Trading for, or signing an existing NBA player is making a decision on a known commodity. Drafting a player is completely different, they have no prior NBA performance to judge on.
Signing Grant Hill can't be a gamble in context of him being the best small forward in the league at the time. Orlando could not have known his career would be nearly derailed by a recurring medical problem and that he would nearly die. (Remember, it's not like he had any history of injury problems.) There was basically no gamble when they made the move, Hill had a badly sprained ankle he was recovering from during the offseason, and he was the dominant player at his position.
When you sign Kwame Brown to $8 million a year in 2005, it is a completely different situation than drafting him in 2001. Kwame had played in zero NBA games in 2001, by 2005 he had four years of data. Kurt Thomas practically averaged 30 and 15 his senior season in college. When the Knicks resigned him in 2004, and when the Suns traded for him a year later, they had ten years worth of data on how he performs in the NBA. On the other end of the spectrum, Michael Redd's college career was pretty unremarkable.
Of course there is a "gamble" in any transaction or change. The point is, draft choices are almost entirely crap shoots as you have no idea truly how the player will perform in the NBA. When you are trading for that same player five years later, you do.
When you evaluate a GM's decision, of course we use some hindsight. In a draft choice, we know 100% more than the GM knew at the time, we know how the player performed in the NBA. With a trade or free agent signing, we can evaluate the context in which the transaction was made with an understanding of the knowledge that was available to the GM about the player.
Just evaluating trades based on hindsight is boring and ignores the nuances of context. Take the Dirk Nowitzki-Robert Traylor deal cited in this thread.
Pitino assured Nowitzki that he would draft him with the Celtics' first-round draft pick on position 10. However, his plan was foiled by Nelson, whose Mavericks had the sixth pick. Nelson worked out draft day deals with the Milwaukee Bucks and the Phoenix Suns: the Mavericks wanted Nowitzki and Suns reserve point guard Steve Nash; the Bucks desired muscular forward Robert Traylor, who was projected to be drafted before Nowitzki; and the Suns had set their sights on forward Pat Garrity, who was projected as a low first round pick. So, the Mavericks drafted Traylor with their sixth pick, and the Bucks selected Nowitzki with their ninth and Garrity with their 19th pick. Then, the Mavericks traded Traylor to the Bucks for Nowitzki and Garrity, and they in return traded latter to Phoenix for Nash.
Of course, years later we can state the obvious and say the Bucks made a bad decision by wanting Traylor. But was the trade itself really a bad decision? The Bucks wanted a player, he might not have been around, so they dealt to get the guy they wanted. Was Steve Nash for Martin Muursepp, Pat Garrity, a guy named Bubba and a first rounder also a bad decision? Especially when you factor in they made the trade partly because Donnie Nelson was friends with Nash?
We need less hindsight and more context. As evidenced by the context of the Kobe draft pick completely changing ones judgement on what on face looks like a disastrous decision.
