Neil DeGrasse is a badass
This Tyson ain't ear-bitin',
but he ain't no chicken, neither
(Tyson... Mike... Tyson... Chicken... never mind)
No, this is not another poem.
For those of you who haven't seen the amazing TV series "Cosmos" this past season, hosted by preeminent astrophysicist Neil DeGrasse Tyson, you're missing out on an amazing achievement on many fronts, including lending visual sensibilities to topics that tend to elude most minds ability to conceptualize: space and time.
The Cosmos visual arts crew lend a sense of scientific reality to wondrous topics, such as visualizing the history of man as a subset of the history of the universe, and plotting it all out onto a lunar calendar year. The adage is fairly common, that humans don't enter the picture until the last increments of that "calendar year", but Cosmos gives you visuals of the enormity of that statement. Gets my geek on.
In my day gig, one of my many other geeky pleasures is studying ways to communicate difficult data concepts to audiences to make them more digestible and understandable. My background in graphic design and math made me a weirdly natural fit for the gig, and I've used it for topics as diverse as product sales to clients, employee satisfaction, and the sale of a multi-million dollar corporation by accurately portraying their long-term financial health in a simple data chart. I sit around reading books on data visualization by gentlemen like Edward Tufte and Stephen Few (though one of the two is decidedly NOT a gentleman). You should see the centerfolds...
After Andy's excellent article on our schizophrenic Denver Nuggets a couple days ago, I started to have palpitations over yet another Rorschach test of a Nuggets season, and wondered how our current campaign fits into the history of the franchise. Maybe a broader view would calm my nerves a bit.
I thought about how effective Cosmos' graphic had been in contextualizing that larger timeframe, and wondered if I might be able to do something a little radical to help myself visualize the Nuggets history in the only absolute-and-binary terms that matter at the end of the day: wins and losses. So... I took a few hours to comb through and paste some data (thanks, Basketball Reference) from the entirety of the Nuggets NBA history. Fitting all of that data into the column size of this page was a bit of a trick, but... Here it is...
"But... I can't read that, Mike..."
I hear you. I cannot read it either. That's part of the problem I'm trying to solve here. And still, the data has some interesting properties here, even before we start to manipulate it. I'm certain you can look at the spreadsheet above and spot the two shortened seasons since the Nuggets entered the league. Easy enough, right? Another interesting view is to look across the upper rows of the data. Even when you can't see the exact numbers, you can tell when a row goes from single to double digits. How fast one sees double digits in the win or loss column tells everything about the start of that season, and also lends to that weird "skyscraper" effect across the top of the numbers (the wins in the 11-71 season don't reach double digits until game 77 - yikes). You can visually see all of that without even being able to read the numbers. Let's try something more:
I'm going to change all of the wins to a green number, and all of the losses to a red number. This is to show the typical accounting "surplus/deficit" color schema, and also is very easily differentiated by most anyone but the approximate 2% of people who suffer from red-green color blindness. To that two percent... sorry for the grey visuals. This change should show a few more patterns. A few more hours of "coloring" later...
Huh. Interesting. I do see those patterns starting to emerge, but the entirety of the data is too visually stimulating to still make out any patterns. Let's try something really crazy... Annnnnnd a few more hours. The wife reminds me I do this for free.
If you didn't scroll down too far, the next step in my mind was to "turn up" that color differentiation. Yes, it's going to be ugly. That's desired. I'm not shooting for pretty, I'm shooting for easily read/differentiated. So, I filled in the backgrounds of those red numbers with a red box, and the green with a green box. The cool part is there's still data in the cells, it's just "camouflaged" by the background color matching. It's visually disturbing to a degree, but it clearly shows wins and losses over the course of 38.25 seasons. And there are patterns galore. Shield your eyes!
WOW, THAT'S UGLY... But... what does it all mean? There are a few things I see in this universal historic view...
The stretches of wins and losses this season, while frustrating, are nothing compared to a few of the streaks seen over the team's history.
The 11-71 season (just to the left of the first strike-shortened year) is obvious and depressing in its strands of red.
The 57-win season is beautiful in its green-ness.
The more visually jumbled an area is, the less streaks (wins or losses) the Nuggets experienced through that timeframe. Duh, I know, but... Still a factor.
The Nuggets early NBA history, and the Carmelo Anthony/George Karl years were their most fruitful (look for green).
Vertical data is vastly more important than horizontal data here. Horizontal data will give you an idea of the long-term health of a franchise, and it's good and bad periods, but the vertical streaks are what spell out the history of each unique season.
There are more observations, but I want to hear more about what any of you see first. There are many ways I could have portrayed this data, but this is a slice, looking at a single method of data visualization. What do you see in the eye test above? Does the broader view help contextualize our current hit-and-miss state, or only further confuse troubled minds? Talk to me Nuggets Nation!
P.S. In the 38+ seasons the Nuggets have been in the NBA (including last night's win against the Miami Heat), Denver's collective regular season record is 1510-1580, if my calculator is correct. Yes, there is a little bit more red in that picture.