“It was the best of times, it was the worst of times…” Charles Dickens’s famous opening lines set the stage for a tale about 18th century Europe, but he might as well have been talking about the 2015-16 NBA season. It’s incredible to think that right now, we have two teams approaching historic starts to the season, but on polar opposite parts of the spectrum.
The Philadelphia 76ers are breaking new barriers in terms of ineptitude. They’ve lost 23 straight games dating back to last season, and their 0-13 start to the year is inching closer to the all-time worst mark of 18 straight losses to start a season by the 2009-10 New Jersey Nets. Meanwhile, the team that relocated from Philadelphia over a half century ago, the Golden State Warriors, are an absolute joy to watch and are redefining ways in which the game of basketball can be played. Coming off a championship season, the Dubs have reached even new heights, starting the year 14-0 and on the cusp of the best start in NBA history, a 15-0 record shared by the 1993-94 Houston Rockets and the 1948-49 Washington Capitols.
All of which begs the question: which will happen first: a Philadelphia win or a Golden State loss?
It’s truly a great philosophical conundrum. Is it easier for someone to topple the ultimate Goliath or for the blindest of squirrels to eventually find a nut? To find an answer, I decided to do a deep-dive into the world of statistics.
I applied Justin Kubatko’s win probability formula to calculate both team’s probability of winning each of their upcoming games. The model uses Basketball Reference’s Simple Rating System (SRS) statistic, normalizes it based on the fact that we’re using in-season numbers, calculates the difference between the SRS of the home and away teams, and then runs that figure through a logistic regression model Kubatko developed. Obviously, it’s a broad statistical formula and doesn’t account for things like number of rest days, injured players, etc., but it provides a good baseline for the team’s chances in each game. I would encourage you to read his full breakdown if you’re interested, but the formulas used were as follows:
Adjusted SRS: aSRS = (Games * SRS) / (Games +12)
SRS difference: dSRS = aSRS[home team] – aSRS[road team]
Win Probability: p[home team] = 1 / (1 + e^-(0.61323 +0.167546 * dSRS))
Starting with the Sixers, the model yielded the following win probabilities for their games over the next couple weeks:
Good news for Sixers fans: the model actually has them favored for the Lakers game! Between that contest and the Nuggets game being basically a toss-up, I feel confident the team will have won at least one game by Dec. 5, with Dec. 1 being the day I’ll peg for that elusive first W. Unfortunately, the Sixers will likely have tied that historically terrible 0-18 start along the way.
On the other side of the fan spectrum, I performed the same analysis for the Warriors, with the following results:
With relatively easy games upcoming against the Nuggets and Lakers, it would seem the Warriors are very likely to set the all-time record for wins to start the season at 16-0. Then, things do get dicier, however. Aside from the home contest against Sacramento, Golden State has four tough matchups on the road, for all of which the model places the Warriors as just slim favorites. It seems probable that for fans in the Bay Area, the magical carpet ride will come to an end in one of those four games. I’ll say Golden State picks up that first L against either Phoenix or Utah, and with both games coming before that Dec. 1 meeting between the Sixers and Lakers, I believe we have our answer.
Conclusion: Golden State loses first! Sorry Warriors fans. And sorry Sixers fans.