payitforward wrote:1. Although sometimes there's a surprise in there (Bennett this year, for example), usually the top picks are obvious. So there's not much skill or pleasure in speculating about them.
Usually the very very top guys are obvious near draft day. It's definitely difficult to identify the best talent in the class earlier in the process. And its always difficult to sort the talent towards the middle and end of the lottery. Usually a class only has one or two obvious studs from the very outset and its been my experience that it's unusual to have a wire to wire consensus #1. Davis, Oden, and LeBron are the only recent examples I can remember.
I think it is pleasurable to discuss the top guys in the class even after their pecking order materializes. They are the future game changers who will make or break teams so they are the most relevant players in every class. It's interesting to me to follow the process of their development from before day one.
And yes I realize superstars and HoFers have come from the late first and second or been undrafted. Very very few in the history of the game. Few enough that making it your goal to identify the next one every year is a fool's errand IMO.
2. You don't mean it's impossible to predict success -- people do it all the time. You mean it's nearly impossible to predict success w/ a high degree of accuracy. But I disagree. Or, more accurately, there are ways to improve the quality of the picks you make at those spots.
The kind of predictions we tend to make about a prospect's future are impossible IMO. "This guy is going to be a future All Star/This guy will never be an All Star." It's a myth. Often our predictions are based on factors outside the player's control. When we're right, it's mostly a lucky coincidence. For one thing, we make lots of predictions because it costs us nothing. We don't have nearly enough information from where we sit to make well founded predictions.
We predict mostly based on past performances assuming the future will somewhat resemble the past. But for the most part, we're not even sure what the past was from where we sit. We have difficulty isolating the factors that led to certain results which may be relevant to future performance.
The other thing we try and do is isolate the tools each player has to determine whether or not he is capable of succeeding in x hypothetical NBA scenario. Even this is incredibly difficult and our info is limited to the point of making our evaluations relatively shallow. I'd venture that none of us truly has the expertise to properly evaluate and contextualize NBA level playing. We see a video of a player working out, hitting every single shot and think, "damn, he can shoot!" When maybe an NBA scout sees that from every player in every workout and better knows what information can actually be gleaned and translated into an NBA context from that.
So maybe we think we know what a player's tools are and we're flat out wrong. More often than not, this is probably true. Maybe we didn't see what the player was doing in practice and underestimated his tools. Maybe we overestimated his abilities based on limited sample sizes or limited anecdotal evidence. Maybe we didn't contextualize his performances properly or account for how they would translate to the NBA. Maybe we didn't acknowledge that player's ability to learn and develop wholly new skills. Maybe the players body changed in ways we didn't predict.
And the future doesn't always resemble the past of course. There are so many unknowables before draft day. Chief of which is what team will pick the player in question. How can you make even the most broad of predictions about a player's future without knowing that? "X player will definitely succeed/fail."
And even if we do correctly guess what team takes X player, we don't know what their plan for him will be on draft day, one year later, two years later, etc. I'd guess lots of teams don't even know what their long term plans for some of their picks are. What if the team's plan for developing that player is bad? There are players who have just as much talent to succeed as the next guy, but get drafted into the wrong situation. Some organizations are flat out incompetent and set players up to fail. Some change plans after the fact when a better option unexpectedly becomes available. Like most things in life, timing and opportunity matters just as much as talent/ability. How can you account for that in pre-draft predictions?
3. Teams who pick well in those late round 1 spots, especially teams in small markets, seem to have disproportional success when compared w/ teams who don't do a good job w/ those picks (ditto with Round 2 picks and undrafted guys as well), San Antonio being the poster child. To me, that indicates the importance of my point #2 -- increasing accuracy at those spots.
I'm not sure a team like San Antonio's success with late draft picks comes mostly from superior predictive ability. They're good no matter who they get. I actually think it comes mostly from superior organizational conduct. Meaning they make players work for them that other teams simply can't. I think they have such a good organization in place, with all of the right impact players and right coaches in the fold, that they get the puzzle pieces to work regardless of how well fitted they were before. Dallas was like this too--they literally won a championship with a team full of players other teams could no longer use effectively. SA isn't choosing from a secret list of players. They aren't working with secret data. When the guys they want to pick get drafted ahead of them, they move onto the next player on their board. Their system is just better suited for guys to succeed within in the majority of cases. A prediction I'd feel fairly comfortable making is the guy San Antonio picks is going to end up being more successful than the guy Charlotte picks no matter which names you throw up for each.
4. High Round 2 picks are some of the most valuable picks a team can have. Why? Because there is almost never a meaningful talent difference from the guys picked e.g. 27-30 and the guys picked 31-34. They are both pretty low % prospects overall, but the former group you have to guarantee for a few years, and the latter group you get much cheaper, give no guarantee, and usually can sign to a 2d contract earlier (and cheaper) than the former group.
Agreed. But again, I'd stipulate that their value is still very low relative to the high lottery players. The standards of what makes a successful late first/second rounder pick are clearly much lower than those at the top. And the impact those players have is almost never equal to the impact of successes and failures at the top.
5. You improve your accuracy by looking at numbers, and often there are more numbers for guys further down the draft: they are more likely to be 3 and 4 year college guys, not so likely to be 1-and-done. Now, numbers don't provide a guarantee of success; there are false positives (guys who put up good numbers in college but not in the NBA) -- but there are almost no false negatives (guys who put up bad numbers in college but good numbers as a pro). I say "almost", but the truth is I've been unable to find even one example of a false negative. If you can point me to one, I'll be grateful.
Well, where are you setting the bar for determining if something qualifies as a false negative? It's probably going to be somewhat nebulous because our predictions themselves are usually nebulous and flexible claims. For example, Beal is going to be somewhat of a false negative based on his college numbers IMO. They did not indicate how incredible a shooter he actually is. If you predicted Beal would be a mediocre shooter in the NBA based on his college numbers, he very well could shatter that prediction and be a false negative.
Beal certainly had the reputation of being a tremendous shooter though. If you went by that instead of his numbers, you got a more accurate sense of his actual shooting ability. I'd say Westbrook is also something of a false negative. There is no way you could guess the kind of player he'd eventually become just off his college numbers. 30 or 60 or even 90+ games just isn't big enough. NBA players can play nearly that many games, if not more, in a single season. And many players are still growing and changing so rapidly at that age anyway.
But speaking more broadly, I just don't agree with a numbers dominant approach to analysis. I disagree that we have the numbers to capture the events of any basketball game with satisfying accuracy and exactitude. I do not believe we have the numbers to isolate and paint an accurate picture of individual performances within the game. I think when we come up with numbers that happen to bear out predictions of future performance, we've usually just been lucky in finding non-causal correlations, mostly only bearing out over very large sample sizes--their predictive value remains in question throughout a players career.
6. There are plenty of guys in that late round 1 to mid Round 2 stretch who have put up outstanding numbers in college. Hence, if you shorten your odds by looking at numbers and you do what you can to increase your high Round 2 picks, you are likely to get some quite valuable assets. More likely than your competitors.
Those are my reasons. It goes without saying that our fearless leader EG does not value or optimize his Round 2 picks.
No matter what you do to improve your odds, they're still terrible by the standards of early picks. Second rounders are great contracts when they're singles and doubles, and spectacular when they're home runs. But the picks are still basically the least valuable team building assets out there. You're far more likely to get something of use from UFA dollars and early picks.
I think we fans often have unrealistically high expectations of the value of second round picks. No team regularly finds hits there. No team typically gives those picks the same chance to develop and be successful that they give high firsts--they don't value them highly. Saying Ernie doesn't optimize second rounders is probably true for every team. You've probably got about a 6 or 7 percent chance of getting a useful long term rotation player out of the second round. I'd bet even the most successful FOs don't get more than one in ten of their second round draft picks to reach even that fairly modest goal. Their odds of finding useful players in that range are better. But it's still only a very marginal advantage. That's why I don't enjoy toiling away over dozens of fringe prospects each year trying to unearth the next gem.
-- you do remember hoping we'd get him, don't you?
. But no there's really only one "I been saying for years..." dude around here. Fun for those who like to engage and argue, but occasionally tedious in their inability to admit flaws, or errors, or sense.
