Checking 2013-14 NBA Win Predictions & Projections

With 2013-14 NBA regular season in the books it’s time for some fresh projections about the NBA playoffs!
But wait a second, aren’t we forgetting about similar predictions for the regular season? As usual I’d like to expand the effort of APBR community in an attempt to measure how successfully basketball writers, bloggers and authors predicted what will happen during the regular season regarding teams’ wins.
So which writer, blogger or author had the most accurate projections before the 2013-14 NBA season?

To answer this question I saved on my hard drive predictions in early November [it's a reason why not all of the links below work, they've changed since then] from most popular sites and some googled sources [but it's not a closed list so if you want to add your's predictions let me know in the comments] and compared them to actual wins for each team by Root Mean Square Deviation which is essentially average difference between each prediction and final total wins in terms of absolute value. Results of this exercise are in the table below…

But please keep in mind two important notes:
1) don’t think or assume it’s all about the skills of people involved because randomness plays a huge part here… injuries and trades in most cases simply couldn’t have been predicted before the season but they affected some of the results in a big way [for example, Gay trade anyone? Or Horford's shoulder...].
Personally I treat it as fantasy basketball on a team level and I’m more interested in the year-to-year patterns than in actual results for one particular season because it’s basically a sample of one.

2) in most cases entries next to each other are interchangeable in terms of accuracy because difference between them can be translated into “2-3 more/less wins in the right direction would flip them”.

Onto the table…

A couple of notes about the results in the 6th season measured:

- Phoenix Suns [on average projected 25.6 wins below actual 48 achieved] shattered previous record for being undervalued [18.7 wins for Cavs in 2007-08]. It really is amazing how team pegged as a tanking bottom-feeder before the season should have been in the playoffs if not for the strength of conference.

- Other most undervalued teams were Blazers [by an average of 16.1 wins], Bobcats [15.2 - similar story to the Suns!], Raptors [10.3] and Spurs [6.0]. I think it’s notable that top three undervalued teams had a big addition before or during the season… BTW, Spurs were undervalued 4th season in a row…

- The most overvalued teams were Bucks [by an average of 16.8 wins] who were on the opposite spectrum to the Suns [viewed as a fringe playoff contender which actually tanked], Pistons [12.0 - subtraction by additions?], Nets [9.2] and Knicks [8.8].

BTW, if you want to see similar information as a graph you should visit statsbylopez.com.

- after multiple strong years from Vegas odds-makers and crowd-wisdom approach by ESPN’s forecast it was their off shooting year. Not a bad one per se but more average than usual top notch.

- members of APBR forum very strongly rebounded after last year’s disappointment… to such a degree that they collected all three medals – congratulations to HoopDon, Robert Eckstut and Nathan Walker!

- Joe Schaller had another strong season in a row so I think it’s a good time to ask some questions or explore whatever he is doing ;-)

- on the other hand simulations by BoxScoreGeeks again landed way closer to the bottom of the table than to the top so… I don’t know whenever it means anything but I would at least consider revisiting each step of the process in search of the most consistent cause of errors.
Also I think it’s notable that BoxScoreGeeks Fans as a group had a better result than the authors who offered them initial numbers – my best guess would be it happened because of crowd-wisdom effect but it may have been simply a fluke. We’ll see next year?

Have I missed anything?

P.S. it may have been my last post of this nature because more and more sites try to bring accountability to predictions which I think is a good trend but it makes this perspective less unique than it was when I started.