Where Do Teams Get Their Stars?

For those that didn’t check out his piece a year ago, James Brocato from Shut Up and Jam tackled this topic about a year ago. It’s one of my favorite pieces, and why I’m revisiting the topic.

It’s really as easy as this.

You need a superstar to compete. Of the many arguments stat heads and conventional fans get in to, this one everyone agrees on. Here’s some basic stats:

Since 1978* every championship team, with the exception of the 1979 Seattle Supersonics was lead by a player in the top 10 as measured by Wins Produced for the season.
Since 1978* every team to make the finals, with the exceptions of the 1979 Seattle Supersonics and the 1999 New York Knicks, was lead by a player in the top 15 as measured by Wins Produced for the season.

A question is, what type of players become great? I broke it down into the following categories for what type of player they were last season:

Rookies (this season)
Below Average Players (WP48 < 0.100)
Average Players ( 0.150 > WP48 >= 0.100)
Good Players (0.200 > WP48 >= 0.150)
Great Players (WP48 >= 0.200)

Let’s see how that shakes out:

Category	Top 20 Players based on Previous Season
Rookies	25
Below Average Players	31
Average Players	62
Good Players	147
Great Players	415

Of the 680 players we have the previous season’s data for, we can see a very odd curve. The odds that a great player will be a rookie or have had a below average season the season before? Terrible! In fact, the same is true with average players. If a player will help your team contend, the best bet is for them to already have been a good or great player!

This post isn’t that optimistic for teams without great players. The basic rule for success in the NBA is to get a superstar and use them to be great for multiple years. And the data supports this.

As a last note, many people will discuss strategies for getting great players. The simple truth is that the problem is not about getting star players, it’s about scarcity. A team tanking, clearing cap space, or making bold claims will not create more stars. There will always be a limited amount. LeBron James can only play for one team. The salary cap rules make most of them look the same. Unfortunately a lot of success in the NBA will have luck built in (Will your draft pick pan out? Will a team trade you a star? Will your market be appealing to a star?) That’s why my basic advice to any team is not overpay for players, to try and wins games and take advantage should the opportunity arises. However, I suspect that advice will not always be followed.

-Dre

*1978 is the first year we use as it is the first year the modern boxscore was used in the NBA and this is needed to calculate Wins Produced.

17 Responses to "Where Do Teams Get Their Stars?"

Al S says:

September 28, 2012 at 7:35 am

Obviously, your graph shows that the best strategy for ” getting great players” is to already have them on your team. You don’t really answer how this is accomplished. I suspect you basic advice is wrong, and that the best way to get a great player is to lose games and draft him with a high draft pick, but we don’t really have the empirical evidence to to test that.
Hasani Davis says:

September 28, 2012 at 7:59 am

Cant go out and buy players with high WP… unless you are the celtics of 08, heat of last year and lakers this year?
oh and YAY my favorite cinderella 99 knicks for making one of those statistical anomalies!!
James Harman says:

September 28, 2012 at 8:02 am

To really make this post interesting it would be informative to show where the good to great players already on the team came from (draft vs. trade). Perhaps including how long they were on the team before they were great for them would be an interesting chart too.
Thomas Baldwin says:

September 28, 2012 at 8:54 am

I’d say we do, Al. Regress draft pick number against career WP…
Jon C says:

September 28, 2012 at 10:05 am

I would guess that well over half of all great players leave the team that drafted them about in the middle of their prime. And unless you have the number one pick, you have a fairly low chance of a drafting a great player. Losing games is very expensive and painful and while it increases the odds of having a high draft pick, it lowers the odds of attracting the great players who are in their prime and who are ready to leave the team that drafted them. I think the strategy depends a lot on what market you play in. The attractive destination teams probably shouldn’t do that the tank and hope strategy, but maybe the less attactive teams should.
Kai Xiong says:

September 28, 2012 at 10:41 am

@ Dre, gotta agree with AI S on this one, answering the questions with “Where do teams get their stars” with “already on team” isn’t really answering the questions because Stars don’t pop out of thin air. Eliminating the non-answer leaves us with off-season acquisitions as the best chance from your graph. I’m not 100% sure but fairly certain we can call this the Big Market/Warm Weather factor. I wonder how that would look if you controlled for Los Angeles. “Already on team” and “Have >2 WP” players are poor answers to how to get a star. Using a 1 season framework is too shallow. Trace the player roots and see how they got on the team. Either drafted/stayed with team, traded stayed with team, or acquired stayed with team, not just already on team.
The Diss says:

September 28, 2012 at 11:01 am

Looking at the top 20 players by WP in 2011:

10 were traded to their teams
6 were drafted by their teams (I’m counting Kevin Love as drafted by the T-Wolves)
4 were signed as free agents

Look at the top 20 players by WP in 2010:

12 were drafted by their teams
6 were traded to their teams
1 was signed as a free agent
1 was drafted in an expansion draft

If I have time later I will do this for all years in which we have data
Al S says:

September 28, 2012 at 11:52 am

I am at work and supposed to be working, not fooling around on the Internet, but I took a look at Top 20 (based on Wins Produced, as listed at NBA Geek’s site) for each of the last 2 years. Of those total 40 player/years:
- 19 were acquired by their team by being drafted in the lottery and stayed with that team
- 2 were drafted later in the 1st round and stayed
- 4 were drafted in the second round and stayed
- 5 signed as free agents
- 10 were acquired by trade
(Of course, some players show up twice – LBJ counts for 2 of the free agents, Tyson C counts for 1 trade and 1 free agent, etc.)
Plenty of caveats apply, but it seems to me to say that the most likely route to having a Top 20 player on your team is to have drafted him in the lottery. It doesn’t say what a team should do in a given situation – not every team should be acquiring lottery picks – but if your team doesn’t already have a player likely to be Top 20 next year, a more likely route to getting one is draft lottery.
Al S says:

September 28, 2012 at 12:30 pm

The other thing we can do is a very rough calculation of how likely a given player will turn out to be in the top 20 next year. A Top 20 list for a given year should contain on average:
0.74 rookies
0.91 players who were below average the prior year
1.82 players who were average the prior year
4.32 players who were good
12.21 players who were great
We can then calculate the percentage chance, on average, a player will end up on the Top 20 list, based solely on their prior year WP – it is the expected # of that type of player on the list (i.e., 0.74 rookies) divded by the total universe of those type of players (i.e., last year there were: 81 rookies, 287 below average players, 83 average players, 63 good players, and 45 great players). Gives us the following percentage chances, if you knew nothing else except their WP the prior year:
Rookies: 0.9%
Below Average Players: 0.3%
Average Players: 2.2%
Good Players: 6.9%
Great Players: 27.1%
xavier.quach says:

September 28, 2012 at 12:46 pm

It’s also worth noting that in the lottery era the only #1 picks to win a championship with the team that drafted them Duncan, Robinson, and Hakeem. Along with multiple lines of research here, it can be concluded that tanking is not a good strategy.
The Diss says:

September 28, 2012 at 1:03 pm

I decided I had the time and created the graph I mentioned above. It isn’t necessarily perfect, as some judgement calls had to be made. I decided that players drafted in an expansion draft (ie Gerald Wallace) should count as “drafted” players. For sign-and-trade deals, if it were simply a player for picks (LeBron James) I classified them as “free agent” players. If it involved any other player I classified it as a “traded” player.

Without further adieu, the graph: http://i.imgur.com/uKh3V.jpg
Kai Xiong says:

September 28, 2012 at 1:27 pm

@ AI S Good analysis, I believe with a couple of marquee destinations, teams are better off doing things the old fashioned way and building through the draft. At least as far as acquiring that first chip goes and trading building around him (SAS, Houston, Dallas, the texas model). Funny case is LA, they’ve seemed to have been doing things in reverse, trade/sign Top 5 player then draft transcendent Top 5 player after.
Jon C says:

September 28, 2012 at 1:47 pm

Actually, now that I think about it, the answer as to what you should do if you want a great player is not the lottary. That is hard (predicting great play) and requires losing. If you want a WP top 20, most teams can probably get one just by trading for him or offering more money. Any team that fully committed to WP model could offer above market value for such players and get one. For examply, Dre Iggy, Noah, Nash and others are hardly untouchable players. A good offer would have gotten those players. The Knicks could have traded Melo for Dre Iggy, for example. Nash could have been had for a good draft pick. Rondo signed for far less than a max contract. Current NBA teams may not often be acquiring WP stars in this manner, but any team certainly could get them if it put its mind to it.
The Diss says:

September 28, 2012 at 2:29 pm

I think this entire debate is interesting, but really the more salient point gets back to the heart of Yay! Points! and most other WoW discussions: teams value players very improperly.

In 2011 Landry Fields and Kris Humphries were both top 20 players (according to WP), and nobody was knocking down their doors for them. Sure, Humphries signed for 1 year $8 million, but given his standing, you’d think someone would’ve thrown a 4 years $30 million offer at him.

In 2010 Troy Murphy was a top 20 player and he could barely get any run the next year.

It seems to me that the number one way to acquire stars has to start with even knowing who they are in the first place.
Where do teams get their Great Players? Al S and The Diss add even more! | The Wages of Wins Journal says:

September 28, 2012 at 2:31 pm

[...] Post navigation ← Previous [...]
Todd Bogart says:

September 29, 2012 at 4:23 pm

And approaching the acquisition of talent from the other angle—which teams have been the best at cutting their losses and dumping their blunders and which teams have held on to their mistakes too long?
While We’re Waiting… Antonetti’s Letter, Capturing A NBA Star, And Browns’ Snap Counts | WaitingForNextYear says:

September 30, 2012 at 7:00 am

[...] From The Wages of Wins Journal, here’s a look at how NBA teams have acquired Top 20 players over the past 25 years, “Of the 680 players we have the previous season’s data for, we can see a very odd curve. The odds that a great player will be a rookie or have had a below average season the season before? Terrible! In fact, the same is true with average players. If a player will help your team contend, the best bet is for them to already have been a good or great player! This post isn’t that optimistic for teams without great players. The basic rule for success in the NBA is to get a superstar and use them to be great for multiple years. And the data supports this.” [The Wages of Wins Journal] [...]

September 2012
M	T	W	T	F	S	S
« Aug				Oct »
					1	2
3	4	5	6	7	8	9
10	11	12	13	14	15	16
17	18	19	20	21	22	23
24	25	26	27	28	29	30

Share this:

Leave a Reply Cancel reply