The Harris-Wall Conundrum: NBA Trade Mechanics Through A Formulaic Approach

Once it became apparent that the New Jersey Nets were the single worst team in the National Basketball Association and that John Wall was going to be the number one overall pick by light years, the questions immediately began.

If they do land the overall pick, do they trade Devin Harris?

When Harris hasn’t been in a suit due to injury this season, he has watched his performance plummet on the worst team in the league. Points, PER, assists, rebounds, all have dropped dramatically. Given the fact that Wall is considered by many (including people who aren’t completely delusional like myself) to be one of the best picks of the last decade, even with only a 25% chance of actually winning the lottery, the seeds for moving up a Harris-move timetable were planted.

That line of thinking started evolving once it became apparent that the Nets were broken beyond repair and it was worth considering a complete blowup.

It’s a minefield. Trade him too early and you risk ending up without a point guard next year. Ask the Sixers how that’s working out. Even if they were able to get back some reasonable talent for Harris, that could potentially complicate their summer free agent plans. Then again, waiting till after the lottery could result in Harris’ value dropping between now and then. Keeping the two of them is simply too tricky of  a situation, not to mention like hoarding beef jerky in the desert instead of trying to find water.

The complex nature of the problem has created conflicting reports, with the latest outlining a full recline from aggressive trade talks, if there was ever considerable intensity in the first place.

But exactly how do you figure this kind of a question out? What are the parameters that need to be considered and ignored? How can we approach this in a more subjective manner? My first thought was to attack the question using a game theory model. After all, a trade by necessity involves interaction with competition in a marketplace, revolving around a primary decision. But I needed help. A lot of help.

With a big ol’ hat tip to Henry Abbott, I contacted Brian Tung of the advanced theoretics blog The Null Hypodermic and asked him for help with a conceptual model. I gave him the following elements to consider.

• The Nets will undoubtedly finish with the worst record in the league, giving them a 25% chance at winning the lottery. John Hollinger’s adjusted probabilities actually put it at 25.5%.
• Wall’s value is more than just being widely considered not only the best prospect in the draft, but according to many (including myself) the best prospect to come out of college since Dwyane Wade.
• Harris has shown a significant downturn in production this season.
• Trading Harris lowers the Nets’ payroll, assuming they trade for assets of lesser or expiring value, which is a premium for them.
• At the same time, trading Harris costs them a valuable young veteran point guard, which could be an incentive for free agent prospects this summer, their ultimate goal.
• There are no other outstanding point guards in the upcoming draft, which would make losing out on the #1 and falling to 2-4 disastrous without Harris.

Having provided Brian with the basic elements with which to construct an outline, here’s his response:

From the outset it must be acknowledged that any model will leave out all kinds of things.  It is best therefore to be clear about what we’re leaving out, so that they can be addressed outside the model.  Also, I’m writing as I think (and vice versa), so this may only be half-baked.  I’d have to give it some more thought to bake it outright.

One difficulty in modelling most of these kinds of decisions is determining how to value basketball assets.  Advanced statistics like PER and APM have improved matters somewhat, but as the dispute over them demonstrates, we still have a long way to go.  But a more significant difficulty is determining how to compare basketball assets to fiscal assets.  Since the Nets are probably going to get the first pick with a 1/4 probability whether or not they trade Harris, this decision simply cannot be made on the basis of basketball assets alone.  Fiscal assets must be taken into account, and they must somehow be valued on the same terms as basketball assets.  I’m going to assume without evidence–either for or against–that the Nets are able to do this.

I’m also going to assume for the time being that the Nets will not get Wall unless they get first pick.  I’m not sure how valid that assumption is; Minnesota could easily get the first pick and it’s not clear to me that they would pick Wall, or that if they did pick Wall it would be with the intention of keeping him.  Of course, if they did pick him to trade him, they might want to trade him for Harris, so if the Nets have already traded Harris away at that point, it might conceivably diminish their chances of getting Wall via the trade.  But for now I’ll assume that’s not a viable avenue anyway.

If that’s the case, we’re not really talking about game theory because the Nets are not competing against anyone other than themselves (irony duly noted).  There are four possibilities:

1. Trade Harris, then fail to pick up Wall in the draft. The Nets gain some fiscal assets (call it F), but lose significant basketball assets (call it H).
2. Trade Harris, then pick up Wall in the draft. The Nets gain fiscal assets F (might be a little less, depending on how closely we’re modelling fiscal effects), and basketball assets are about a push (you gain W – H).
3. Keep Harris, then fail to pick up Wall in the draft.  Fiscal assets and basketball assets largely unaffected.
4. Keep Harris, then pick up Wall in the draft. Fiscal assets mostly unaffected (again, like 2, might be a little reduced), but gain basketball assets along the lines of (H, W) – H.  This mumbo-jumbo basically means that having both Harris and Wall is worth more, basketball-wise, than Harris alone, and the difference between them is what you gain.  Because they can’t play point at the same time–not in the usual sense, anyway–this difference will, however, be less than Wall’s worth on his own.

So, if you trade Harris, you gain fiscal assets F and “gain” basketball assets worth W/4 – H; we divide W by 4 because there’s only a 1/4 chance of picking up Wall, but you always lose Harris’s worth if you trade him.  If you keep Harris, your fiscal assets remain basically unchanged, but you gain basketball assets worth [(H, W) - H]/4.  If you can somehow place F, H, and W all on the same scale (i.e., compare fiscal assets to basketball assets), you can then compare your gains from the two options:

F + W/4 – H vs  [(H, W) - H]/4

Do a little algebra and you can convert that into

F vs  [(H, W) - W]/4 + 3H/4

Let’s keep things straight now: the left side is the “trade Harris” side, and the right side is the “keep Harris” side.  Now, just as adding Wall to Harris gains you less than Wall’s worth on his own, adding Harris to Wall would also gain you less than Harris’s worth on his own.  We can represent that as

F vs  (H – a medium-sized bit)/4 + 3H/4

or just

F vs  H – a little bit

The “medium-sized bit” basically represents the cost of having only one basketball to share amongst the two, in case both Harris and Wall are on the team; the “little bit” is just one-fourth of that cost.  What does this tell us?  It tells us that if you feel that you can in a sense break even on a Harris trade (i.e., gain fiscal assets equivalent to Harris’s basketball worth), you go ahead and pull that trade.  This may seem a little odd because it’s (to first order) independent of Wall’s basketball worth.  There are a few counterpoints to that observation:

• Harris’s basketball value includes, as Matt says, his ability to attract free agents in the summer, in addition to whatever value he brings strictly on the court.
• The “little bit” depends on how good Wall is.  If he’s a draft dud (unlikely, but it could happen), it’s essentially zero.
• On the other hand, if he’s amazing, as in a significant upgrade from Harris out of the box, then that little bit in the final comparison could be one-fourth of Harris’s worth–possibly including Harris’s ability to draw free agents, if Wall is so good that free agents would want to play with him, too.  Of course, we wouldn’t really see Wall’s performance until after this summer’s free-agent bonanza.  (Frankly, I’m not sure it’s going to be that much of a bonanza, but we’ll see.)

If the last point is true–if you are certain Wall is that amazing–then you pull that trade if the fiscal assets you gain are worth even three-fourths of Harris’s basketball value.

I must emphasize that this cold-blooded analysis leaves out factors that may be of great importance.  For instance, I’ve assumed that the cost of losing Harris is limited to on-the-court value.  In a sense, that is true, but it discounts the psychological cost of being a historically terrible team, and the fiscal cost of diminishing gate receipts.  You could probably account for the latter, but the former is something you sort of have to feel with your gut.  Also, I’m not convinced that Harris’s downturn at this stage of his career is much more than a blip.  He could get better next year, easily.  Then, too, it might be a result of playing on a crappy team; for a ball-handler, that’s sort of a minus, statistically speaking.  If he stays with the Nets and draws some free agents, his own performance could improve as a result.  On the other hand, Wall represents youth.  If by some chance the Nets could surround him with some established talent, they could be playoff contenders at some level for years to come.  There are probably quite a few other factors that I’m not thinking of at the moment.  Probably you can come up with more.

As I said, not wholly baked.  Hope this gives you some idea of the kind of analysis that could be done, though.  A more thorough analysis would be much more time-consuming, of course.

Here’s a graphic representation of our decision choices modeled:

While obviously a base of a soup and not what you’d ladle up for paying customers, Tung’s analysis does give us a reasonably stout set of elements to consider. First and foremost is the effect of the unstable probability of Wall in the current atmosphere has in simplifying the Harris trade value to a monetary one. Given the possibility of Wall on the horizon, even at the relatively low chance they currently have, if they find the right economic model to offset Harris’ value, it would make sense under this structure.

While Harris is undoubtedly an asset not to be traded brazenly, there is some support for this kind of an approach. Jettisoning Harris provides even more freedom for what is already one of the most unburdened teams in the league come July, provided they can avoid taking on long contracts in return. And the value of Wall, determined by New Jersey’s scouts and management, is off-set by the Lottery and the devastating effect a premature trade would have on the franchise if it were to miss out on Wall.

There’s more to this question than simply “How desperate are you?” and hopefully in the weeks to come, we can shed more light on it.

Thanks to Brian Tung for his help with this post, as well as Henry Abbott for lighting the way.