Wednesday, February 11, 2015

The Eye Test Debate

And like I say, the Rockets for a long time. So they went out and paid James Harden a lot of money — they got better. Then they went out and got Dwight Howard — they got better. They had Chandler Parsons, and now this year, they went out and got Ariza. The NBA is about talent.

Charles Barkley made a big splash last night when he called analytics people nerds who didn't get a lot of girls in high school, but he did have a point tucked away under all the name-calling and vitriol. He was really only saying something we can all agree on - the teams with the best talent and the best coaching staffs are going to win the most games. That is the case now, that has always been the case and it will always be the case. If there's a debate, it's about the methods by which teams can acquire talent.

After all, you don't need analytics to tell you Dwight Howard and James Harden are good. You can look at a box score and figure that out. Sir Charles discounts all the work the Rockets had to do to get those players but he's right in that you can give most of the credit for Houston's success to Dwight and Harden. Once you get foundation guys in place, all a front office can do is work around the margins and try to put them in a position to succeed.

Where the Rockets have been ahead of a lot of teams is in how they have used analytics to identify guys who aren't Top 3 picks. The only eye test you needed to figure out that Dwight Howard and James Harden were going to be good was do my eyes work? Where the rubber meets the road for "eye test" vs "analytics" debate is guys like Beverley and T. Jones.

Personally, I don't really need any statistics to tell you that Patrick Beverley and Terrence Jones are good players. I can watch them play the game for awhile and figure that out.



Jones has got good size and length for a PF and he's a freak athlete who can play above the rim and get down in a stance and slide his feet on the perimeter. He's a really good ball-handler for a guy with his size and he can even push the ball himself after a break and create a shot for one of his teammates. If you give him a lane to the rim, he can put the ball on the floor and then finish over the top of your center. His jumper isn't great but it's getting better every season and it puts defenses in an impossible bind. He's more of a face-up guy than a post-up guy, so the better his shot is, the higher his ceiling as as player. On top of all that, he can fight on the boards, protect the rim and get his hands into passing lanes. He's still a really young player so I don't hold his defensive lapses against him too much, but it is something he will need to improve to be a 30-35 minute player on a good team.

I look at Terrence Jones and I see a guy who should be a high-level two-way starter at his position. I'm really not sure how many people on Basketball Twitter would agree with me and I don't particularly care. The point is there is a right answer and it really doesn't matter how you get to it. If you need per-minute stats, on-court/off-court numbers and PER to convince you that Jones is a good player, more power to you.

Here's the thing - it's easy to say I'm an eye test guy but if all those NBA front offices are such great eye test guys, how come Terrence Jones fell all the way to No. 17 in the 2012 draft? How come the Rockets needed analytics to tell them what should have been obvious to everyone?

The eye test guys spent most of Jones sophomore season at Kentucky quietly bad-mouthing him and insinuating that he had attitude and confidence problems? Why? Because his per-game stats were down across the board from his freshman season. Apparently none of them were able to put together the connection that adding Anthony Davis and Michael Kidd-Gilchrist to your college front-line (in place of Josh "Jorts" Harrellson and Darius Miller) was going to decrease the number of opportunities you would have to rack up points, rebounds and assists.

Jones was just as big, just as skilled and just as athletic as he is now, but most of the "eye test guys" were really just scouting the stat line. Their eyes saw that his points, rebounds and assists were down so they made up some risible just so stories about his game.

This is the 2012 NCAA championship game between Kansas and Kentucky, where Jones was matched up with Thomas Robinson. Knowing what we know now, I defy you to watch this game and think Robinson was a better player. He scored 18 points - but he needed 17 shots to do it! Jones, because he had a smaller role on his team, had only 9 points, but he only needed 7 shots to get it.



Yet if it was obvious then how come Thomas Robinson went 12 picks ahead of Jones in the draft and no one said anything? That's because when most people say eye test they are really just saying I saw this guy take a lot of shots and then I looked at the box score so I guess he must have had a really good game. Thomas Robinson passed a lot of people's eye-test but the problems he had in NBA were evident when he was in NCAA if you knew what to look for. Instead of feeding himself a bunch of self-serving lies about his own intelligence, Morey ran the numbers and found a diamond in the rough.

It's the same thing with Patrick Beverley. After the fact, it seems pretty obvious that a guy with his size, athleticism and defensive intensity and his ability to knock down spot up 3's would be able to find a spot in the NBA. Just about every team in the NBA needs someone who can ball-hawk Russell Westbrook and then stand in a corner on offense. Then how come it was the Rockets who found Beverley out of Europe? The same month Houston signed him, the Mavs signed Mike James to shore up the PG position.

I don't watch enough European basketball to know for sure, but I kind of doubt Patrick Beverley was the only guy to slip through the cracks. Look at what just happened with Hassan Whiteside! The world is full of tall, athletic guys who have played enough professional basketball to where they could handle the transition to the NBA. The problem is that most of these eye test guys in the NBA don't watch a ton of Euro ball either so they never know who to sign. I'm guessing the Rockets ran Beverley's numbers in the computer, liked what came out and then went out and scouted him.

People want to say Houston is this super genius analytic team but they are really just taking advantage of basic inefficiencies in how other teams evaluate talent. Remember when everyone was freaking out about their bench? Daryl Morey and Co. knew there was plenty of talent floating around the NBA and they were confident they could find it. Why? Because they have some analytic models they trust which can spit out answers.

I'm a person with very little interest in math (I have a college degree yet can do nothing more complex than long division) yet I like most of the personnel decisions the Rockets make. That's because there are different ways of getting at the right answer, just as there are in every other walk of life. All either of the eye-test vs. analytics debate is yelling at each other about the best way to re-invent the wheel.


The irritating part to me about so much of the analytic discussion is the seeming obsession everyone has with using formulas to rank things. If we have 400 wheels, can't we assume that wheel #345 and wheel #367 are fairly identical? What exactly is the difference at that point?

Every article you read these days seems to about whether Player X is Top 5 at his position or Top 10 or whether he should get X amount of money if Player Y got Z amount. It's a bunch of accountants arguing about actuarial tables who have convinced themselves they can find the meaning of life if they can crunch the numbers hard enough.

If we can all agree that Terrence Jones and Patrick Beverley are good players, let's try to come to some interesting conclusions. How much should another team value Jones, given the smaller role he has had in Houston? Is a young guy better off growing into a small role on a good team or being thrust into a big role on a bad one? Is having a guy with Beverley's defense at the 1 and running more offense through the 2 and 3 a more efficient way to build a team? How many guards in the European leagues can play defense like him? How can we find out?

The only problem with letting "stats nerds" write about basketball is that they tend to be very well-educated so they bring a lot of the problems with academic writing with them - the needlessly pedantic debates about issues of no practical relevance to anyone in the real world. That's what John Roberts said about most law school writing these days.

Instead of sitting around all day and arguing about the best way to find the optimum wheel, let's string a bunch of wheels together and try to go somewhere.

2 comments:

  1. That quote is by Juelz Santana not Cam'ron, cmon dog

    ReplyDelete
  2. This comment has been removed by a blog administrator.

    ReplyDelete