Gary Payton, Shawn Kemp, Hersey Hawkins, Detlef Schrempf: between 1992 and 1998, each of these players had several seasons that would have made them the most valuable player on most other team’s rosters. Seattle, though, in this era, had a dominant starting five–not just one star, but a unit of extremely productive, extremely valuable players at every position.
Payton and Kemp served as perfect complements to one another, both carrying some of the scoring load, one stationed at the perimeter, the other interior, while Hawkins and Schrempf (as indicated by their lighter, more pastel shades in the graphic) did a little bit of everything for the team. Even Kendall Gill and Nate McMillan brought multifaceted perimeter games to the table, while no major player specialized exclusively in scoring, making it difficult to shut down the team’s offense.
Even when Kemp left, Vin Baker stepped up to fill his role on the team, but in the strike-shortened 1999 season, Baker disappeared, and neither he nor the Sonics were able to recapture their former ability. The team of the 1990s, underappreciated because it failed to win the championship, gradually dismantled, and the Allen/Lewis combo was too scoring-focused to be the centerpiece of a very successful club.
Reviewing the history of this illustrious franchise just reemphasizes the disappointment many fans feel at the prospect of this team moving from Seattle. Kevin Durant, in his rookie year, lead the team in BoxScores, and despite a poor showing this year, the future looks bright for the Sonics–on the court.
I would be interested in hearing what true Sonics experts see in the BoxScores graphic: Was Payton truly the MVP of the team for all those seasons, or ought Kemp be at the top? How about some of the older teams? Seattle fielded some pretty good teams in the 80s, lead by Gus Williams (whose yellow color reflects his prototypical combo guard style) and Jack Sikma (the sweet shooting rebounding center (light blue))–do the colors and orderings mesh with your memories of those teams? What do you think?
The Arbitrarian 
Long-time admirer, first time commenter.
What would be nice is if the vertical axis was stretched/shrunk depending on how many wins the team racked up. [or am I misunderstanding how it works?]
Keep up the good work,
-Jawad A. Bhatti
First, it’s great to have a long-time admirer, and second, thanks a lot for taking the time to comment–I have no accountability, if people don’t write and tell me whether this stuff is any good or not.
Stretching/shrinking the vertical axis isn’t a bad option, it would end up looking like a stacked bar graph, where the height of each bar is equal to team wins, and each bar would be partitioned into chunks according to the percent of valuable contributions–certainly this would be an easy way to instantly appraise team success and player BoxScores.
I elected, however, to make the vertical axis a percentage axis that sums to 1, while putting team wins on the horizontal–my reasoning is that this makes it easier to compare the importance of any given player to their specific team (with the bar graph option, a player contributing 20% of valuable contributions for a 50-win team would be twice as tall as a player contributing 20% to a 25-win team–you’d see the difference in BoxScores, but not so much in PVC.) Also, the width of the vertical bars, scaled according to team wins still, I think, gives a pretty good idea of relative team success. BoxScores are thus conveyed by area, and the actual number of BoxScores are explicitly written, so as to aid such comparisons… I suppose that a more vertically-oriented graph like you suggest would be just as valid, but I suppose I preferred the emphases on comparability of relative importance to teams, while still preserving important team and individual information… Does anyone else have any comments on this subject?
My husband is a big Sonics fan, so I’ve absorbed some things by osmosis (and by living in Seattle). I can’t comment with much insight on whether Payton was really the MVP of the 90s Sonics. My own interests revolve around information presentation; I find your information and presentation really interesting.
I do find this way of presenting the info too distracting. I think that the eye is first drawn to the colors. Even if one ‘reads’ the information left-to-right (and so apply the time scale), the information first reads as a jumble, since that’s the way the colors look. IOW, you have to know too much about the underlying data & organization to be able to make sense of the graphic. Once you do understand things, it’s quite rich.
I’m trying to think about how to present the time series of your information in a way that (a) someone can make sense of info at first glance, (b) is more visually pleasing, and (c) encourages more exploration. (Those are the goals of any infographic, of course.)
It seems that you are trying to show a time-series of pie charts. I’m sure there’s a way to present this out there. I’ll keep pondering this.
Meanwhile, my husband is greatly amused that I’m investigating NBA data in such depth.
Admittedly I am a Suns fan, but I think those Sonics teams of the mid-9o’s were the second best to never win a ring. Of course they were second to the 92-93 Suns.
Shawn Kemp in the 96 finals: 23 points, 10+ rebounds, 2 blocks and a steal per game against the greatest defense in my lifetime.
Gary was most important. Shawn may have been most efficient but he needed Gary to set the table.
First of all, I think this is fascinating work…in fact, it’s a good step to understanding the dynamics of good teams…
Ashley makes a good point though — the colors reference information that is not immediately evident to the viewer. You mentioned in a previous post developing some sort of player taxonomy.
Maybe refining that and then defining a few shades that are relatively similar would help to describe what pieces made those teams good and how they complemented each other?
Not sure… I think it’s great work though.
Also, if there was some way to indicate that those units in the 90’s were probably some of the best defensive units ever, that would be interesting…
I’m a big fan of the color/archetype taxonomy idea. It’s a creative way of adding another dimension of information into the chart (and it also doesn’t distract me – it’s actually easier to process (for me) because the boxes are more clearly highlighted).
Really neat work here.