There's No Escaping Malcolm Gladwell

Malcolm Gladwell has a number of public responses to the sort of thing I ranted about the other day-- not to me specifically, mind, but to the same general points-- on his own blog and on ESPN's Page 2. It's pretty much the same argument others made in the comments to my post. Taking a fairly representative bit from his blog post:

The press is not for everyone. But then the piece never claimed that it was. I simply pointed out that insurgent strategies (substituting effort for ability and challenging conventions) represent one of David's only chances of competing successfully against Goliath, so it's surprising that more underdogs don't use them. The data on underdogs in war is quite compelling in this regard. But it's also true on the basketball court. The press isn't perfect. But given its track record, surely it is under-utilized. Isn't that strange?

The problem here is that I'm not sure what track record he's talking about.

I've been following college basketball closely for twenty-five-ish years, and I just don't see that much support for the claim that pressing basketball is the best way for untalented teams to beat better competition. The more common pattern is actually the other way around-- teams that press tend to whip up on inferior opponents, but struggle to beat teams that are good.

There are only a handful of championship-level teams over the last quarter-century that have won by playing pressure basketball, and two of them were absolutely stacked: Pitino's 1996 Kentucky team, and Jerry Tarkanian's 1990 UNLV team. The best example of a pressing team to win with arguably mediocre talent would be the 1994 Arkansas team, who edged a Duke team consisting of Grant Hill and a bunch of stiffs-- talent-wise, Duke was probably the underdog in that one.

The more striking pattern is in the losses. Essentially the same UNLV team that won in 1990 lost in the semifinals the next year to essentially the same Duke team that they had run out of the gym the year before. Arkansas made the title game again in 1995, losing to UCLA. Billy Tubbs took Oklahoma to the title game in 1988 playing pressure basketball, and lost to Danny Manning and a bunch of stiffs. Georgetown under John Thompson played a tough full-court game, and went 1-for-3 in title games with Patrick Ewing, losing to UNC under Dean Smith, who was (in)famous for the "four corners" slow-down game, and a Villanova team with no great talents.

If you look at the recent-ish history of NCAA basketball, you're more likely to find less-talented teams winning by playing slow than by pressing. Think Princeton over UCLA in '95, and almost over Georgetown in '89. Kansas in '88 played Oklahoma even for a half at high speed, and beat them playing slow in the second half. Villanova beat Georgetown by limiting possessions. Davidson last year got within a game of the Final Four last year not by making it a frenetic full-court game but by playing under control at all times. Ben Howland has made it to three of the last four Final Fours by playing some of the slowest, ugliest basketball you could ever hope to see. The one loss that 1996 Kentucky team had was to a UMASS!!! team with Marcus Camby and a bunch of role-players, who won by not trying to run with the loaded Kentucky team. And on, and on.

What really ought to be the main example of the effect he's purporting to find is Loyola Marymount under Paul Westhead in the early 90's. They played a ridiculously up-tempo pressing game using players who couldn't make it at better-known schools, and managed to get themselves on ESPN pretty much every night for a couple of years there. They're not in the piece, though, because that's pretty much all they did-- they got a lot of attention from the sports media, and won a lot of games in their terrible little conference, but whenever they had to play good teams-- teams with talent as good or better as their own-- they got beat. Sometimes they got blown out.

Similarly, Jerry Tarkanian made a name for himself at UNLV with an intense up-tempo pressing team, when he was working with really talented players and lots of shady recruiting. When he left under a cloud, and had to work with rejects and head cases at Fresno State he didn't do much of anything. For that matter, Westhead has had a couple more cracks at it, and never managed to match his Loyola Marymount success.

This is why the piece bugs me. It's not just that he's flubbed a few details here and there in a basketball analogy-- the whole thing is built on a handful of cherry-picked examples, and falls apart if you look too closely.

Gladwell is sufficiently wrong about his claims regarding basketball (at least as I see it) that I don't feel I can trust what he says about the other pieces of evidence he offers. The broad claim-- underdogs do better by changing the rules of the game-- ought to be fairly uncontroversial, but the stuff he says about the main example around which he's constructed the whole article is so far off that it makes me wonder whether he's reporting the other evidence correctly, or if he's garbled that as well.

More like this

I'm never quite sure what to make of Malcolm Gladwell. Lots of smart people seem to be favorably impressed by his writing and ideas, but whenever I actually read anything by him, there doesn't seem to be much there. Take, for example, this New Yorker piece on basketball as a metaphor for innovation…
Kate and I went to the two games of the "semifinals" of the 2K Sports Classic Supporting Coaches vs. Cancer, Your Name Here for a Prince pre-season "tournament" last night (the scare quotes are because the four teams playing last night were guaranteed to be playing last night, regardless of what…
Two sports posts in one day? They're going to revoke my membership in the Association of Intellectuals. But hey, college basketball is one of my great passions. ESPN has been discussing and debating who would go on the Mt. Rushmore of college basketball for both coaches and players during the ESPN…
Well, my Blue Devils just lost their first game, to Georgetown on the road by 3. What made this a fascinating game is what it tells us, both good and bad, about this Duke team. This game simultaneously showed why this Duke team will have a difficult time winning the national championship this year…

The more common pattern is actually the other way around-- teams that press tend to whip up on inferior opponents, but struggle to beat teams that are good.

What are you using to choose which teams are superior or inferior? It must be something other than win-loss record because that would make this argument tautological. It does seem like disentangling "talent" from "performance" is a rather tough thing to do for making these kinds of arguments.

You're using the examples of champion level teams, but if he's talking about Davids vs. Goliaths then those don't seem very relevant to the argument, because it would be pretty unlikely for Davids to win enough games against superior competition to get to the point where they were at the championship game. (You also use examples in your next paragraph to support your thesis like Princeton-UCLA, which was of course a 1st round game and not anywhere near the championship.) Further, if the press is a comparatively rare strategy, then arguing that there are more famous upsets that involved a slower pace doesn't do much to show the relative success of the strategy. If there are 10 times fewer teams pressing, then it's a relative success if they have only 5 times fewer upsets, obviously.

I don't know what these numbers look like, but my point is that it seems like your rebuttal to his anecdotal evidence is merely offering different anecdotes and doesn't have much more obvious claim to rightness or wrongness.

I think you may be a bit harsh here, he didn't go THAT far. Sure, his examples are a bit cherry-picked, but the overall premise isn't terrible basketball logic.

If a team is losing, what do they do? Press! It heightens the chance that the team will get blown out, but it also is understood that it increases the chance of a comeback. Stretch that logic back to warm-ups, if you know your team is likely to lose, as a coach, you should be looking for new strategies to try (different zones, traps, and presses, gimmick defenses and offenses) to give you a greater likelihood of winning. Increasing that likelihood is what Gladwell is talking about when he says it makes teams better.

Sure, there are plenty of examples where a pressing team failed, it's a much riskier strategy. but do you really think a team of knuckleheads like UNLV had (excusing Stacey Augmon) was going to play a basic basketball game and beat Duke regularly? The odds would favor Duke. What happened is what we should happen, UNLV blew them out once and then lost once as well. While we all were shocked by the loss the second year, we probably shouldn't have been, UNLV's strategy was inherently risky. Which is why coaches in college don't do it a lot, like most humans, they are risk averse. Don't win the title, OK. Don't keep my job because I lost games I shouldn't have? Not OK.

Gladwell isn't so off base with the basketball stuff that you should throw everything else out.

In the quote you pulled you explained Gladwell's point exactly.

But then the piece never claimed that it was. I simply pointed out that insurgent strategies (substituting effort for ability and challenging conventions) represent one of David's only chances of competing successfully against Goliath, so it's surprising that more underdogs don't use them.

Those slowdown strategies you outlined are insurgent strategies just like the press is. Some less talented teams don't have the talent to use the press. Or a team tires too quickly. A good recent example is last year's Chicago State team, a Division I independent which stayed with Northwestern and Marquette for a half, but then tired while playing the press.

I made the argument today on Chicago College Basketball that Northwestern is doing exactly what Gladwell suggests by running the Princeton Offense and 1-3-1 zone.

Those slowdown strategies you outlined are insurgent strategies just like the press is. Some less talented teams don't have the talent to use the press.

See, right there, you turned the whole thing inside out. One of the central claims Gladwell makes about the story if that the press is a great underdog strategy for teams that don't have talent. In his telling, there shouldn't be any team that "doesn't have the talent to run the press."

I have no objection to the fairly banal claim that if you don't have as much talent as the opponent, you shouldn't try to play the style of game they favor. The problem I have with the article is that he goes farther than that, making two linked claims that I don't buy.

The first is that the press is only rarely used, which just isn't true. It's not difficult to come up with examples of coaches who aren't Rick Pitino who have used the press effectively, and there are more who have run the press and not gotten very far. Full-court pressure is a standard part of any decent team's basketball arsenal, and there is no shortage of teams who make it their primary weapon.

The second claim is that the press is exceptionally successful, making it "under-utilized." That, too, I find dubious, given NCAA history. The vast majority of top-flight teams play a sort of mid-tempo game-- running if they can, playing half-court if they can't-- but looking at the moderate number of teams who either press all the time or deliberately play slow, it's not at all clear that pressing provides a great advantage. Slow-playing is at least as effective as pressing, arguably more so.

As an illustrative example of the way that changing the style of play can be effective, the anecdote about the girls' team is fine. But Gladwell goes beyond that, making claims that there's some sort of basketball orthodoxy that leaves most coaches too short-sighted to see the wisdom of playing "40 minutes of hell." That's a fairly broad claim, and not one that I think can be supported. In reality, most coaches play a mixed style not merely from inertia, but because a rational consideration of the game leads to the conclusion that that's the right thing to do.