Last week, ESPN published an opinion piece on where each of the professional sports leagues in North America should and would expand next. In the piece, Kevin Windhorst asserts that there are possible issues about the NBA moving into a redeveloped KeyArena citing concerns over revenue streams for a potential NBA franchise. I found it an interesting assertion considering David Bonderman has been identified as a major investor in the Seattle Center Arena, (potential) NHL Seattle franchise owner, and has stated his intent to pursue any opportunity to bring an NBA team to Seattle. Additionally, Bonderman is a minority owner of the Boston Celtics, so if you think anyone has an inside track to landing a franchise in Seattle, it would be him.
I asked local NBA writer, Kevin Pelton, via Twitter if he could explain the contradictory statements of Windhorst’s claims and Bonderman’s intent to become an NBA owner in Seattle. My intent was not to discredit Windhorst, only to understand how both Windhorst’s claim and Bonderman’s plan could coexist. The Twitter dialogue was healthy and constructive, but never answered. If you are interested, the whole thread is here. Kevin mentioned that if the NHL is first, then it might have scheduling advantages that would put the NBA at an implied revenue disadvantage.
As I stated in the thread, this was the first time I had heard of a potential prioritized scheduling issue between NBA/NHL in the same arena. Being the data nerd that I am, I wanted to dig into this to see if there was any truth to this.
Note that this is a narrow and high-level analysis of league schedules based on days of the week. I fully recognize that every market and team is different. I am sure that there are scheduling requirements with leagues, TV schedules (local and national), and operating agreements that make this much more complex than I spell it out to be. Additionally, each arena and team scheduling requirements could be more collaborative or restrictive from market to market.
Last season there were 11 arenas that shared an arena with at least 1 NBA and 1 NHL team. Clearly, dual anchor tenant arenas already do exist. Additionally, 4 metro areas (Minneapolis/St. Paul, Phoenix/Glendale, Miami/Sunrise, and SF/San Jose) have a separate arena for their respective NBA and NHL teams. It would appear that separate arenas for NHL and NBA teams is the exception, rather than the rule. Counting San Francisco and San Jose as the same metro is a bit of a stretch, but I wanted to be conservative in my research.
I looked at regular season schedules over the last 5 years across the NBA and NHL to see if there were any noticeable advantages or disadvantages to either league in a dual anchor tenant building compared to a single anchor tenant building.
Looking at the distribution of games league wide by day of the week already has a few interesting anecdotes. The two most popular days for the NBA are Wednesdays (21%) and Fridays (19%). In the NHL, ~60% of the NHL games are scheduled on Tuesday, Thursday and Saturday vs ~33% of NBA games on those same days.
To determine if there are conflicts or disadvantages, we need to look at the difference between single anchor tenant and dual anchor tenant arenas to determine if there is any significant change in the scheduling.
As you can see, there are no material changes when comparing the schedule of dual anchor tenant arenas vs. single anchor tenant arenas when looking at schedules by day of week. The distribution of games only varies by a couple percentage points between single anchor tenant and dual anchor tenant arenas.
I wanted to make sure there weren’t other biases in my analysis, so I excluded the Clippers because they are a third anchor tenant in Staples. I even excluded the Canadian NHL teams in case there were any Canadian Market TV requirements (i.e. Hockey Night in Canada), to see if that made a difference in the numbers. In both cases, the numbers did not significantly change between single anchor tenant and dual anchor tenant anchor arenas.
My conclusion is that concerns about revenue streams affected by prioritized scheduling is unfounded.