Handicapping Home Court Advantage
-Matt Cox
*EDIT: This article has been refreshed to reflect updated 2019-20 season results (which can be found here)
Whether it’s the Cameron Crazies at Duke or the Oakland Zoo at Pittsburgh or the Paint Crew at Purdue (I just triggered Virginia fans everywhere for mentioning the Paint Crew instead of the ‘Hoo Crew’ ), there’s simply no place like home…
From a fan perspective, the perpetual debate over ‘who has the best home court advantage in college basketball?’ makes for good banter, sure, but let’s be honest here - is this really an argument that can be won definitively? As a fraudulent frontrunner who considers himself both a Duke fan and an Indiana fan, I find myself moderating the ‘Assembly Hall versus Cameron Indoor’ dispute far too often. My personal opinion? Assembly, Cameron and Allen Fieldhouse at Kansas are the gold standards of college basketball venues, but I refuse to anoint any single one with official GOAT status. Regardless of where you stand on that debate, when it comes time to remove our “fan” hat and put on our “handicapper” hat, there’s no place for these fan-driven biases.
Estimating the value of home court advantage is as much an art as it is a science, thanks to a myriad of variables that fluctuate in importance from game to game. The process of determining home court values for each and every matchup is hard enough as it is, but consider all the information and analytics now at the bookmakers’ disposal. In this ‘Big Data’ era, the smartest guys in the room are getting smarter, faster. The almighty quest to find an edge against the books was already an uphill battle to begin with, but that slope is become steeper and steeper by the day.
Listen, I’m not trying to scare or intimidate here. Rather, I’m reinforcing the criticality of accurately estimating the value of home court, an often glossed over component of the tedious handicapping process.
Analysis Overview
I’d be remiss if I didn’t acknowledge Ken Pomeroy’s inspiration on this analysis, who dove into this same topic with a similar approach a year ago to the day. Effectively, my analysis below is a refresh and an expansion of his work, and probes deeper into the following two domains:
The declining value of home court advantage: This phenomenon has been well publicized in college basketball, so don’t think I’m declaring a new revelation here. As Mr. Pomeroy revealed in that aforementioned analysis, home court value has been undeniably shrinking over the past few years. While my main goal was to see if that trend continued in 2019, I’m more intrigued by where that trendline is heading moving forward. My hypothesis is that there is some arbitrary “floor” of how low home court value can go and I think we’re approaching, if not already at, that threshold.
The home court value of low-majors: The exciting part about refreshing Pomeroy’s analysis at this particular point in time is that the 2018-19 college basketball campaign marked a ‘guinea pig’ season for oddsmakers. Before last year, books would only produce spreads for a subset of the nation’s 32 conferences, so if you wanted to bet on the SWAC or the MEAC you typically had to wait until conference tournament time to do so. That all changed last season, as most major sportsbooks published lines on every single Division 1 game.
As the chart below indicates, there were nine conferences last season which were consistently lined for the first time, giving us degenerates an unhealthy amount of games to bet on each day.
Before going any further, it’s important to call out the biggest caveat or limitation of this entire analysis: all data input and analysis output is specific to conference games and does not consider non-conference results. This is because non-conference games rarely feature head-to-head rematches, which serve as the foundational data points for extrapolating both actual home court advantage and oddsmakers perceived home court advantage. While there are certainly other approaches we could take to evaluate the value of home court advantage for both conference and non-conference results, it is a massive undertaking and would require extensive data massaging to arrive at reasonably accurate estimations - let’s save that for a rainy day, shall we?
For this conference-specific home court value assessment, it’s important to understand how to interpret the output from the actual home court advantage (‘actual HCA’) and oddsmakers perceived home court advantage (‘oddsmakers HCA’). Results for both outputs are derived from the head-to-head matchup pairings within conference play. The difference is that actual HCA is based on the actual scoring margin variance for each head-to-head matchup pair, while the oddsmakers HCA is based on the spread variance for each head-to-head matchup pair. Overlaying these two approaches allows us to compare the difference in what oddsmakers’ believed the home court advantage should’ve been versus what the home court advantage actually was.
Ok, enough about the process… let’s talk RESULTS…
Macro Findings
The average value of home court advantage hovered around 3 points last season, a .35 point drop off from the year prior…
After a brief spike in average HCA during the 2017-18 campaign, that figure corrected back on course last year, finishing below 3 points for the third time in the last four seasons. Per the chart below, the actual 2019 HCA value of 2.98 brought the 7-year running average to 3.19.
…while this dip aligns with the longer-term trend of diminishing home court value, the rate of decline appears to be leveling off…
In five of the last seven seasons, the average value of home court advantage has declined from the year before, but the broader trendline reveals that those year-over-year dips are flattening out. After home court advantage bottomed out in 2017, the rebound in 2018 implies that an arbitrary floor between 2.5 and 3 points likely exists. In other words, there is likely an expiration date on this home court advantage decline, as we may begin to see these year-over-year values hover between 2.5 and 3 for the foreseeable future - though, only time can prove or disprove that hypothesis.
…oddsmakers were slow to react to this trend in its early days, but they’re now beginning to catch on
With how quickly home court value declined from 2012 to 2014, the chart below allows us to examine how fast oddsmakers adjusted. The actual home court advantage trendline from the chart above has been duplicated in the graph below for a side-by-side comparison of the actual home court advantage (yellow shaded line) to the oddsmakers’ defined home court advantage (brown shaded line).
Unsurprisingly, oddsmakers slowly and deliberately curbed their average home court advantage value over the past seven years but appear to have finally narrowed the gap. Bettors who were ahead of the curve may have exploited this market inefficiency up until 2018, but the discrepancy between actual and oddsmakers’ home court edge looks to be nearly extinct.
It’s worth noting that some of the dip in HCA from 2018 to 2019 could be due to the inclusion of the aforementioned nine low-major conferences, many of whom don’t carry the same home court advantage as the big boys. We’ll take a glance at that dynamic shortly, but the chart above visualizes an important dynamic in college basketball handicapping. Getting a sense for the rate of change in both the actual home court advantages and the oddsmakers’ perceived home court advantages helps us set the goalposts for where standard home court values sit now and how those goalposts may be shifting going forward.
Home Court Advantage by Conference
To stay true to my aforementioned itch of understanding bookmakers’ perception of low-major home court value, lets hone in on the nine leagues that were lined on a consistent basis for the first time this year…
In appraising home court value for the nine newly lined conferences, oddsmakers played it safe for the most part...
Per the chart below, oddsmakers kept the average home court value for these nine leagues fairly uniform and none of the nine conferences ended up at the extreme ends of the spectrum. A telling figure is the small range between the highest and lowest average home court value among this group: 3.39 (MEAC) - 2.96 (Patriot) = 0.45
Anecdotally, this seems quite logical. Given oddsmakers likely have less robust data on these smaller conferences, they opted to assign HCA values close to the mean of the other 22 conferences, a reasonable and conservative assumption.
…however, this conservative approach may have resulted in the overvaluation and undervaluation of the ‘true’ home court advantage for a few low-major leagues
Using only one seasons worth of actual HCA values can be highly volatile. This is why looking back multiple years gives us a better baseline to assess oddsmakers’ accuracy in their home court value assignments for these nine newly lined leagues. With a larger sample of in-conference head-to-head pairings included in the analysis below, we can be more confident in our assertions that oddsmakers may be over or undervaluing value home court advantage in certain conferences and teams.
Such is the case with the Southland, the America East and the NEC. By broadening the actual HCA data points to span multiple years, we can see that the average Southland home court advantage has bounced around 4 points for the last 7 seasons, almost a whole point higher than the 3.09 average HCA oddsmakers assigned in 2019. This implies that the Southland’s true home court value may be stronger than what the oddsmakers believe - thus, home court advantage in Southland conference games could be underrepresented in the betting markets.
On the flip side, I suspect the oddsmakers may be giving too much respect to the home courts of the NEC and the America East. The actual HCAs for both leagues have consistently finished under 2 points on average, more than a point below the average HCA assigned by the oddsmakers to NEC and A-East conference games in 2019.
My hypothesis for the discrepancy in actual versus oddsmakers HCA for all three conferences revolves around geography. The Southland covers a fairly large geographical region, which makes travel especially tough for a low budget league that can’t often accommodate chartered flights and luxurious hotels. In contrast, both the NEC and A-East are concentrated in the dense northeast part of the country and the average travel experience is typically much less strenuous. Such is the case with the MAAC and the Ivy, two other conferences that span a relatively small geographic radius, both of whose average HCA values have consistently ranked among the lowest in the country.
Home Court Advantage by School
We’ve talked big picture macro-level insights and conference-specific insights. Now, we reveal which teams hold claim to the best and worst home court advantage in college basketball. The two charts below show the strongest and weakest home courts, as defined by the oddsmakers’ average home court value.
*NOTE: Teams from the nine conferences covered in the section above have been omitted due to sample size constraints