• Changing RCF's index page, please click on "Forums" to access the forums.

15 Year NBA Draft Analysis

Do Not Sell My Personal Information

Matches

Eloquently sarcastic
Joined
Oct 13, 2009
Messages
544
Reaction score
708
Points
93
I've been on record here a couple times as steering away from the "blow it up" model, also known as "Look at the Thunder! Let's just do what they did!" I point out the Memphis Grizzlies who had similar W-L records over a similar stretch and instead of coming out with a future MVP and several nice pieces around him like the Thunder, came out with Mike Conley, OJ Mayo and Hasheem Thabeet.

But I thought I should get some hard data on the draft. What is the success rate of picking in various locations? If we tank get a top 4 pick, what's the likelihood of that player becoming an all-NBA talent that we can rebuild the organization around? If we're in the high lottery (1-7), what are the chanced that the guy busts anyway and never even starts more than 80% of his games? How frequently can you get solid players in that 14-20 range? I didn't want to just spout assertions and pull anecdotes. I wanted hard data. And I didn't want to hoard it.

So here it is. (The link is to the company gracious enough to host the image)

Data is from the 1994-2008 drafts (since I could not adequately gauge success rates yet for 2009 or 2010). The criteria are this:

Gray line: Played in 100 NBA games or more, or will break that mark this season, or would have if he had not gotten into a motorcycle accident, or is named Greg Oden. To not succeed here is considered a complete bust.

Green lines: MPG for career. The dark green is players who averaged or have averaged at least 16 mpg, the light green mark is 22 mpg. I chose these figures because I consider that a player who does not average 16 minutes a game must be considered a fringe NBA player and anyone drafted in the lottery who misses that mark is a bust. In the low lottery, even this low mark is only about a 75% proposition. The 22 mpg mark represents for me someone who is at the very least, an important bench contributor.

Red lines: The next tier up in success metrics: percentage of games started. I wanted to see players who were borderline starters (50% or more) and fairly regular starters (80% or more). The scariest stat here is that even drafting #2, you're only slightly above a 60% chance of getting a true NBA starter.

Blue lines: All-stars. Dark blue line is players who've been all-stars at least once. Light blue is players who made All-NBA teams at least once.


My quick takeaways:

Please don't say the draft isn't luck. Once you get outside the top 5, you only have a 50% chance of drafting a guy who's simply a starter. That gives credence to the "Tank Now!" camp. On the other hand, if you do "Tank Now", remember that you only have around a 22% chance of getting the #1 pick, and even if you get it, you're more likely to hit on Red or Black in roulette than getting a truly elite talent. And in the top 4, it's only 46.6% that your pick will make a single all-star game.

That sound you hear is me, in the corner, in the fetal position, crying myself to sleep.


Esoteric note: The graph here does not represent actual percentages. Since 15 years is not a very large sample size, I normalized the graphs to smooth them out a little. The vagaries of why a player was chosen in a specific slot are not as important as the general trend. For example, for whatever reason, players drafted #11 or #12 over this 15-year stretch have been HORRIBLE. It's a litany of suck and you can still see the dip in the graphs. But this graph smooths things out somewhat - for example the data for the #18 spot is actually an average of the actual 18th spot results, the results from 17 thru 19, and the results from 16 thru 20. Since there's nothing particularly special about the #18 slot - a player picked there could have easily been taken a couple spots earlier or later. This is also why the #30 spot jumps up. Because I was lazy and did #31-36 as an aggregate, I couldn't actually to the average of #28-32 for the #30 spot -- shown here is the actual success rate for #30.
 
Can we add a PER and a OPP PER line to this graph?
And also would be interesting to see where teams are picking most over that timeframe.
Are 7-8 picks bad because the Clippers kept landing there?

My initial takeaway: It's good to draft 1-6, don't take the risky euro, white, or young player 7-8 instead take the sure thing that is at 9-10.

And trades for picks that fall in the 17-22 range, Sleeper players a plenty.
 
I would say that writing everything off as "luck" is taking the easy way out. There is no doubt that the Chris Wallace, the Kevin McHales, the Mike Dunleavy's, etc. of the world have completely blown their drafts. However, was it bad luck or bad talent evaluation? Probably a combination of both.

The problem here is that there is luck in every aspect of the game. Those promoting the Pistons model have to realize how insanely lucky the Pistons were in winning that title. Will a player of Rasheed Wallace's caliber be available for pennies on the dollar in the one year you need him the most? Where was our Rasheed last year? We really could have used a prime Rasheed last year... Ultimately, when you let all of the percentages play out, if you ever want to win an NBA championship, you want to take the path that gives you the most UPSIDE, and the draft, especially for a market like Cleveland, is that path, it's the only real path.
 
Can we add a PER and a OPP PER line to this graph?
If by "we" you mean, "Matches"...not likely. :tongue: This was a Saturday afternoon and almost all the stats I used were on the lovely draft history pages at basketball-reference.com. To go into each player and lookup their PER would be ... another Saturday.

Another striking thing is how relatively flat the lines are from about 12 through the late 20s, especially for all-stars and 80% starters.

Also, I'm not writing everything off as luck, I'm just pointing out that it plays a huge factor. The message here is not just "we need to get lucky", it's an educational message about what teams can really expect from the draft. If you "expect" to get an all-star at the #5 spot, then there's a good chance you'll be disappointed because only six #5 picks in the last 15 years have made an all-star team. Only seven at that spot have ended up as regular starters in the NBA.
 
http://numbersdont.com/?p=672

Wednesday Q&A: Chances of Winning a Ring as a Late Pick.
Written by: Mark Cameron (Camakazee)

Dear Numbers Don’t,

I know this isnt really related to the Cavs, but what are the chances of a random player, a designated top 10 pick, or a second round dude ever winning the ring in his career?

Thanks,

Sock.

Well, Sock, that’s a good question. Unfortunately, it would be far too complicated to crunch all of the data for every single player that has ever won an NBA ring. Instead, I’ll take a look at the NBA champions over the past decade since it just officially ended.

Furthermore, I’ll limit the scope to only players that played in at least 3 of every 4 playoff games the year of the championship, weeding out benchwarmers in the process. I figure if a player plays in 75% of his team’s games then he’s considered a rotation player by that very liberal definition. On average, 8-10 players play in 75% of their team’s playoff games in a championship season.

Ultimately, I will look at these 8-10 players from championship teams each year for the past 10 years to see where each player was drafted on average.

Picture1.png


*Indicates a that a player in this grouping won NBA Finals MVP.
**While P.J. Brown was the 29th pick of the 1992 NBA Draft, he was the 2nd pick of the 2nd round, thus will count as a 2nd Round pick rather than a pick 16-30, despite being picked 29th.
***Undrafted players will be considered as pick 61, one spot after the usual last pick of the draft.

Okay, so what does all of this mean, if anything? Well, first it’s interesting to note that 18 out of the 93 rotation ring-wearers over the past ten years were top 5 draft picks. That’s 19.4%, which not only shouldn’t baffle you, but shouldn’t be too surprising considering that the top five picks of a draft are supposed to be the best of the class (thus, more likely to win a ring on their team).

However, you must also consider that since the average draft has 60 players, the top five picks only account for 8.3% of the players drafted each season. Furthermore, the top five picks typically go to the teams with the worst records, barring trades and/or fluke lottery drafts, making it nearly impossible for these players to start out their careers on playoff contending teams. In this light, the fact that nearly 20% of the rotation players on championship teams over the past ten years were top five picks is fairly impressive.

Adding to that, it’s very interesting to notice the trend of Finals MVP over the past ten years. For the first seven years of the decade (2000-2006), the Finals MVP was awarded to a top five pick. On five occasions it was a number one pick (Shaquille O’Neal x3, Tim Duncan x2), once it was a third pick (Chauncey Billups), and most recently it was a fifth overall pick (Dwyane Wade). However, the latest trend has the Finals MVP trophy moving away from top five picks. The last three winners of the award, Kobe Bryant, Paul Pierce, and Tony Parker, were picked 13th, 10th, and 28th, respectively in their draft. For the sake of Cleveland Cavaliers fans, hopefully this trend doesn’t last too long.

But what I found most impressive was the distinction, or lack thereof, between top 15 picks and late first rounders or second round picks. Top 15 picks accounted for 42 of the 93 rotation players, or 45.2%. On the other hand, picks 16 and on, even ignoring undrafted players, accounted for 43 of the 93 ring bearers, or 46.2%. Now some of this is skewed since picks 1-15 only account for 25% of that year’s draft picks, yet 16-60 make up 75% of the draft, but many second rounders aren’t even expected to make an NBA roster. In fact, second round picks aren’t guaranteed a contract like all first rounders are, making the prospect of a second rounder sticking on an NBA roster a slim chance. Ultimately, while the numbers are slightly skewed due to the percentage of picks and the quality of teams picking at those spots, one would be correct in stating that over the past ten years a player drafted 16th or later is just as likely to be a rotation player on a championship team as a top 15 pick.

It’s also cool to see that there have been eight instances (8.6%) where an undrafted player has been a rotation player for an NBA championship team. Over the past ten years, there have been six players to accomplish this feat; Bruce Bowen (x3), Darvin Ham, Udonis Haslem, Mike James, Fabricio Oberto, and Ben Wallace. Lastly, if you were to look at the odds, a player is almost as likely to be a rotation player for a title winner when being undrafted (8.6%) as being picked 6-10 in the first round (10.8%).

The final area to cover is the average draft position. In the past ten years, the 2007 San Antonio Spurs were the only team to win a championship with rotation players whose average draft position places them in the second round (average pick of their 10 rotation players was 32.3). This should make Cavs fans feel better knowing they were swept by a bunch of second rounders. In a similar fashion, the 2004 Detroit Pistons were the only team to win the Larry O’Brien trophy with three undrafted rotation players (Ham, James, and Wallace), boasting the next highest average draft position (28) after all three Spurs teams (’07 – 32.3, ’03 – 29.4, ’05 – 28.8).

The good news is that when you look at the ten rotation players for the Cavaliers so far this season, they have a high average draft position (26.6), three are second-round draft picks (Gibson, Varejao, and Mo Williams), and one is an undrafted player (Moon), just like those three Spurs teams. So keep your fingers crossed for now, Sock. Or should I say toes?

Make sure to email us at numbersdont@gmail.com for a chance to have your stat analyzed by our staff and check back every Wednesday to see if your question was used!
 
Last edited:
I dont give a shit about historical draft analysis because every draft is different and it is one part of the equation.
 
I would say that writing everything off as "luck" is taking the easy way out. There is no doubt that the Chris Wallace, the Kevin McHales, the Mike Dunleavy's, etc. of the world have completely blown their drafts. However, was it bad luck or bad talent evaluation? Probably a combination of both.

The problem here is that there is luck in every aspect of the game. Those promoting the Pistons model have to realize how insanely lucky the Pistons were in winning that title. Will a player of Rasheed Wallace's caliber be available for pennies on the dollar in the one year you need him the most? Where was our Rasheed last year? We really could have used a prime Rasheed last year... Ultimately, when you let all of the percentages play out, if you ever want to win an NBA championship, you want to take the path that gives you the most UPSIDE, and the draft, especially for a market like Cleveland, is that path, it's the only real path.


In addition to Rasheed, I'd also like to point out that the other 4 were unusual cases as well. Billups was the #3 overall pick in the '97 draft, but didn't really do much until his 7th league in the year. In fact, after he was traded to Denver (the first time), they thought so little of him, that they traded him to Orlando as salary filler (he never played a game for the Magic). It is pretty rare to have to wait 7 years for a top-3 pick to start producing.

Ben Wallace is the ultimate "where the hell did that guy come from???" player. Everyone talks about how drafting in the top-5 isn't a surefire way to get a superstar, but the odds of an undrafted player from a D-II school becoming the best defensive big man in the NBA are ridiculous. That was like winning the lottery, especially since the Pistons traded Grant Hill for him (for those that don't remember, Grant Hill was really good back then. The trade was considered insanely one-sided).

Tayshaun Prince, it could be argued, was the best college player in the country his senior season, yet he fell to 23rd in the draft. Again, the odds of a guy like Prince falling to 23, then accepting his role as the 4th option aren't very good at all.

Rip Hamilton ended up in Detroit because Michael Jordan decided to come back and stupidly traded him for Jerry Stackhouse. Why? Because he felt Stackhouse would be a better fit around himself (Stackhouse's career was never the same after the trade). That one won't be too difficult as long as some team's GM decides to come back and play again and agrees to trade his best player for an overrated, older, overpaid, me-first scorer.


There's a reason the "Pistons model" has only produced one championship...
 
I dont give a shit about historical draft analysis because every draft is different and it is one part of the equation.

Agreed, 15 years is too small to draw any definitive conclusions. Drafts like 2003 come around from time to time, and then drafts like 2006 come around. You tank for a draft like 2003 where the situation is overall excellent, you trade picks around and target specific players you like on draft day for one like 2006 (Boston and Portland), where your superior scouting/managment lets you find diamonds in the rough.
 
In addition to Rasheed, I'd also like to point out that the other 4 were unusual cases as well. Billups was the #3 overall pick in the '97 draft, but didn't really do much until his 7th league in the year. In fact, after he was traded to Denver (the first time), they thought so little of him, that they traded him to Orlando as salary filler (he never played a game for the Magic). It is pretty rare to have to wait 7 years for a top-3 pick to start producing.

Ben Wallace is the ultimate "where the hell did that guy come from???" player. Everyone talks about how drafting in the top-5 isn't a surefire way to get a superstar, but the odds of an undrafted player from a D-II school becoming the best defensive big man in the NBA are ridiculous. That was like winning the lottery, especially since the Pistons traded Grant Hill for him (for those that don't remember, Grant Hill was really good back then. The trade was considered insanely one-sided).

Tayshaun Prince, it could be argued, was the best college player in the country his senior season, yet he fell to 23rd in the draft. Again, the odds of a guy like Prince falling to 23, then accepting his role as the 4th option aren't very good at all.

Rip Hamilton ended up in Detroit because Michael Jordan decided to come back and stupidly traded him for Jerry Stackhouse. Why? Because he felt Stackhouse would be a better fit around himself (Stackhouse's career was never the same after the trade). That one won't be too difficult as long as some team's GM decides to come back and play again and agrees to trade his best player for an overrated, older, overpaid, me-first scorer.


There's a reason the "Pistons model" has only produced one championship...

Not to mention the Lakers team they beat lost their third best player (Malone) when he blew out his knee and were in the midst of the Shaq and Kobe feud hitting a breaking point and boiling over into a on-court chuckfest where they refused to pass the ball at any point in the games/
 
I dont give a shit about historical draft analysis because every draft is different and it is one part of the equation.

If we want to go that route, we definitely should not be tanking this year. Everything changes from December to June, but right now I don't see any franchise players in the college ranks, nor do I see a blue-chip seven footer to bring along. Montiejunas sounds pretty intriguing, but I haven't seen him play yet.

BTW: Loved the research, Matches. You did the old SCT crew proud.
 
I don't understand the "debate?" For those of you "debating," do you not understand this team sucks? This team will be picking in the top 5 of the lottery as currently constructed. There's no need to try to get worse... we're putrid. We have 7 wins, mostly against garbage teams and one emotional opening night victory. This team will be LUCKY to win 25 games.
 
The problem with any analysis is that to be a contender you must get elite players. Since no one can argue that Cleveland will add that guy as a free agent, they have to trade for them or draft them. Teams on occasion make an error and trade a future elite talent before he turns the corner, but it doesn't happen often. So the best chance, and it is still very tough, is to draft one.

History clearly shows that you are far more likely to get them at or near the top of the draft but even with those high picks you are still unlikely to get the guy that takes you to the top level. The Cavs got lucky and grabbed the top pick when LeBron was the no-brainer pick. Toronto got the first pick in the year that Bargnani was the top choice. Luck is clearly necessary no matter what to build a team from the basement to championship contender.

Ownership has a tough decision to make. It is certainly possible in the Eastern Conference to spit and glue together a team to 35-40 wins and slip into the back of the playoffs. To some fans that will be entertaining enough, for others that will get old real quick after the recent run of success. It is the NBA equivalent of being a fan of Purdue in Big Ten football. They hover in the lower middle of the conference and every now and then will do a bit better. But they will never be an elite team. I suspect Gilbert is not content with being Purdue. I will be surprised if they don't gut the team at mid-season.

Ultimately it may be that Cleveland has a 1 in 20 chance of getting very good again if they gut it, and 1 in 50 if they don't. Either way the odds suck but one method improves your chances.
 
This is a great post. It is common practice to factor in talent over blind luck when evaluating people's success in their jobs. It gives us a feeling that we have much more control over our success than we really do. Remember, those of you who are considering this "too small a sample size," that you must also consider sample size when evaluating a "draft guru."

Note: for the rest of this post, I will use impact player = 85% starter, all-star, or all-NBA

If we manage to grab a top-5 pick for five years in a row, assuming Chris Grant can hit on an impact player 70% of the time, no matter where he is in the top-5, there is still a 46% chance that we would come out of this with three or fewer impact players. So, there is a 46% chance that Grant would come out of FIVE years of drafting with a record of 3 out of 5, worse than his true ability (70%).

There is also a 15% chance that we only get two or fewer impact players. That's for FIVE full years of picking in the top-5. With a GM that hits on picks significantly better than average.

So, even with someone who is quite above average at picking in the draft, there is still that very statistically significant 46% chance of only getting 3 impact players (and remember, these impact players could all be just serviceable starters...) out of FIVE full years of tanking. If we are unfortunate enough to land in that 15% after tanking for five years, and all we have are two serviceable starters and a bunch of role players, I don't think many people will be very happy. Chances are, Grant would be fired after 2 or 3 screw-ups, and Byron would be fired after 2 years, even though neither would have made it long enough to truly evaluate them. My point is, Gilbert needs to keep asses in seats, and with the downside of tanking being that people lose interest, I don't think that it makes much sense to deliberately go for the top-5 pick for 4 or 5 years straight. The chances just aren't good enough to do it.

Are you guys willing to spend five years of total suck just to have a (almost) 50/50 shot at getting a really good team with an All-NBA player, an all-star, and a good, quality starter?

Or, would you rather have 7 years of the "treadmill of mediocrity," where you are picking in the 11-20 range? Assuming Grant is also good at drafting in the middle of the draft, he will get an impact player (starter or better) 20% of the time. This gives a 36% chance of getting 3 or more impact players over those 7 years of drafting. I would much rather watch 7 years of decent basketball and have a 36% chance of being really good than watch 5 years of absolute shit basketball for only a 54% chance of being really good.

The best - and most fun - thing we can do is make smart moves, not overpay, acquire more picks to increase probabilities, keep asses in seats, and get lucky!

The biggest thing that people should get out of this is, that there is very little difference in the probability of getting an impact player at 11 and the probability of getting an impact player at 28, 29, or 30. The flip side is that there is a clear difference in probabilities of having a top-5 pick and the rest...but the Cavs still need to make money, even while rebuilding. The other flip side is that we appear to be awful and are headed for a top-3 pick anyway :(

Also, Matches, the probability of hitting a Red or Black in roulette is 36/38 ;) great post though!
 
This is a great post. It is common practice to factor in talent over blind luck when evaluating people's success in their jobs. It gives us a feeling that we have much more control over our success than we really do. Remember, those of you who are considering this "too small a sample size," that you must also consider sample size when evaluating a "draft guru."

Note: for the rest of this post, I will use impact player = 85% starter, all-star, or all-NBA

If we manage to grab a top-5 pick for five years in a row, assuming Chris Grant can hit on an impact player 70% of the time, no matter where he is in the top-5, there is still a 46% chance that we would come out of this with three or fewer impact players. So, there is a 46% chance that Grant would come out of FIVE years of drafting with a record of 3 out of 5, worse than his true ability (70%).

There is also a 15% chance that we only get two or fewer impact players. That's for FIVE full years of picking in the top-5. With a GM that hits on picks significantly better than average.

So, even with someone who is quite above average at picking in the draft, there is still that very statistically significant 46% chance of only getting 3 impact players (and remember, these impact players could all be just serviceable starters...) out of FIVE full years of tanking. If we are unfortunate enough to land in that 15% after tanking for five years, and all we have are two serviceable starters and a bunch of role players, I don't think many people will be very happy. Chances are, Grant would be fired after 2 or 3 screw-ups, and Byron would be fired after 2 years, even though neither would have made it long enough to truly evaluate them. My point is, Gilbert needs to keep asses in seats, and with the downside of tanking being that people lose interest, I don't think that it makes much sense to deliberately go for the top-5 pick for 4 or 5 years straight. The chances just aren't good enough to do it.

Are you guys willing to spend five years of total suck just to have a (almost) 50/50 shot at getting a really good team with an All-NBA player, an all-star, and a good, quality starter?

Or, would you rather have 7 years of the "treadmill of mediocrity," where you are picking in the 11-20 range? Assuming Grant is also good at drafting in the middle of the draft, he will get an impact player (starter or better) 20% of the time. This gives a 36% chance of getting 3 or more impact players over those 7 years of drafting. I would much rather watch 7 years of decent basketball and have a 36% chance of being really good than watch 5 years of absolute shit basketball for only a 54% chance of being really good.

The best - and most fun - thing we can do is make smart moves, not overpay, acquire more picks to increase probabilities, keep asses in seats, and get lucky!

The biggest thing that people should get out of this is, that there is very little difference in the probability of getting an impact player at 11 and the probability of getting an impact player at 28, 29, or 30. The flip side is that there is a clear difference in probabilities of having a top-5 pick and the rest...but the Cavs still need to make money, even while rebuilding. The other flip side is that we appear to be awful and are headed for a top-3 pick anyway :(

Also, Matches, the probability of hitting a Red or Black in roulette is 36/38 ;) great post though!

With the treadmill of mediocrity, you get 7 years of it, get your 11-20 range picks, and then the players that made you mediocre get old, the 11-20 range guys become overpaid role players elsewhere, and you're heading straight towards being trash anyways, just 7 years of boring uneventful basketball until you ultimately start the true process.

If the team has no interest in building for an NBA championship, what's the fucking point? Some people may be content with "decent" all the time, but I'm not. I have yet to see a "treadmill team" that was fun or even interesting to watch. The only mediocre teams I have seen that I enjoyed are teams who made the jump from bad to mediocre because their newly drafted studs finally started winning them some basketball games here and there and you know that their ceiling is sky-high.
 
With the treadmill of mediocrity, you get 7 years of it, get your 11-20 range picks, and then the players that made you mediocre get old, the 11-20 range guys become overpaid role players elsewhere, and you're heading straight towards being trash anyways, just 7 years of boring uneventful basketball until you ultimately start the true process.

If the team has no interest in building for an NBA championship, what's the fucking point? Some people may be content with "decent" all the time, but I'm not. I have yet to see a "treadmill team" that was fun or even interesting to watch. The only mediocre teams I have seen that I enjoyed are teams who made the jump from bad to mediocre because their newly drafted studs finally started winning them some basketball games here and there and you know that their ceiling is sky-high.

Look, you can argue the entire first paragraph to be against the 5 years of sucktitude by just changing a few words. If your argument is that players are going to leave, then that negates any argument for building a team.

Here's where our viewpoints may differ: I believe that watching a team compete for a playoff spot, and maybe making the playoffs a few times in 7 years will be much more fun to watch than a team competing for the number 1 pick for 5 years. The probabilities of building a successful team in each of those two scenarios are close enough to justify choosing the more entertaining one!

Also, do note that there is also a 50/50 shot that you won't see that "jump from bad to mediocre," even after 5 years of picking in the top-5 of the draft. There is, however, an almost 40/60 shot that you will see a jump from mediocre to very good (or even great) after 7 years of picking in the 11-20 slot.
 

Rubber Rim Job Podcast Video

Episode 3-12: "Max Strus Juice"

Rubber Rim Job Podcast Spotify

Episode 3:11: "Clipping Bucks."
Top