on the roster had yet to develop yet and did so with other teams down the line. Jermaine Dye played fairly well in Oakland but it seems like it all came together in Chicago the last few years. A bit more protection in the lineup and usually players hit a point where they understand the game and their game better. IN 2002, the strentgh of that team was the had a front 3 of Zito, Hudson, and Mulder who all logged over 200 IP’s and from their records had good run support. Their middle relief was horrendous and Koch saved 44 games but had 11 wins as a closer. Very suspect stat because it seems a good offensive team bailed him out of some blown saves.
Playing in Oakland keeps a lot of pressure off the players because of the small market and games being on too late back east. This is also before the MLB package so most of us on the east coast only watched these teams when they played our team. I think there’s something to be said for a general consenus of little expectations. In the grandier scheme, Oakland isn’t NY or Boston where head’s role with anything less than a World Series Ring. Success in these markets is being competitive and making the playoffs if your lucky.
This is an example of the Beane’s plan for the 2002 draft vs who they actually drafted. Keep in mind that the top off is who Billy would’ve drafted in a perfect world where he could pick every single player he wanted without another team drafting that player
Pitchers
Jeremy Guthrie – Cleveland, #22 (1st round)
Joe Blanton – Oakland, #24 (1st round)
Jeff Francis – Colorado, #9 (1st round)
Luke Hagerty – Chicago Cubs, #32 (1st round)
Ben Fritz – Oakland, #30 (1st round)
Robert Brownlie – Chicago Cubs, #21 (1st round)
Stephen Obenchain – Oakland, #37 (1st round)
Bill Murphy – Oakland, #98 (3rd round)
Hitters
Nick Swisher – Oakland, #18 (1st round)
Russ Adams – Toronto, #14 (1st round)
Khalil Greene – San Diego, #13 (1st round)
John McCurdy – Oakland, #26 (1st round)
Mark Teahen – Oakland, #39 (1st round)
Jeremy Brown – Oakland, #35 (1st round)
Steve Stanley – Oakland, #67 (2nd round)
John Baker – Oakland, #128 (4th round)
Mark Kiger – Oakland, #158 (5th round)
Brian Stavisky – Oakland, #188 (6th round)
Shaun Larkin – Cleveland, #274 (9th round)
Brant Colamarino – Oakland, #218 (7th round)
[edit] Oakland’s picks
#18 – Nick Swisher – successful major leaguer, traded to Chicago White Sox after 2007
#24 – Joe Blanton – successful major leaguer, traded to Philadelphia Phillies
#26 – John McCurdy – out of professional baseball since 2006
#30 – Ben Fritz – had Tommy John surgery, currently in minor leagues
#35 – Jeremy Brown – has had small stints of time in the majors, retired prior to the 2008 season.
#37 – Stephen Obenchain – retired after a lack of success at the A and AA levels
#39 – Mark Teahen – successful major leaguer, traded to the Kansas City Royals and is the second baseman for the Royals
#67 – Steve Stanley – topped out in AA for the A’s
#98 – Bill Murphy – traded three times and debuted with the Arizona Diamondbacks in 2007
#128 – John Baker – traded to the Florida Marlins and debuted and became the starting catcher in 2008.
#158 – Mark Kiger – traded several times, now with the AA affiliate of the New York Mets (Binghamton Mets)
#188 – Brian Stavisky – currently with the AA affiliate of the Los Angeles Angels
#218 – Brant Colamarino – currently with the AA affiliate of the A’s
If you look at the list, a couple of nice mid range players in Swisher, Baker, Teahen, and Blanton. But, nothing jumps out here to me of this method uncover a star. Nice players that will help you win a title but you wouldn’t want any of these guys being the focal point of your offense or pitching.
August 10th, 2009 at 9:32 pm
right but they did get 4/13 draft picks to become successful, about 33%. Which is what you hope for in a draft
August 11th, 2009 at 3:33 am
Mike,
You’re half right about your inference that WAR is a be all and end all statistic. It’s not. No statistic is. It is far from perfect, and any sabermetrician would be inclined to agree with you. However, it is a very valuable stat. To my knowledge, it’s the only stat that encompasses all the facets of the game (hitting, fielding and baserunning) and does it with a high confidence rate (standard deviation). WAR just documents what happens on the field, which is a matter of record. It attempts to neutralize the effect of the team, and focuses solely on the player in question. This much you know, I’m sure. With a straight face, there is no way you would be able to tell me that traditional stats such as RBI, BA, Fielding % and SB are reliable metrics and tell the complete story of a player. For instance, a guy hitting 100 RBI for the Phillies is a lot less impressive than if someone did it on the Pirates. Sure, hitting .300 is great. But you’re OBP is .325 and your SLG is .350 you’re not providing that much value. Aaron Miles is a good example of that. Fielding percentage just counts the errors and divides that by chances. We all know how subjective errors are, and it doesn’t even factor in range! Last year, Ryan Theriot stole 22 bases last year, yet he was thrown out 15 times. While he succeeded 7 more time than he failed, that’s 15 outs that he made trying to steal a base. If it helps you to think of it this way: 15 outs is the difference between a .260 and a .300 hitter as you posted above.
By now it’s quite obvious you’re eager to respond to what I just wrote, but listen to this first. The above traditional stats I cited are good. They say something. But compared to some of the sabermetrical stats it just sounds like gibberish. Sabermetrics, and WAR in general, try to quantify individual player achievement. When evaluating a player using traditional metrics, it’s difficult to weigh its importance. It all ends up being an arbitrary judgment.
Lastly, baseball players are being paid millions of dollars. It would be foolish to not invest in the most objective ways to value a player. Oh, and sabermetrics do care about the intangibles. They just don’t quantify it because once you do so, they lose their intangible status. Why invoke a supernatural cause when a material one would suffice.
August 11th, 2009 at 2:30 pm
TJ,
Very well said and I think this is the best arguement anybody has come on to make. Please feel free to share your knowledge with the community at any time. Seems like you one of the few that understood our point on this matter.