We sabr-ists need to focus on simple facts that are easier to digest than "but Ryan Howard's WAR isn't that good, and King Felix has a better FIP than C.C.!" We need to put data out there that others will be able to see for themselves, as opposed to bashing them with stats.

BBWAA voters love RBI? Fine. Point out that RBI are a function of lineup, and list Howard's RBI opps next to his (and others' ) RBI.

Point out that outs are bad; everyone likes "innings" for pithcers - lots of innings are good, right? Outs = negative innings, so OUTS ARE BAD. List Ichiro's and high OUTS totals next to his hits.

Don't argue ERA > Pitcher wins. It's often a loser, since games are won, "earned runs" are not won. Just list run support in a pitcher's record, and everyone can see for themselves why C.C. has so many "wins". They'll think they figured it out!

Park effects: list avg runs scored in each home park. Don't write articles about it. Don't confuse issue by quoting what individuals do in home/road splits. We simply need to find ways to get a few basic facts into MSM stat sections.

List batter outs, and RBI opps.
List pitcher run support.
List runs by parks.
Wait for MSM to trumpet their discover their new findings. :)

First Rule of Sabermetric Marketing…28Sep10…is don’t talk about sabermetrics. So I’m going to talk about being kicked in the balls, instead. Then I’m going to explain how my being kicked in the balls is totally relevant to marketing sabermetrics. OK? Let’s go:

I once wrote a blog entry on my old Catfish Stew blog about being kicked in the balls during an indoor soccer game. It went a little something like this:

Somebody forgot to give the goalie the message. Instead of easing up when we got close to contact, he came at me like some freakish combination of Ronnie Lott and Scott Stevens. He ran full speed for the ball, jumped as high as he could to knock it away from me, and in the process, sent his knee full force straight into my groin, and slammed the rest of me right into the hockey-style boards.

I'll let the Canadian R. Jones take it from there...

Point out that RBI are a function of lineup, and list Howard's RBI opps next to his (and others' ) RBI.

This might be a harder sell than is generally appreciated among sabrists (among whom I would count myself). RBI may be a very bad way of calculating who is driving in the runs in situations where it makes a real difference to a game's win expectancy. But it is still the most familiar way of assessing that very important contribution.

I don't understand why 'outs' as a category doesn't get more mileage, though. It's easy to calculate, so that's not an excuse in the way it might be for not taking into account pitcher run support.

Yes, we should get basic stats listings to sub "outs" for AB. A simple AB - H. If you want AB, add outs and hits.

MLB Outs leaders

Jeter D.. 474
Pierre J.. 457
Weeks R 457
Young M. 455
ScutaroM 454
Lopez J . 449 (who?)
Ichiro S. 448

Lopez J . 449 (who?)

Danks for nothing.

Danks for nothing.

:-)

What makes it particularly difficult in this case is that sabermetrics is essentially about removing emotions from the equation.

I guess this guy doesn't read too many of the discussions around here.

BBWAA voters love RBI? Fine. Point out that RBI are a function of lineup, and list Howard's RBI opps next to his (and others' ) RBI.

I tried this once on a Yankee board a few seasons ago. The response (repeated by multiple posters) was that I had just proven that A-Rod did in fact suck and Matsui was much better in the clutch, because 24.6% is more than 23.9%.

I don't understand why 'outs' as a category doesn't get more mileage, though.

The argument against using raw Outs as a counting statistic is the same as the argument against using RBIs: they're a function of opportunity. Basically, TomH.'s list of Outs leaders is mostly leadoff hitters.

So, just like with RBIs, you want to control for opportunities to have Outs be meaningful. But, of course, we already do that: Outs per PA is just 1 - OBP, and Outs per AB is just 1 - BA.

I'll let the Canadian R. Jones take it from there...

Here you go.

The argument against using raw Outs as a counting statistic is the same as the argument against using RBIs

Yes, but the mainstream does use RBI. Why doesn't it use Outs? In the context of this season's mood, Jeter's position at the top of the list would be another talking point about his downward career arc.

Why doesn't it use Outs?

I think because it's implicit in batting average. I think the mainstream's use of RBIs is more the anomaly. The mainstream doesn't use runs scored, for example, and I've heard them use the same argument against it as we use against RBIs - it's a function of lineup position and lineup quality. On the pitching side, writers use ERA instead of runs (or earned runs) allowed. I think that most mainstream guys, if they think about it, get the idea that stats are more informative when they control for context. For some reason, they just seem to have a blind spot on this subject with respect to RBIs (and pitcher Wins).

I would be hesitant to push Outs as a meaningful "sabermetric" statistic, because, as I said, its usefulness lines up in many ways with the usefulness of RBIs. There's a bit of tension, I think, in trying to simultaneously argue that RBIs need to be divided by opportunity (or ignored entirely), but we should look at total Outs instead of just looking at OBP.

In the context of this season's mood, Jeter's position at the top of the list would be another talking point about his downward career arc.

According to BB-Ref, Jeter's finished in the top 10 in the AL in Outs Made 6 times (including this season). He's 4th among active players in Outs Made, and 59th all-time. Jeter makes a lot of outs in large part because he's a leadoff hitter who gets tons of plate appearances (top-5 10 times, 3rd among active players, 47th alltime).

His career-low .265 BA and .335 OBP tell the story of his downward career arc just fine.

Didn't Jimmy Rollins lead the league in outs in his MVP year? Wasn't it the only thing he led the league in that season? Wasn't it still his best year by a mile?

[1] I agree that, with all the PBP data now, you can evaluate many aspects of player performance using transparent manipulations of familiar stats. One favorite example of mine is first base defense. Instead of the "black box" defense metrics, I'd rather look at unassisted putouts by first basemen, assists to bases other than 1B, 3-X-3 DPs made, etc. (B-R lists these stats plus GIBP% and RHB% for player and league PAs defended to adjust for context.)

Chris Jaffe, in the new Managers book, makes great use of simple, clear manipulations of stats to illustrate differences in managerial strategies.

For some reason, they just seem to have a blind spot on this subject with respect to RBIs (and pitcher Wins).

In which case, I return to my RBI observation. It does tell us whether a player is, not exactly clutch, but the guy who drives in that specific run that wins you the game.

Both these statistics do tell us something important, but they are not a very efficient way of doing so.

Which brings up the marketing matter — will 'Got Efficient Stats?' sell sabermetrics? Somehow I think 'Got RBI?' or 'Got Wins?' work better. In other words, on some matters sabermetrics is flogging the proverbial dead horse. We are not going to win here.

Now, 'Got Runs?' and talking about Runs Created (or whatever) begins to work.

Both these statistics [RBIs and pitcher Wins] do tell us something important, but they are not a very efficient way of doing so.

I think you're exactly right here. I would phrase it slightly differently, however. Both of these statistics tell us something "real". They identify real runs and real wins.

Now, 'Got Runs?' and talking about Runs Created (or whatever) begins to work.

And the problem, here, is that Runs Created aren't "real"; they're theoretical. The more context is removed by sabermetric stats, the more informative those stats become in terms of understanding true talent but in some ways the less informative they become in terms of what actually happened. I think this gets to a fairly large part of the problem with sabermetric stats from a mainstream perspective. When we talk about Wins (above replacement, above average, as shares, whatever), sabermetricians typically aren't talking about real, actual team wins. They're talking about how many wins a team SHOULD have won, not how many the team actually won.

The fact is, those numbers aren't very different and, to the extent they are different, the difference typically doesn't really tell us much of any use. And that's an important point to make and sabermetricians should definitely try to sell that - that WAR does line up very, very well with real wins. But that's still a step removed from the extent to which RBIs line up with actual runs scored or pitcher wins line up with team wins - they do so perfectly, by construction.

And the problem, here, is that Runs Created aren't "real"; they're theoretical.

OK, but that's why I think the way forward has got to move from something like WPA, than Win Shares or WAR. Then you're down to 'Got Wins?'. But I suspect that's a controversial position.

People tend to be too invested in ideologies to switch them like they might switch to a new brand of dish detergent. (Though I'll grant you some people have an unnatural allegiance to their dish detergents.)

I'm not sure history would show that marketing campaigns are particularly effective at changing ideologies. Paradigm shifts tend to happen when enough of the old guard die (literally). And honestly there's enough of a competitive advantage to being more right than the other guy that you really don't have much incentive to correct his mistaken beliefs.

The question in my mind is why do sabermetricians care so much if the BBWAA hands out its awards to undeserving players? I suspect most of them don't really care at all. Who wants to go through the trouble of having to lead the horse to the water and then trick the ############ into drinking it?

Let Sabermetrics be for none and all, to steal a phrase from Also Sprach Zarathustra.

The question in my mind is why do sabermetricians care so much if the BBWAA hands out its awards to undeserving players? I suspect most of them don't really care at all. Who wants to go through the trouble of having to lead the horse to the water and then trick the ############ into drinking it?

This would be the group that's just about to award Felix Hernandez the Cy Young award. That BBWAA?

Perhaps it's best to realize that more and more baseball people are generally coming around to sabermetrics, doing so at the pace they feel comfortable. My only suggestion would be to let that happen, and not to try to prod it along by telling the still-unconverted how ####### stupid they are. That would be the extent of my marketing plan.

I'm just glad that Ken Arneson is writing something, anything.

This would be the group that's just about to award Felix Hernandez the Cy Young award. That BBWAA?

Hence my use of the word if and not when.

1. Sabermetrics has already followed one of the "geek" marketing paths -- slaying the sacred cow. RBIs are huge, bunts are good, clutch hitters exist, defense matters, pitcher wins are huge, Jim Rice was a god, etc. (Of course now it's in the awkward position of "OK, defense does matter ... and a LOT ... well, one stat says it's a lot, the other one says this guy is average but you know what I mean.") This is similar to Freakonomics -- which, intentionally or not, followed the sabermetric marketing plan -- and even to an extent Malcolm Gladwell.

2. That leaves the second path of "geek" marketing success -- humor with an appropriate mix of self-deprecation and other-deprecation. Think Daily Show with stats.

3. As to a "serious" marketing strategy and comments in this thread ... I think you guys have got it wrong. Outs aren't good, RC and WAR aren't good, etc. The successful marketing point, if one exists, is "we can predict the future." Of course one problem with that is that we can't do that well at all we can only (hopefully) do it better than the folks who rely on old-school stats.

3a. There is a potentially successful retrospective path along the lines of "why teams win." Just make the points that teams score through OBP and SLG; that they prevent runs by limiting others' OBP/SLG via pitcher strikeout/walk/HR rates and good defense. If folks buy that then they'll easily buy that players with those characteristics are the best players.

Alas there is the very real possibility that the "edge" sabermetrics gives you is so small as to be meaningless to the fan (and possibly even teams).

People tend to be too invested in ideologies to switch them like they might switch to a new brand of dish detergent.

But when did this all become ideological? AFAIK, the only reason people objected to Bill James back in the 1980s was because he wrote stuff like:

Many athletes truly believe that they are successful...because they're better people than the rest of us...That's where all the ######## about clutch ability comes from.

Throw in a dash of mgl and a pinch of BPro's superciliousness and you've got reasons to get annoyed with saberist nerds.

Bill James and sabermetrics would not have got nearly so far without that kind of attitude, but I don't see it as an ideological struggle. The mainstream doesn't make as much use of sabermetrics as it could because saberists had no qualms about pillorying their sources, and they could curry favour by shouting back.

It isn't simply 'we can predict the future.' It's 'we can predict the future so you can win your fantasy baseball league and talk smack to your friends.' That motive more or less gave birth to DIPS.

But when did this all become ideological?

I think there's a difference between those who are receptive to sabermetrics and those who are not that is reflective of an ideology about how new knowledge is acquired.

I stay away from serious marketing - I'm not a very serious person.

This would be the group that's just about to award Felix Hernandez the Cy Young award. That BBWAA?

We'll see. I'm sure he'll get a lot more votes than any other .500 winning percentage SP has ever gotten, but I still don't think he will actually win.

more and more baseball people are generally coming around to sabermetrics, doing so at the pace they feel comfortable. My only suggestion would be to let that happen, and not to try to prod it along by telling the still-unconverted how ####### stupid they are. That would be the extent of my marketing plan.

I do agree with that part.

We'll see. I'm sure he'll get a lot more votes than any other .500 winning percentage SP has ever gotten, but I still don't think he will actually win.

Really? At this point, I think he'll win pretty comfortably.

Alas there is the very real possibility that the "edge" sabermetrics gives you is so small as to be meaningless to the fan (and possibly even teams).

The "edge" sabermetices gives "to the fan?!" What does this even mean? Sabermetrics is useful in exactly two scenarios. First, fantasy baseball, as long as the league is saber-friendly and counts OPS instead of AVG and SLG instead of RBI; and second, roster construction for actual baseball teams. If you're not constructing a baseball roster, or pretending to via fantasy baseball, there is no "edge" to be had. All that "edge" would amount to is the feeling of smug superiority that you lord over your "ignorant baseball friends." There is no other use for sabermetrics, or performance accounting of any kind.

Similarly, what exactly does "marketing sabremetrics" mean? That's like "marketing analytic processes." To whom do you "market" such things. Sabermetrics isn't a product. It's a series of analytic processes that claim to hue more closely to scientificly and statistically valid evaluations than traditional methods. You don't market that. You might market your book, built on "sabermetric principles", that you want people to buy every year before Fantasy Baseball season. Or you might want to market your new stat, either for the egotistical joy of having people use "your stat" as a measuring stick for performance or in order to leverage a paying gig doing analysis for a sports team or sports website. But again, you're not marketing sabermetrics. That phrase doesn't even make sense.

As for the author's point, however weird and irrational it is at heart, the best "marketing" sabermetric ever got was being hated by Murray Chass and company.

The First Rule Of Sabermetric Marketing:

Is you Do Not Talk About Sabermetric Marketing.

The Second Rule of Sabermetric Marketing:

If this is your first Market, you must Sabermetrize.

We'll see. I'm sure he'll get a lot more votes than any other .500 winning percentage SP has ever gotten, but I still don't think he will actually win.

My running total has King Felix with 8 first place votes and Sabathia with 2. Of course this is subject to change after tonight's 3-headed battle...or some writers were just bullshhitting.

Alas there is the very real possibility that the "edge" sabermetrics gives you is so small as to be meaningless to the fan (and possibly even teams).

The "edge" sabermetices gives "to the fan?!" What does this even mean?

I don't think he's talking about sabermetrics "giving" an edge to any particular fan -- I think he's allowing for the possibility that some differences between "sabermetric" thought and "mainstream baseball" thought might be subtle enough that your average fan out there wouldn't really care.

Otherwise, I agree with what you're saying.

The question in my mind is why do sabermetricians care so much if the BBWAA hands out its awards to undeserving players?

It's just a more sophisticated way to engage in a tradition as old as the game itself - second-guessing. Baseball fans do that all the time; sabermetrics just gives them another outlet for doing it.

-- MWE

I don't think he's talking about sabermetrics "giving" an edge to any particular fan -- I think he's allowing for the possibility that some differences between "sabermetric" thought and "mainstream baseball" thought might be subtle enough that your average fan out there wouldn't really care.

Okay, if this is more the point being made, I'll walk back from the ledge a bit. Still, I'm a little confused as to what "marketing sabremetrics" would actually mean. The gestalt seems to be "how can we convince more people to think like us," which isn't really a marketing campaign as it is a political campaign. Sabermetricians believe they have a deeper knowledge of the game, based on statistical analysis and scientific management theory. They believe that this deeper knowledge should give them more intellectual cache within the game's managerial and critical space than they currently have. Sabermetricians are basically liberal technocrats of baseball. Whenever you guys figure out how to "market" the "no, trust us, we know this better than you, we've done *regressions*" thing to an emotionally driven population deeply distrustful of "experts" and "elites" and their "scientific models" for governing behavior, you could make *billions* by offering said marketing process to the Democratic National Committee.

I'm just glad that Ken Arneson is writing something, anything.

I guess I need to be kicked in the crotch more often.

The gestalt seems to be "how can we convince more people to think like us," which isn't really a marketing campaign as it is a political campaign. Sabermetricians believe they have a deeper knowledge of the game, based on statistical analysis and scientific management theory. They believe that this deeper knowledge should give them more intellectual cache within the game's managerial and critical space than they currently have.

I think this was where Will Carroll was coming from initially, and Tango's response (and MGL's in the comments) is that the second part of that statement is dead wrong, at least as it applies to what they do.

-- MWE

I didn't realize there was a difference between a marketing campaign and a political campaign.

I don't have a clue who the "market" for "sabermetrics" is nor exactly what they're "buying." Probably better to ask Will Carroll. My point is that, other than marketing fluff, if you want to "sell" "sabermetrics" then you need to convince people that it does what they want better than the current "product."

My point on "edge" was along the following. I don't know why people want to "buy" projected standings but given every baseball writer and every saber projection system spits them out, presumably these are a popular "product." Now it might well be true that, if measured by Spearman rank correlation or some such, ZiPS, PECOTA, DM, etc. do better than lunkhead sportswriters, gamblers and drunks in the Wrigley bleachers but not with such uncanny accuracy that the public will be wowed. And if the crackhead on the corner tells you the 2005 Sox are going to win the Series while PECOTA tells you they're gonna finish last in their division ... you're going to remind them both that it's now 2010.

So while I don't have a clue what the "average" or "casual" or "non-sabery" or "anti-stathead" fan is "buying" in the baseball statistic "marketplace", I do suspect that "sabermetrics" has to offer a "product" that is either superior to the current "product" or does something no existing "product" does.

The identification of this "market", the existing blah "products", the exciting "sabermetric" "products", the demonstrated superiority of the new "proucts" over the "old", etc. are left as an exercise for someone with even more time on their hands than me.

My own personal opinion is that the war (so to speak) to "market sabermetrics" is over, and we won. The opposition has reached the point of aggressive, hate-filled, defensive counterattacks, much like the Tea Party's Social Darwinists, and no one can get through a broadcast of even one game without some broadcaster commenting on some sabermetric stat. The current level of infighting within sabermetrics is good evidence for this. If you are confronted with what appears to be an unbeatable enemy (the British or conventional baseball opinion), you band together (the Revolution, early sabermetrics). You put aside most of your own conflicts with each other. When you finally win, that's when the fragmentation (political parties, Burr/Hamilton duels, arguments over WAR methods) starts up.

There were two marketing "strategies" that worked, neither intentionally, as far as I can tell. One was John Benson and others' takeover of the fantasy industry by providing better prediction stats than anyone else could do. The other was to provide good fuel for sports bar arguments. The way I first figured this out, in fact, was when my little column in St. Louis' Riverfront Times started to get serious publicity. The RFT is an independent moderate left-leaning political paper at its core, but my column was up on every sports bar's bulletin board the day the thing came out, every week. And you know, if you're in a sports bar argument and someone throws a saberstat at you and you lose, you start to learn about sabermetrics.

BTW, the reason for people's fascination with homers and RBI is visual. You see the guy hit the sac fly to drive in the guy who just hit that triple. You see the triple hitter score the run "on the fly ball." You've forgotten about that triple. You end up associating RBI with runs, because that's what happens visually. Same with homers. You see the homer go over the fence. You see the batter and all the baserunners scoring. You forget about the three walks that loaded the bases, but you remember the homer and its RBI. And hence my continual comment on Moneyball: The only real contribution that the book makes is to point out that taking walks is a tool, not a skill. Traditional scouts use only the five other tools because those are what they can see (running speed, throwing arm power). The ability to take walks isn't nearly so visual, especially if you're only seeing the prospect for a game or two. To a large extent, sabermetrics is simply the science of measuring the impact of baseball events that aren't visually immediate in terms of their scoring power.

- Brock Hanke

The First Rule Of Sabermetric Marketing:

Is you Do Not Talk About Sabermetric Marketing.

The Second Rule of Sabermetric Marketing:

If this is your first Market, you must Sabermetrize.

Primey.