« Canadian Healthcare, redux | Main | Thoughts on Freakonomics »

Comments

Bob Calder

So does this apply to using graduates of the USMA in critical positions in the Army as well?

Hueristics like that work as long as the playing field doesn't change. There is absolutely no reason for it to fail.

On the other hand, a highly ritualized selection process will fail to adapt to rapid change by selecting those who are able to change radically in a way that is not within normal parameters.

So the sports analogy is perfect for the theory. One place I think ordinary thinking fails is the assumption that a school or a training process can make outstanding people. You can apply all the methods you want to the cohort of trainees in order to train them to perform like stars. Unfortunately there are aspects of character, mental abilities, and physical abilities that are outside the norm. So when you drop these well trained people into the situation and expect them to perform like the star, they don't. Because they can't.

U Conn has somebody who is able to look inside in a way that is unusual. I wonder if most of the stars pass through the hands of one particular recruiter? It could also be a matter of incentivising players and their families.

Ben Hyde

Cialdini's work on influence is really about how marketing perverts otherwise extremely effective heuristics; though I'm not sure he knows that.

jrw

But is it possible to know less than Isaih Thomas?

Sean

You're exactly right KoreyL.

However, perhaps, the Kentucky program just isn't up to snuff.

Destroyed the Celtics.

Adam

You say that "in off-years for Duke and UConn I can trade down and stockpile picks." But how will you determine whether it's an "off-year"? You cannot defer to Duke and UConn here because every year they recruit players and every year there is obviously a best and second-best player.

Similarly, if you get the number one pick in the draft, you would automatically take the "best" UConn/Duke player. But how would you determine who was the best? Would it be based strictly on college performance? (Even then, would it be the highest scoring average? rebounding? assists? some combined measure?) If not, then it must at least be based on their NBA potential, which is another way of saying draft experts' views? So the question is whether you will have your own staff do this or just rest on others' views. If so, the most interesting thing that you'd be doing seems to be saving resources by trusting that the market for evaluating NBA potential is basically efficient, at least for players from two high-profile colleges.

Adam Hoff

I am intrigued to find this argument, especially here, since it is basically the same argument I am making about USA basketball. Rather than beat ourselves up trying to piece together a team through tryouts and interviews (sending the message that we are desperate), we should just get all the best Duke, or UConn, or Roy Williams (Kansas and UNC), or Arizona players, put them with their old coach, and go win the gold.

You can find my column on an all-Duke USA team here: http://www.whatifsports.com/insider/default.asp?article=20060307, and the "roster alternatives" on the accompanying blog at http://wisinsider.blogspot.com/2006/03/alternatives-to-duke-plan.html.

It's probably borderline inappropriate to post those links here, but I am very interested to see what some takes might be on this method.

At the very least, I think between Gladwell's theory and my own, it dispells the notion that Duke players always fail in the NBA.

Cheers,

Adam

Adam Hoff

Whoops, I screwed up the links by including punctuation.

Here they are:

http://www.whatifsports.com/insider/default.asp?article=20060307

http://wisinsider.blogspot.com/2006/03/alternatives-to-duke-plan.html

Brett

This reminds me of something I heard (that I have absoultely no way of verifying the truth of) when I was an intern at Goldman Sachs about 10 years ago - people were talking about how other investment banks were hiring away people Goldman with higher offers after they had worked at Goldman for a short time.

The theory being, of course, that they could spend less on recruiting, and piggy-back on Goldman's extensive recruiting and interview process. Paying someone an extra X% after knowing that Goldman already hired them was worth it.

tom moon

Isn't the NBA supposedly rigged anyway?

LarryT

For a related discussion of player loyalty, see expat professor David Allen's (Instituto de Empressa)blog on soccer team Real Madrid: http://www.davidbruceallen.com/

Luke Middleton

While Gladwell's team would be better than the Knicks (unfortunately, possibly just for reason of not having inconsistent, demeaning, and detached Larry Brown overseeing), we must also factor in the concept of "team".

While positions were paid attention to, there's also the recent US Olympic team side of things (another time a random group of players was put together and handed to Brown). What style of play do you subscribe to? Who's the focal point of the offense? Where are your weak links on defense? Up-tempo or half court? These are all things that can obstruct good players and make previously subpar players shine (see: the Suns offense).

And then you throw in the coaching style of someone like Hubie Brown, and the list of factors increases...

Nutlow

I think if you look at the players drafted out of the University of Arizona, they significantly outperform those players selected before them. It may not solely be about the best programs but about the coaching and the program that best prepares players for the NBA. From 1985-2004 these are UofA players drafted in the 1st round vs the previous pick in the draft listed in parenthases:
04:9-Andre Igoudala(RafealAraujo)
01: 13-Richard Jefferson (Vladimir Radmanovic)
99: 19-Jason Terry (Shawn Marion)
98: 14- Michael Dickerson (Keon Clark)
95: 7- Damon Stoudamire (Bryant Reeves)
94: 12- Khalid Reeves (Carlos Rogers) 13th Jalen Rose
93: 22- Chris Mills (James Robinson)
91: 10- Bison Dele (Stacey Augmon)
89: 3- Sean Elliot (Danny Ferry)
Notable second round picks:
88: 3- Tom Tolbert
88: 53- Steve Kerr
01: 30- Gilbert Arenas

Lester Spence

There's a problem here...and I couldn't put my finger on it until I thought about it a bit. Isiah Thomas' problem has nothing to do with his ability to draft. Including his stints in Toronto and in Indiana, Thomas has made generally good decisions in drafting players. Much better than his backcourt partner Joe Dumars for example.

But where he sucks (and where Dumars excels) is in evaluating PRO talent. Now your bottom line argument--that Thomas thinks he knows more than others--is on point.

But you'd have to come up with another heuristic to do better....because I don't see how yours works in the free agent game.

Josh

I love this blog!

Jesse

So now extend your brief mind exercise a bit further. More and more coaches begin thinking of throught through the lense of heuristics. Assuming these same coaches will only pick from regularly-attending Sweet 16 teams, the pot of schools for which these coaches pick remains small and demand for such players shoot through the roof. Now, how easily will you be able to "wine and dine" for Battier or trade up for such solid draft picks as you suppose? My point: this idea of heuristics is cute and sounds great! The average person, with limited knowledge, could put a team on the court that could rival any mediocre team. Just as economic models are not only stripped down and simplified but rarely work in the real world. However, I would love to study further. Thanks for introducing me to the idea.

Drew Margolin

It seems to me that the idea of the heuristic is a good one but that this particular heuristic is not so good. I am skeptical for two reasons:

1. Ratios of stars to busts.
The goal of any draft decision is to sort wheat from chaffe. Thus, any analysis that does not explicitly compare their ratio is useless. The question is not how many Duke or UConn players are good in the NBA, but what ratio of players that were thought to be good from these programs actually ended up being good? These programs may still be better than average, but this is the analysis that needs to be done.

The whole point of the heuristic is that it eliminates other decision-making tools, so you can't look at Mike Dunleavy or Christian Laettner and say they were worth less than Brand or Hill, even though they clearly are, because this assessment is based on some other "basketball knowledge" whose unreliability is the purpose of the heuristic in the first plac.e

2. Meta-effect of hype.
Even if Duke and UConn could be historically shown to produce a better ratio, the very act of producing this analysis would begin to erode the ratio. Why? Because the denominator of the ratio is based on how many players are "draft-worthy" from a particular school, a number which is not fixed by some independent assessment but rather a fluid figure based on the school's reputation. Thus, if Duke and UConn are shown to have more successful players, more of their players will appear "draft-able."

This may seem convoluted, but this is precisely the logic at work when we assess smaller school products. For example, Marquette has, as far as I can remember, 100% ratio in the last 5 years -- Dwayne Wade, that's it. There are other schools that only get one player into consideration every year or so, and thus have a very high chance of achieving 100% ratio. Does this mean we should consider their best players every year for the draft? Of course not. This is because we know "intuitively" that there is a degree to which conventional wisdom informs us of which players are legitimate for consideration. Marquette has a high ratio because we "know" that even very good Marquette players are not "draft-able." That is, Marquette's mediocre reputation as a program keeps their denominator low. By the same logic, we accept that Duke and UConn will have lower ratios because even some of their non-stars are "draft-able" and therefore get drafted.

We can do this analysis at any level -- just players worthy of top 5 consideration, or top 10, or total draft. The problem will always be the same. Either you have some objective standard for assessing who is draftable, or you have a conventional wisdom/reputation standard. The former requires you do some additional analysis beyond your heuristic, the latter means that using your heuristic screws up its value as you will evaluate more duds as valid at any given level.

Over time, no meaningful conclusions can be reached by looking at only one variable and ignoring all others as the process/system being observed will shift toward that variable and abandon the "common sense assumptions" that made it meaningful in the first place.

Yuhau Lin

So Bill Simmons, Malcom Gladwell and Chuck Klosterman as well as Len Pasquarelli all seem to converge on basically the same theory, that GMs overthink things and that's why they make seemingly obvious mistakes.

While it's tempting to think that this is the primary cause, and I would really like to think so, I'm starting to think that the real reason so many GMs screw up all the time and get fired might be that it's a completely unreasonable and impossible job to do.

Making roster and/or coaching changes on a 13 person team in a game that plays only five people at a time is inherently more volatile than any decision that, say, a CEO or even regional VP of a company would make. More volatile by several orders of magnitude. On a basketball team, you typically have 8 guys that matter, on a really deep team like the Mavs, maybe 10. I don't have any stats to back this up, but I would imagine that a good 80-90% of your salary cap is dedicated to these 8-10 people. So every meaningful roster move you make as a GM is going to have significant ramifications on your team performance-wise, as well as on the long-term financial health of the franchise. Every single one.

The real (non-sports) world is totally different. People are not generally expected to make multiple pivotal decisions on a yearly basis. Take the example of a CEO like Dick Parsons, who certainly holds a position that most would consider quite challenging. He's a successful CEO. Everyone is pretty much happy with what he's done. He's becoming an American business icon. But all he's really done is let decent assets show their true value, meaning he's basically gone around making sure everyone is happy and continuing to do the things they're doing. That's certainly an oversimplification, but the point is, he hasn't made any moves that could be considered pivotal or the equivalent of trading a rotation player and/or signing a significant free agent for a basketball team. If he was an NBA GM, by now he'd have spun off AOL, sold the movie division to sony, bought assets in China, signed partnerships in Latin America, all the while being held to a standard by shareholders to not only outperform the S&P on an annual basis, but to be the highest valued stock in the industry.

The sports equivalent of Parsons' success at Time Warner would be Jerry Colangelo coming into the wasteland that Babcock left him and then just doing nothing. Just letting the team figure stuff out for a few years before making a decision on what to do with the assets. Could that ever fly in sports? I mean is it even remotely reasonable to think that he'd be considered a competent GM if he did that?

And what's more, that Colangelo situation is actually impossible, given the nature of the draft and free agency itself. Colangelo already has two monumental decisions in front of him that, in a manner of speaking, he did not choose to pursue but were forced on him -- signing Chris Bosh to a max contract and the #1 pick in the draft. He's been on the job for a few months and he already has to make possibly the biggest consecutive decisions the franchise has ever faced, not to mention how he's going to spend the cap room, what he's going to do with Mike James, how he can fleece Isiah before the rest of the league gets to him, etc. And this all continues through to the trading deadline and starts all over again next offseason.

Another aspect of this phenomenon is that not only do GMs have to maintain unreasonable levels of accuracy on decisions that are disproportionately pivotal, but they also are dealing with assets that are orders of magnitude more volatile themselves than the typical business asset. A substantial business asset with relatively good financial health cannot and does not, unless under extraordinary circumstances, lose all of its value overnight. But professional ball players can and do with great frequency. Every 7 footer with a big contract. Every head case in non-contract years. Every aging former all-star signed by the Knicks. They can go from assets to liabilities literally the second they put on their new jerseys.

Take the real world (counter)example -- Carly Fiorina. She pushed through the Compaq acquisition upon her arrival. But she lasted 6 years. They gave her six years to prove that it was a good acquisition. And that's the only significant move she made throughout her tenure. And it took 6 years for her to be made accountable for it. Could you imagine Jerry Buss waiting six years to see if the Shaq trade (and only that trade) resulted in a championship? Or better yet, the Orlando Magic holding onto Steve Francis, Cuttino Mobley and Kelvin Cato for six years to see if they would eventually equal or surpass the value of Tracy McGrady?

That's obsurd in sports, but it's totally reasonable in the real world because Compaq is a viable asset. No matter how value dilutive to the group it turned out to be, it still had and still does have value. A franchise like that cannot fail overnight, where a guy like Stevie Franchise can and does all the time. Even if Compaq is an obvious drag on the total value of HP's assets, it still contributes value in its own right and the process of realizing that it will never turn itself around to drive an increase in group value is a very long one.

So all of this seems to mean to me that maybe it is nearly impossible to be a good GM. Joe Dumars? Darko. Popovich? Brent Barry and Nick Van Exel. Cuban? Petrie? Colangelo? They'll all have their share of failures. But maybe they should be allowed to do so. Take Dumars for instance. Let's say, and this is an entirely realistic possibility, that they don't win the finals this year, then they overpay to keep an aging Ben Wallace (or let him go entirely), and Darko improves upon the decent start he's established in Orlando to become even 75% of the player he was once thought to be. Well, they traded Darko for basically nothing in order to clear room for Big Ben, so if that doesn't pan out, not only will they have blown the 2003 draft by not taking wade or bosh or carmelo, but they will have either commited $10+ million a year to an aging one-dimensional player, or will have gotten nothing in return for the #2 pick in the 2003 draft, one of the deepest and most star-filled drafts of our lifetimes. Where does that put Dumars? How long before the "he got lucky with Billups, the Wallaces and an amazingly good string of health for three years and all he got to show for it was one lousy championship and a maxed-out Tay Prince" articles litter the press? Sure, he's been one of the better GMs so far, but he's facing (as every GM does every year) career altering (and possibly threatening) decisions this offseason. It's really just too much pressure to put on one person. It's an impossible job. That's what I think.

Which brings us to Isiah. Who's to say he's not the next Elgin Baylor? Dolan obviously doesn't care about costs, so really, all Isiah needs to do is win, regardless of salary. How do we know he's not going to come around in a few years? I mean, Elgin's trade for Elton Brand turned out to be fantastic. His trade for Sam Cassell was pure genius. It was the single most brilliant and lopsided trade of the offseason. How do you get Sam and a #1 for Marko Jaric? So all of a sudden that erases what he did over the previous few years? Well, it certainly seems so! So maybe all Isiah needs is time! Actually, no, I think Isiah's still an outlier -- he's Enron -- but you get my point.

The comments to this entry are closed.

My Photo

Bio

  • I'm a writer for the New Yorker magazine, and the author of four books, "The Tipping Point: How Little Things Make a Big Difference", "Blink: The Power of Thinking Without Thinking" and "Outliers: The Story of Success." My latest book, "What the Dog Saw" is a compilation of stories published in The New Yorker. I was born in England, and raised in southwestern Ontario in Canada. Now I live in New York City.

    My great claim to fame is that I'm from the town where they invented the BlackBerry. My family also believes (with some justification) that we are distantly related to Colin Powell. I invite you to look closely at the photograph above and draw your own conclusions.

My Website

Books

  • What the Dog Saw

    buy from amazon

    Outliers

    buy from amazon

    buy from amazon UK

    Blink

    buy from amazon

    buy from amazon UK

    Tipping Point

    buy from amazon

Recent Articles

Blog powered by Typepad