Kevin Myers- FOX 9 Creative Services Manager
"How do meteorologists keep their jobs when they are wrong half the time?"
"It's the only profession where you can miscalculate so consistently and still remain employed."
This is the tired "joke" we've all heard, and some of those who repeat it aren't really joking.
This sentiment, of course, is wrong on several levels. While forecasts are often off in a range of about five degrees, this variance can only be considered "wrong" by irrationalists expecting unattainable perfection. Furthermore, any sensible meteorologist usually refrains from saying "it absolutely won't rain" or "I guarantee a downpour", except in the most extreme circumstances. Almost all forecasts come with a percentage chance of precipitation, giving you a sense of the potential risk.
It is more accurate to say that meteorologists get the forecast right 75-80% of the time. In my book, this is an acceptable rate for predicting a future outcome using a science-based method.
It would also be an acceptable rate for NBA general managers to make good draft picks. However, every year, as a group, they fall short of this success rate. A generous approximation of draft "success" rate would probably be closer to 50%, depending on how you define success. Expectations decrease the further you move along in the draft, with little to no assurances as you move into the second round. However, even the lottery is littered every year with future busts who are identified by many as "can't miss" players and compared to All-Stars and Hall of Famers.
Why does this happen? Of course, there is the human factor. We develop in odd and unexpected ways, with our better and worse selves often rising and falling sporadically. This is not something a talent evaluator can be expected to consistently predict.
The biggest culprit for the poor talent projections, however, is an old school mysticism that still pervades many front offices. For decades, talent evaluation has been dominated by former pro players, protected by the conceit that if you didn't play the game at the highest level, you don't possess the secret knowledge that allows you to identify the hidden truths about a prospect's potential. While there is some credence to this, it has also been inflated to ridiculous heights, used as a country club codeword by the elite to retain their privileged position. If there were some other way to predict the future besides rare, lived experience, the roundball oracles who get it right half the time would not seem so precious and need to be retained at high compensation levels.
There is another method for evaluating athletic talent, and it's become so mainstream that the general public has at least a passing knowledge of it. Many call it "Moneyball" after the titular movie, which is actually not a very accurate name for it. It doesn't have much to do with "money", but it has a lot to do with numbers. Every year, more sports teams continue to embrace advanced statistical analysis. Several years ago, 16 of the 30 NBA teams had at least one employee crunching numbers for analysis purposes. Of those 16, 14 of them made the playoffs. Do you think maybe there's a connection?
There is still resistance within many front offices to analytics as a valid (and even equal) form of talent evaluation. Much of this resistance comes from the fact that the old school simply doesn't understand it, and we often fear and dismiss that which we don't understand. Additionally, if this method is validated, it diminishes the "rare skill set" that the privileged few have been profiting from.
This brings us to my beloved Minnesota Timberwolves. They recently dismissed the cock-sure and clueless David Kahn. Mr. Kahn did not possess first hand experience playing basketball at high levels. And, he did not embrace (or even, apparently, even understand) statistical analysis as a means of talent evaluation. The result was predictable. Kahn possesses one of the worst draft track records in the history of the league. He had Ricky Rubio fall into his lap for him to take with his first draft pick, a choice that the worst of GMs couldn't have screwed up. He then proceeded to miss on virtually every other pick he made during his reign, and in every conceivable way.
The failures wouldn't have been so painful for me as a fan if they weren't so obvious at the time they were happening. Instead of the pain of hindsight and regret, Timberwolves fans with even a modicum of interest in draft prospects and access to minimal information in the public domain could have (and did) predict his failures in real time. It was a sports fan horror show, conducted by a smug charlatan who thought he was the smartest guy in the room.
You may question whether Kahn's failures were really so apparent at the time of their execution. However, a Timberwolves fan website like CanisHoopus.com consistently came to group consensus among its users as to who the team should select and would have chosen a better option (if not the best option) had they been in Kahn's chair. It's all written in the archives of the internet for you to review. Curry over Flynn. Keep Lawson. DeJuan Blair over Ellington. Cousins over Wes. The Pacers will trade the #10 pick for Flynn? Great! We can take Paul George at 10 and then go with either Cousins or Monroe at #4. Derrick Williams? Sigh. His success at the college level was predicated on a small sample size of hitting three pointers, which is a skill-set that doesn't reliably translate to the NBA. If he can't succeed from 3 point range in the pros, he will be little more than average. Probably not worth the #2 pick.
It was all said and agreed on by a group of intelligent, but common, NBA fans. How did they arrive at these conclusions? Outside of sporadic viewing of college games and highlight reels, several of the users on the site developed their own mathematical formulas for player evaluation. Over the last four years, these models have gone from crude to fairly refined and are now based on regression models that when applied to previous drafts give outcomes that would have outperformed the decision makers at the time. All of this was accomplished by accessing info that can be found in 10 minutes on the internet.
Luckily for these fans, Kahn has been shown the door and there is a new/old sheriff in town. The Timberwolves have brought back "one of us" in Flip Saunders, a member of the secret society of Eye Test Experts. Flip once looked up and down a high school kid with no college experience and said "that guy, I'm drafting that guy." His success in identifying the potential in Kevin Garnett bought him a lot of credit at the Talent Evaluator Store, credit that his co-conspirator in that first pick, Kevin McHale, proceeded to mostly scorch over the next 12 years.
We are left with the question of how much of a role Flip played in McHale's decisions. He obviously had input, but was he involved in pegging Ndudi Ebi, Paul Grant and Will Avery? We don't really know.
I have full confidence that Flip Saunders knows more about basketball in his pointer finger that he so skillfully spins a ball on than David Kahn has ever and will ever know. But, that is a very low bar that anyone who has read this far into this article could probably meet.
What we are about to find out is if Flip is a member of the 50% Failure Club, or if he can join the ranks of front offices like San Antonio, Oklahoma City and Houston who consistently identify superior talent, even deep into the second round. The biggest hindrance to Flip proving himself elite is that he is only using half the tools at his disposal. He may be able to size up talent and have a real eye for it, as he supposedly proved in his days of running CBA teams. But, his adversaries have their own Eye Test Experts, along with departments full of mathematicians developing statistical formulas that will compliment the traditional scouting report. The best NBA offices use both methods hand in hand, using one to check against the other and vice versa.
I actually sent a tweet to Flip about the importance of statistical analysis and, to his credit, he responded.
Kevin Myers @MyFansanity25 May
"@mcuban: Analytics Is Not a Strategy http://dthin.gs/13M0Dzj " @Flip_Saunders please, PLEASE read this- Analytics use in player acquisition.
Flip Saunders @Flip_Saunders25 May
@MyFansanity analytics gives confirmation and information. Totally agree!
First of all, huge props to an NBA GM who takes the time to respond to fans on Twitter. It is very endearing and an amazing contrast to David Kahn's "methinks" attitude.
I was initially heartened to see Flip apparently embracing analytics. But, as I began to parse the tweet, I came to suspect that he was saying that the role of analytics is to confirm what his eye is telling him. This is not how it should work.
Subsequent public statements from Saunders have solidified my fears, revealing him to have an awareness of analytics, but not a realization that they serve as a valuable check against the eye test. Results can expose areas where your eye may be overwhelmed by the sizzle of a player, causing you to miss the fact that there is no steak. There are benchmarks that a player can fail to meet (such as 1 steal per 40 minutes) that have historically correlated with failure at the next level. If a player reaches certain aggregate statistical levels, there is a good chance that he will succeed in the NBA if given the chance, even if he isn't projected by the "experts" to go in the second round (see Jeremy Lin).
If Saunders doesn't at least recognize that he needs to bring in an analyst to compliment his old-school eye test skills, the Timberwolves will likely be doomed to mediocrity, at best. Rumors have it that Flip may hire Milton Lee, an executive with the Brooklyn Nets, to serve as his right hand man. Lee is a former stock trader, and sounds like exactly the kind of forward-thinking analyst who will utilize every form of research to maximize the returns on his investments.
However, this won't happen until after the draft, if at all. For now, heading into another excellent opportunity to elevate this seemingly cursed franchise out of the longest playoff drought in the league, we are relying on Flipnosis. Saunders is the equivalent of the guy who says he can tell you whether or not it's going to rain based on a feeling in his bones. There are a lot of people who operate on feelings like this. Some of them can actually get it right most of the time. Many of them don't. Is Flip in the minority of prognosticators who get it right more often than not? We can only hope.
Meanwhile, the majority of teams are using their old guys with feelings in their bones in conjunction with a team of modern "meteorologists" who are utilizing historical trends, scientific data and advanced mathematical projections.
Who do you trust to get it right more than half the time?
I have my own forecast for the 2013 NBA draft. Here are a few things to keep in mind as we head into my analysis.
"Success" is relative to where a player was drafted.
Picks 1-7: Expected to become stars, but the more realistic expectation is that they turn into at least above average starters.
Picks 7-14: Eventually become starters for at least two seasons.
Picks 15-30: Must receive regular minutes on a team for longer than the length of their rookie contract.
Picks 31-60: Must make a roster and contribute in any meaningful way.
While I've been critical of the failure rate of many general managers, expectations for player performance should be moderated by the reality that on average only nine players from any draft ever serve as starters for two seasons or more. While general managers could be making better decisions, the unstoppable hope and optimism of the draft also puts unrealistic ceilings on fan's projections of a player's potential.
Here are my tiers, based mostly on statistical models by "vjl110" and "Madison Dan" at canishoopus.com, James Brocato at shutupandjam.net and Ed Weiland at hoopsanalyst.com. Any variance from the results of their models is based on "the feeling in my bones" I got from watching video scouting reports on DraftExpress.com.
These predictions don't take into account the advantage high draftees receive in being automatically given playing time and extra attention from coaching staffs.
The percentages next to their names indicates my forecast for the chance that they "make it rain" in their respective categories.
1. Otto Porter- 85%
2. Nerlens Noel- 82%
3. Victor Oladipo- 72%
4. Cody Zeller- 63%
5. Trey Burke- 54%
1. C.J. McCollum- 83%
2. Kentavious Caldwell-Pope- 81%
3. Ben McLemore- 65%
4. Steven Adams- 58%
5. Nate Wolters- 55%
5. Sergey Karasev- 53%
6. Rudy Gobert- 51%
1. Kelly Olynyk- 83%
2. Michael Carter-Williams- 78%
3. Mike Muscala- 76%
4. Giannis Adetokunbo- 74%
5. Jeff Withey- 73%
6. Lucas Nogueira- 69%
7. Shane Larkin- 67%
8. Reggie Bullock- 65%
9. Dennis Schroeder- 65%
10. Tim Hardaway Jr.- 64%
11. Jamal Franklin- 64%
12. Glen Rice- 62%
13. Tony Mitchell- 59%
14. Allen Crabbe- 56%
15. Gorgui Deng- 54%
16. Andre Roberson- 52%
17. Arsalan Kazemi- 51%
NOT WORTH THE POSITION THEY'LL BE DRAFTED AT
1. Alex Len- Historically awful steal rate projects him as the next Olowokandi.
2. Anthony Bennett- Tweener forward with injury and weight issues. Who is his success model?
3. Shabazz Muhammad- High volume, inefficient scorer with dreadful assist rate, bad attitude.
4. Mason Plumlee- Didn't improve throughout college, didn't stand out in any one area.
1. Ricky Ledo- Ineligible in college. Very talented, but just as troubled?
2. Archie Goodwin- Youngest player in the draft. Can get to the rim, but can't shoot. Could he develop?
The consensus seems to be that if the Wolves stay at #9 that they will likely select Kentavious Caldwell-Pope. I've been a KCP been for over a year, when an observant poster at Canis Hoopus first pegged him as a possible fit for the Wolves. He didn't come out after his Freshman year, and then improved dramatically as a Sophomore, making him an even more attractive target for Minnesota.
Don't discount, however, Flip gunning for Best Available Talent and taking Cody Zeller at #9. Although Flip saw Zeller get destroyed by Trevor Mbakwe on the same floor he was once a General on, he also watched Zeller arguably have the best season of anyone in college basketball. I have Zeller in a higher tier than KCP, and Flip might as well. Keeping Zeller would only add to the log jam at power forward, so you can expect that Derrick Williams would be on his way out the door if this happens.
Flip could also select Zeller with a trade already lined up. Charlotte GM Rich Cho is reportedly high on Zeller, but his boss (MJ) doesn't want to take Zeller that high. A trade down where Charlotte receives Zeller (at #9), #26 and Williams would be great value. If I could gamble on this scenario and received favorable odds, I would bet on it. For this to happen, Charlotte would likely need Oladipo to fall to them at #4, something that seems less likely by the hour. However, if Oladipo is selected by Charlotte, get excited Timberwolves fans. He is not a good fit with Charlotte small forward Michael Kidd-Gilchrist (both are defense first, offense distant second). It would indicate that Flip is moving up to secure his man-crush, Victor Oladipo.
I'm just an overly-invested fan with an unusual obsession for the NBA draft. My projections may or may not prove to be more wise than how the draft order actually unfolds. I hope that Flip Saunders can outperform my internet-harvested talent evaluations. I hope he is either a truly gifted gut-instinct guy who can make it rain on draft night, or that he realizes the "other half of the game" and brings in a teammate who can set him up to make smart shots. If he can be "wrong" in his draft picks as infrequently as your average meteorologist, his time running the Timberwolves will be a success.