Sunday, January 09, 2005

Baseball America Roundtable: Stats vs. Scouts: The Great Debate 

Baseball America hosts a roundtable discussion between two "statheads" -- Gary Huckabay and Voros McCracken -- and two scouts -- Eddie Bane and Gary Hughes.

There's some interesting stuff, and a couple of the guys come off as being fairly unpleasant folks to deal with, but it is a good read.

Going through it, though, got me thinking about the hostility that seems to exist between the two camps, the way it has become a form of trench warfare with both sides having dug in and become dead set on holding their ground no matter what.

One of the most unfortunate elements is that, as sabermetric principles have started to become more integrated in baseball, a lot of the "statheads" are still stuck in a bomb-throwing, outsider mode. It is like when an opposition party finally starts making gains and getting power after years of being on the outside...the communication skills and political aspects of trying to get things done, rather than just sit on the sidelines and criticize, don't always translate.

Some of the folks at BP coined the phrase "beer and tacos" when discussing the scouts vs. stats debate. They say, why do you have to choose between stats and scouts? Why do you have to have either beer or tacos? Why not both, particularly since they go so well together? The problem is, there are those in the sabermetric community for whom, it seems, ridiculing the traditionalists has become a way of life, part of their schtick, and working towards a compromise doesn't fit in their worldview. I've heard statheads dismiss anything gleaned from watching a player perform as "visual scouting", an archaic form of data-gathering that simply injects unreliable subjective information into what should be an objective process, needlessly confusing the evaluation process. Rob Neyer, after the 2004 draft, wrote a column sneering at the capsule reports filed by scouts on various amateur prospects.

But there are weaknesses in players that cannot be gleaned through the statistics alone, that require personal observation to discover. A hitter's swing may be too long, making him vulnerable to hard-throwing major leaguers and leading to ridiculous strikeout rates and lower batting and slugging averages on the major league level -- think Jack Cust or Carlos Pena. A pitcher may be dominating at high-A because he's polished and has great control, but doesn't have the stuff to get batters out at the major league level consistently -- think Mario Ramos or John Stephens. Fielding, in particular, is difficult to judge using statistical methods right now, leaving it particularly vulnerable.

There's nothing for us "statheads" to be ashamed of about that. The stats may identify a player out there that is being overlooked, but that the scouts think is flawed or can't make it because of physical limitations. It doesn't mean that a prospect should be dismissed out of hand, but the scouting reports should be taken into account when evaluating the prospect. Similarly, a player may scout extremely well, but have K/BB rates or poor isolated power or other statistical keys that suggests that he won't pan out. One has to synthesize that data, incorporate both aspects in order to make a proper evaluation. It would be foolish to ignore the subjective evidence solely in favor of the objective evidence, and reveals exactly the type of ingrained biases that Bill James and Craig Wright and others at the vanguard of the sabermetric movement fought against when they took on the traditional baseball types.

The whole point of the sabermetric movement was to get more data, better data, more comprehensive information to incorporate in your evaluation and decision making process when looking at players and teams. It was about understanding, the quest for knowledge and truth.

Unfortunately, it seems that some have turned it into a pissing match, a holy war to show that they are right at the traditionalists are wrong.

I want to see prospects play. When a pitcher gets called up by the Rangers, I want to make sure I see them pitch on TV as soon as possible. I'll Tivo the game if I'm not going to be home. If the game isn't on in Houston, where I live, I'll ask those who watched for a scouting report. When folks I know are going to see a minor league game, or are going to spring training, I want them to bring back a report on what they observed.

I remember seeing Hank Blalock when he was at AA, getting goosebumps seeing him at the plate, watching how damn professional he looked up there, how, even though he was the youngest player on the field, he carried himself like he was a ten year veteran. I remember coming away even more impressed with him than before I saw him.

But then, why was I watching for Blalock in particular anyway? Because he was putting up stats that were out of this world, when he was just 20 years old. I wanted to see him to verify subjectively the stats -- the objective evidence -- that I had already received. If I didn't know anything about him -- or if he was 26 and putting up a .230/.280/.350 line -- would I have even noticed him? I don't know...

I do think that there is a tendency, even by those who are aware of the trap, to overemphasize what you see, particularly with first impressions. I know fans came back from Surprise this spring raving about Joaquin Arias. Jamey Newberg has him ranked as one of the top two or three prospects in the organization, based, from what I understand, in large part on what he saw from him this spring. And Jamey is not alone...Buck loves Arias, scouts love Arias, the folks who see him love him.

And yet, I look at the numbers, and see a guy with limited power potential, who doesn't walk, and who has great defensive skills but has yet to translate that into great defensive ability. I see a guy who, if everything pans out, is a quality defensive shortstop who will steal bases and hit for average, but won't have the OBP or power to be a premium offensive player. And I also see a guy who, at 19, still has a long way to go, still has a lot of developing to do and a lot of hurdles to overcome before he can even reach that potential. I see a guy who may be getting overvalued by the traditionalists because his skills at such a young age are so exciting, but have only so far that they will be able to go.

And just like the statheads need to have the scouts to point out that the guy who is dominating AA with the 87 mph fastball, the decent but not great breaking ball and change, and the superior control looks to be way too hittable to be a top-of-the-rotation starter in the majors, the scouts need the stathead types to point out that the toolsy outfielder with the great speed and tremendous batting practice power isn't showing the plate discipline or the power development that he needs to be an elite prospect. Done properly, it provides a form of checks and balances, with each side tempering the biases of the other, acting as the devil's advocate to make sure that each side of the situation is being examined.

One of the things in Moneyball I thought was so fascinating was Billy Beane's refusal to go see Nick Swisher, the college outfielder he absolutely loved, and wanted to take with the A's first first-round pick. He wouldn't go because he didn't want to tip off other people that he was on Swisher that hard...but at the same time, he was desperate for the scouty information on him...for the subjective evidence that so many in the stathead crowd want to dismiss. And the two top prospects at this point from the infamous Moneyball draft -- Swisher and pitcher Joe Blanton -- are guys who were coveted by both the stathead and scouting crowds. Beane, the Michael Corleone of the stathead crowd (with Bill James, of course, being Vito Corleone), wanted to have the subjective information along with the objective information, and if he couldn't see Swisher himself, he at least wanted his lieutenants to see him and bring him back the news from the front.

Going forward, the most successful organizations are going to be those that can synthesize both the stats and the scouts. One of the reasons I am such a big Grady Fuson fan is that I feel he's one of the best at combining the two. A thriving organization is going to have the Ian Kinslers and Kameron Loes, but also have the Juan Senreisos and the Joaquin Ariases. A successful organization has to be able to utilize information from both branches to determine who has the best chance for success, who is being overvalued or undervalued by your competitors and potential trade partners. And scouts and statheads alike need to view each other as complimentary pieces of the same puzzle, rather than competitors in a zero-sum game.

Nice work, Adam. Have you shared it with anyone at BA? -- Jamey
Thanks Jamey...

I haven't shown it to anyone, other than the delightful folks who visit the blog here...
Nice read Meno ... can't we all just get along? yuck yuck

I agree with Jamey that the 'beer and tacos' post belongs on a bigger stage. I don't know if there are any NEW ideas there, just smoothly-said ideas for thick folks like moi.

Last paragraph - 'success organizations' should be 'successful organizations' ... I think. Get it edited if you go big time with it.

I totally agree with your take on Arias - great example. As somebody who watched his stats/recaps and listened to lots of his games last summer, my impression is worse than others who've actually seen him and others who actually PAY him. I am certainly encouraged by Jamey and Buck's gushing, but I hope their faith translates to the field this summer as offensive and defensive stats worthy of the hype.

BTW, I'm looking forward to listening to more minorLBB this summer over the web. I figure I like it for the very reasons you state in your writing -> I like to attach the things I hear and the things the announcers relay to my pile of boxscores. Anecdotes, nagging injuries, asides, pre/post-game interviews - it's all part of the puzzle I enjoy piecing together.

Another BTW, a little advanced-scouting at the Bakersfield Blaze official website [high A] calmed my worries about getting to listen to their games over the internet this season - their release ...

The Bakersfield Blaze of the California League are proud to announce that KGEO 1230 AM will broadcast the entire 2005 season, the first time since 2003 the Blaze will be heard on the radio. Blaze Assistant General Manager Brian Thomas will handle the play by play for all 140 games.

All 140 games will also be broadcast Live! over the internet this upcoming year. You can check back to this page for more information and instruction on how to link to the broadcast, as the season approaches.

always lurking,
Post a Comment

This page is powered by Blogger. Isn't yours?