On Sabermetric Rhetoric

Dear FanGraphs community,

This isn’t a post about baseball, per se, but rather about the way we talk about it. Lately, I’ve been thinking a lot about how to improve the quality of dialogue surrounding sabermetrics. Please excuse my rambling, as I tend to get rather emotional and philosophical when discussing this particular topic.

When reading posts and especially comments, I sometimes get the sense that we think we are right merely due to the fact that statistics are objective. In a sense, this is true. As long as the methodology is clearly laid out, stats really are just numbers. But people are biased. All language is persuasive in some sense, and the inherent neutrality of numbers is often hijacked by various human agendas. Sabermetrics are not exempt from this phenomenon.

Most modern discourse surrounding baseball analysis pits “old-school” vs. “new-school” in a largely arbitrary ideological cage fight. These sorts of polemical constructs make for good television, but slow progress. Its easy to get caught up in the excitement of a debate while completely missing out on what really matters. Baseball is a beautiful game and it brings people together. It’s America’s pastime for a reason! It transcends cultural differences, generation gaps, and even language itself.

Statistics help us to understand and evaluate how well this great game is being played. They act as a mental “handle” by which we can intellectually grasp the importance of each individual event and performance. Everyone, regardless of their stance on sabermetrics, wants statistics that are both intuitive and accurate. So let’s set aside our agendas for a minute and think about how to proactively bridge the gap between these two sides that have so much to offer!

For starters, we should minimize our implementation of hostile methodologies. Getting on a soapbox and proclaiming the evils of traditionalism simply doesn’t do anybody any good. It feeds our pride, as well as the opposition’s presumption that we care more about our statistics than we do about, you know, actual baseball. Over the last few years, I’ve begun to think of myself more as a teacher of sabermetrics than a defender of them. This approach has two important ramifications.

First, it dictates that we get along with those who disagree with us. In my experience, people are only open to new information in the context of a trusting relationship. As fellow baseball fanatics, we have an easy point of contact with traditionalists: we both like baseball. Duh! Focus on that first rather than stuffing a lecture on DIPS theory down their throats.

Second, a teaching disposition encourages us to refine and adapt our communication of sabermetric concepts. Next time you want to call someone a nincompoop on a message board, first ask yourself, “What could I have done to explain this idea more clearly.” Chances are, the person isn’t stupid, just unenlightened and/or overly argumentative. Over my next few posts, I’ll get into the nitty-gritty of how we might make this happen.

Contrary to popular belief, numbers aren’t evil. Baseball statistics in particular have come a long way toward being less deceptive. Let’s represent them well, shall we?

Sincerely yours,

KK-Swizzle



Print This Post

I am a medical student from Traverse City, MI. Interests outside of baseball analytics and medicine: distance running, disc golf, Christian theology, strategy/board games, all things Fire Emblem


Sort by:   newest | oldest | most voted
Anon
Guest
Anon

I sometimes get the sense that we think we are right merely due to the fact that statistics are objective.

Hardly. Statistics have human bias in their structure, and raw counting stats have bais in their usage. AB excludes sacrifices and walks, hits exclude reached on error, ERA excludes unearned runs, FIP excludes all non-HR contact, etc.

I agree with the other ideas, which are good life lessons. Being respectful and building relationships will make others more open to your ideas.

KK-Swizzle
Guest
KK-Swizzle

Did you read the rest of the paragraph? I think we are on the same page, but read it again and let me know if there actually is something we disagree on.

Alec Denton
Member
Member

Agreed. My (probably oversimplified) view has been that there isn’t a genuine conflict between these two camps at all:

http://aldland.wordpress.com/2014/02/13/baseball-notes-the-crux-of-the-statistical-biscuit/

Alec Denton
Member
Member

Ok, that didn’t post correctly. Revised:

“Most modern discourse surrounding baseball analysis pits ‘old-school’ vs. ‘new-school’ in a largely arbitrary ideological cage fight. These sorts of polemical constructs make for good television, but slow progress.”

Agreed. My (probably oversimplified) view has been that there isn’t a genuine conflict between these two camps at all:

“The reason this ‘debate’– the ‘eye test,’ wins, and batting average versus WAR et al.– isn’t really a debate is because the two sides have different descriptive goals. In short, the traditional group is concerned with what has happened, while the sabermetric group is concerned with what will happen. The former statically tallies the game’s basic value points, while the latter is out to better understand the past in order to predict the future. The basic stats on the back of a player’s baseball card aim to tell you what he did in prior seasons; the advanced statistics on Fangraphs, Baseball-Reference, or in Baseball Prospectus aim to tell you something about what he’ll do next year based on a deeper understanding of what he did in prior seasons.”

http://aldland.wordpress.com/2014/02/13/baseball-notes-the-crux-of-the-statistical-biscuit/

Paul
Guest
Paul

I disagree. When traditionalists use stats like pitcher wins to describe how good a pitcher is, they are not accurately showing what has happened. A sabermetric tool like FIP is trying to BETTER describe what has happened. xFIP, xbabip, etc. are designed to predict the future, but a ton of metrics are not designed this way.

I’d say the difference is the traditionalists want to stick to measuring players according to tradition, while sabermetric people don’t see an inherent value in tradition; sabermetric people want to constantly be striving for more accurate measures.

Alec Denton
Member
Member

Thanks for your comment, Paul. I don’t know that a metric like FIP actually “better describe[s] what happened.” The name– fielding-independent pitching– itself indicates that it’s describing something other than “what happened.” In reality, pitchers have only their own defenders behind them, and when it comes to an examination of actual outcomes (i.e., team wins and losses), those are the only defenders that matter.

KK-Swizzle
Guest
KK-Swizzle

Nice! This is essentially what I’m going to write about next, and I think it is the source of much of the misunderstandings regarding sabermetrics. In my opinion, the difference between predictive and descriptive statistics simply isn’t talked about enough. In fact, your link there is one of the first instances I’ve seen it clearly laid out.

Alec Denton
Member
Member

Thanks, and hello from a West Michigan native.

Stuart
Guest
Stuart

I find that one of the hardest things to explain to people is probability (and hence believe that teaching probability should be a more fundamental part of math education in this country).

Too often, even someone well versed in statistics will forget that something that only happens 5% of the time, still happens an awful lot if your sample size is big enough. A good player has an awful month and a bad player has a great month all the time. Every once in a while, that month means something (irreversible decline, a new stance). Usually it is random. And yet it is human nature to look for order and reason where there is none. We all do it and more humility when you recognize this in others is always called for.

Brian Henry
Member

Stuart, it’s a good skill, but when do you teach it? I teach basic statistics to undergrads and a fair number of them struggle with it, so it is hard to go very deep in probability even with a fairly bright group.

Stuart
Guest
Stuart

I don’t know Brian. I teach graduate students (not statistics students . . . professional school) and they too struggle mightily. My gut is that it needs to be done (yes, along with a lot of other skills) long before you or I see them.

Paul
Guest
Paul

As an undergrad student myself who considers himself adept at basic statistical methods(t-tests, p-tests, chi squares, etc.), I have also seen smart people struggle with statistics. Basic statistical methods are not hard to understand by any stretch, so it may be finding just the right way to explain them. I have tutored other students on them, so I can commiserate with your troubles.

tz
Guest
tz

Stuart, it’s funny that you mention 5%, because I have what I like to call the 5% principle. Lots of attention in life gets called to the most spectacularly extreme 0.1% of any group, probability curve, whatever you can possibly analyze. Because of sheer mass, lot of attention gets paid to things with a frequency above a necessary critical mass, or tipping point.

And, as a consequence, we are prone to have too little focus on those items that occur around 5% of the time, whether it’s a demographic that’s just too small to throw their weight around or a breakdown in a company’s operations that happens often enough to cause real problems but not often enough to bubble up to the top of the list.

And I think that’s part of the reason why we try to come up with causative explanations for why the dice show snake eyes, rather than accepting that it can happen just by sheer luck. Because we like to feel in control of circumstances, and we like to pride ourselves in being ready to handle the “worst-case” scenarios, we end up trying to pass the buck on those “five percent” type occurences where we just haven’t shifted our focus.

This is why, for example, there are insane traffic tie-ups in the Deep South whenever there is wintry precipitation. These events occur just often enough to attract attention every few years about “them dumb-ass Southerners and their lack of driving skills”, but not often enough to justify the cost of the de-icing, sanding, and plowing equipment needed to quickly make frozen roads passable. If the powers that be decided that “well we got ice this year, so it must be the start of a trend”, they’d overreact and budget for winter weather problems like it was Minot, ND instead of Atlanta or Dallas. But the consequences of accepting that you may have to live with a shutdown of traffic every 5-10 because of such weather is that folks WILL try to find a reason why you got that ice storm so that they can feel control enough to FIX the issue.

I’m convinced that any human mind is capable of only deeply analyzing a finite number of hypothetical possibilities. When a possibility that we’ve barely focused on occurs, our typical response is like the investigative journalist who relentlessly pursues just exactly what caused that man to bite that dog. It’s much rarer to just say “$#!% happens.”

Sam
Guest
Sam

Great post on a great website from a fellow Hope College alum!!!!

kkrueger91
Guest
kkrueger91

Aw yeah!

DavidKB
Guest
DavidKB

Could this post be made standard course material for every public policy student?

tz
Guest
tz

I would highly recommend it. And this should also be a hyperlink for any comment board that wants to encourage healthy discussion of topics.

Kudos to Fangraphs for putting this in their “Best Of” for this past week. And many thanks to Kevin for putting this out there for us all to remember.

wpDiscuz