A couple weeks ago, I took a look at Nick Young in my blog at the Post. Many disagreed with my conclusion that Young is fool’s gold and that the Wizards should move on.  Rook6980 at BulletsForever wrote this excellent piece countering mine, which is well worth reading. So, I looked deeper and came away even more convinced that the Wizards should not invest in him as a starter.

Young would be at his most efficient as a 3rd option. Here’s the problem, though — teams typically need their 3rd options to do more than score. However, Young doesn’t rebound, doesn’t pass, and he struggles to learn the playbook (he never learned the Princeton and it took him two offseasons, two training camps and a full season to learn Flip’s playbook).

But, Young supporters argue, he’s a good defender — just look at his on/off data and his counterpart defensive numbers. The guy has become a good defender.

Not so fast, though. College hoops godfather Ken Pomeroy ran an excellent experiment on +/-, which you can read here. To summarize, Pomeroy designed a player to have absolutely no impact on his team’s +/- then simulated 50 20-game segments. Those simulations yielded wildly divergent results — purely by chance.

That said, collect enough +/- data and get a big enough result, and eventually it means something.  Maybe.  Because by the time you get enough data to get results that are statistically reliable, you’re incorporating several seasons worth of data and you can’t be sure that what you’re getting is an accurate reflection of what the player is doing right now in his current context.

In Young’s case, going back three years teams him with Jamison, Butler, Haywood, etc. and has him being coached by Eddie Jordan and Ed Tapscott. Is that reflective of what he’s doing now?

So, what about those three straight years the team has been better defensively when he’s on the court? Surely that means something right? Maybe it does, maybe it doesn’t. The only result that would even begin to approach statistical significance was the 08-09 season when the team was 6.9 points per 100 possessions better defensively when Nick was on the floor.  “Better” is used loosely here — they allowed 117.9 points per 100 possessions when he as off the floor; 111.0 when he was on.  More accurately, they were less putrid defensively when he was in the game. It’s worth pointing out the team went 19-63 that season.

Last season, the difference was 1.3 points per 100 possessions, which most definitely falls solidly into the fluke range.  This season, the difference is 3.6 per 100, which STILL would fall into the “possibly fluke” category.  Combine the two seasons, and it’s only 2.4 points, which remains an iffy result.  The difference might be because of Young, but it also might just be a fluke.

And, even if the difference is because of Young, the data actually isn’t telling us that Young’s a good defender.  It’s saying that Young is a better defender than his teammates.  Those teammates being guys mostly notable for being bad defenders.

It’s encouraging to see his counterpart defensive stats steadily improving throughout his career.  However, counterpart automated counterpart numbers like the ones generated at 82games miss a lot because they don’t account for things like help defense, switches, zones, cross-matching, or traps.

Stats are extremely useful in evaluating players, and I’m a proponent of using them. But I think it’s important to spend time thinking through what the numbers are actually saying.  It’s also important to compare one set of numbers with the eyeball test, and with other numbers.  In this case, I don’t think the combination of inputs (stats + eyeball test) are saying Young is a good defender.  They’re saying he’s better than his teammates, which is a vastly different statement.

So, looking at it from a team construction standpoint, if the Wizards invest in Young as the starting SG, they have some constraints. They must have good rebounders at every other position — Young won’t help there. He gets balls that bounce directly to him, and that’s it. The team can’t significantly change the offensive system because of the time it’ll take Young to learn it. The team defense has to be designed without counting on Young to be a help defender because his defensive awareness isn’t good.

I could find a spot for Young on my team, but it would be in an off-the-bench role. A 7th or 8th  man, who’d be the first scoring option for a defensive-minded 2nd unit. Put him out there with a couple physical screen-setter types and a safe and steady ball-distributing PG. Let him run off screens and shoot the ball — but fewer long twos.

But, here’s the problem. With Young’s per game scoring average (which explains most of NBA salaries — virtually ALL NBA salary can be explained by per game points, rebounds and assists), he’s almost certain to be too expensive to be a 7th or 8th man. Under a capped system (that’s likely to get even tighter in the new CBA), it’s critical that players be paid at a level commensurate with their contributions.

Young getting starter dollars — even 3rd option starter dollars — moves him out of being a valuable scorer off the bench. At that kind of money, he’ll be overpaid for what he can do to help the Wizards win. If he’s a starter and a 3rd option, he doesn’t do enough else to warrant the money, and if he’s coming off the bench then resources that could be used to pay a legitimate starter are being used on a reserve.

Which brings me back to what I published at the Post. The Wizards should not invest significant dollars in Young. If they do, they’ll regret it.