-Matt Cox

Welp, this is strange…

My offseason college basketball content schedule is officially obsolete. Hope you got your money’s worth, COVID.

March Madness. The Final Four. One Shining Moment. All of it…

Done.

Plug pulled.

Shut up Kemo, I don’t want a new show!

I had no intention of whipping up any 2020 season look back content until the end of April, but here we are. And here I am, painfully trying to distract myself from plummeting any deeper into the abyss of a Tiger King binge watch odyssey.

The meat and potatoes of this article is a refresh of last year’s “Handicapping Home Court Advantage”, which debuted in late May last offseason. Before indulging in the main course – an analysis of how home court advantage value is trending over time with breakdowns by conferences and by teams – I’m serving up a quick appetizer.

Assessing the Impact of a Longer 3-Point Line

As the picture above reveals, last summer’s lengthening of the 3-point arc was a more substantial shift than when the NCAA first pushed back the line in 2008. The prevailing rationale is that the move would free up driving lanes on the perimeter to create more offensive spacing in the half-court. The NBA’s deadliest offenses are prime exhibits of how maximizing floor spacing with pick-n-roll magicians and dead-eye shooters can manifest in orgasmic offenses, first mastered by the James Harden-led Rockets and the Steph Curry-led Warriors.

The NCAA’s long-term vision passes the sniff test, but the abrupt move last offseason was destined to cause some early hiccups for college players, only a fraction of whom possess the talent and skill level of their professional NBA counterparts. Predictably, as the data below will confirm, it took time for college shooters to retune their shooting habits.

Let’s begin with the handicapping evidence. If you subscribed to the notion that college players would need time to adjust to the longer 3-point line (and assumed oddsmakers would be slow to adapt to that dynamic), betting the UNDER consistently would’ve been a wise way to test that hypothesis.

Turns out, blindly betting UNDERS for the first two weeks of the season would’ve netted you a 52 unit return, which cashed at a profitable 56.6% win rate during that span.

Linesmakers were indeed late to the party, forced to react on the fly. The weeks to follow indicate they may have actually overcorrected to the early run on UNDERS – OVERS hit at a 52.7% success rate from weeks 4-7 – but the books finally seemed to stabilize by week 8. Juxtaposing those weekly UNDER win rates with weekly 3-point shooting percentages show the natural correlation pattern we’d all expect: weeks when teams shot the ball poorly from long distance, UNDERS hit at a relatively higher clip (and vice versa).

While the 3-point conversion rate see-sawed throughout the year, the 3-point attempt rate was far less volatile (see ‘3PA%’ column above). Per the ‘2020 Avg.’ row in the table on the right, 37.5% of all field goal attempts last season were from long distance (2nd highest 3-point attempt rate in the KenPom era), proof that the 3-point shot isn’t going away anytime soon. However, that rate of 37.5% was actually a 1.2% drop from the year prior (38.7% in 2019), a sign that the longer 3-point line may have deterred some fringe 3-point shooters from launching trey balls as freely as the year before.

Going forward, the 3-point shooting factor is just one of many changing variables oddsmakers will need to juggle in setting their totals. Scoring is a function of both tempo and efficiency, the latter of which has multiple components to unpack within it. Tempo has remained steady since the shot clock was shortened to 30 seconds back in 2013 (per KenPom, 69.1, 69.4, 69.4, 69.0, 69.1 were the D1 averages over last five seasons), but the efficiency sub-factors appear to be on the move.

Per KenPom, the national offensive efficiency average sunk to 100.8 points per 100 possessions last year, the lowest since 2013. The aforementioned 3-point shooting woes had a hand in this efficiency dip but declines in offensive rebounding rate and free throw rate also contributed to the efficiency contraction.

• Offensive rebounding: It’s been well documented that offensive rebounding is being emphasized less and less in today’s college game, evidenced by a continuous decline in the national offensive rebounding rate since 2008. The importance of offensive rebounding is amplified during shooting cold spells, as teams effectively forfeit critical extra possessions by deprioritizing the offensive glass.

• Free Throw Rate: KenPom tracks free throw rate all the way back to 1997 and last year’s mark of 32.6 was the first time in history that figure dipped below 33%. The importance of generating points from the foul line is especially important for mediocre and bad teams, many of whom score at an efficiency rate of less than a point per possession on average. Rule changes, such as last year’s 3-point line extension, and refereeing points of emphasis, such as flopping and post physicality, areas are key determinants in how frequently the whistle is blown. There’s no crystal ball into what tweaks the NCAA will make in the future, but there’s clear momentum toward opening up the game and reducing defensive advantages, where possible.

“We did a lot of study on shooting percentages, and we don’t think those will change a lot,” said Big 12 coordinator of officials Curtis Shaw of the 3-point line move. “We do think it will result in better spacing and create more offensive flow.”

The other wrinkle worth mentioning is the season-to-season fluctuation in talent. Last year, it was widely accepted that talent was down across the board, as many pundits regurgitated the “there are no great teams” talking point throughout the year. Did the dearth of offensive skill and talent contribute to clunky offense? Perhaps… But based on preliminary roster projections and a quick scan of the 2020 recruiting class, I wouldn’t bet on a significant boom in scoring based on a new player pool.

Handicapping Home Court Advantage

 Ok kids, time for dinner!

The Approach:

For all college basketball handicappers out there, especially those who rely on a power ratings approach, appraising the value of home court is a critical part of the process. Slapping a specific numerical figure to the value of home court advantage is far from precise, but the methodology used to produce the figures below (modeled off KenPom’s approach in a prior analysis project), is sufficient. Please refer to the first rendition of this analysis for additional details on calculation specifics, as well as key limitations and pitfalls of this approach. For now, here’s a high-level explanation of the two sets of figures shown below:

• Oddsmakers’ Home Court Value: The oddsmakers’ implied estimation of home court value, based on the spread differential between home and away intra-conference matchups.

• Actual Results Home Court Value: The ‘actual’ home court value based on scoring margin differential between home and away intra-conference matchups.

Macro Findings:

Last summer, we highlighted the slow erosion of home court value on a national scale, a multi-year trend observed in both actual scoring margin results and oddsmakers’ spreads. Home courts all across America must’ve rallied around my disrespect, because they sure got their revenge last year…

Per the intra-conference scoring margin differential depicted in the chart below, average home court value jumped up to 3.32 last season (up from 2.97 in 2018-19).

As discussed last year, home court value is undeniably on the decline, but there’s an arbitrary floor as to how low it can go. The ease of travel is continuing to improve, and college kids are becoming more immune to the adverse effects of travel, which collectively is diminishing the advantage of home court. But, as long as you have teams consistently traveling across the country and playing at unfamiliar venues, home court value will always exist – it will never shrink to zero. The magnitude of last year’s spike is a little surprising, but the upward correction was to be expected.

That natural trend is better exemplified in the oddsmakers’ implied average home court value trend, which shows the recent leveling off of the slope decline.  

Conference Observations:

If we dive into the conference cross tabs, we’ll see where last year’s upswing in home court value stemmed from. My colleague Jim wrote about the iron clad strength of Big Ten home courts in mid-January, as home teams were routinely demolishing visitors during the early portion of conference play. But, the Big Ten wasn’t the only league where life on the road was miserable. In fact, teams all across the Power-6 conferences defended their home floors with vigor – per the chart below, the Power-6 comprised five of the top-6 largest home court advantages last year based on actual intra-conference scoring margin results (highlighted in salmon):

Moving toward the opposite end of the conference size spectrum, the mid and low majors are where my intrigue lies. Last season marked the second year in which oddsmakers set lines for nine of the nation’s 33 conferences (highlighted in light blue) – prior to that, these leagues were typically only lined during conference tournaments and the NCAA tournament.

Isolating these nine conferences and examining their 2018-19 to 2019-20 home court value deltas, a consistent theme emerges. On average, oddsmakers discounted home court advantage in eight of the nine conferences, with the Southland being the lone exception.

Bookmakers clearly picked up on the unique nature of the Southland, a league with more strenuous travel than other low-major conferences. For the rest of the low-major landscape, books are properly tracking the trend of home / road randomness, as home court advantage here is minimal.

Team Observations:

As we dive deeper down the home court advantage rabbit hole, we find additional evidence that bookmakers are making necessary adjustments to their team-specific home court value inputs.

Below are the top-25 largest home court advantages for the 2019-20 season based on the oddsmakers’ implied estimates (shaded column), with a side-by-side comparison to their actual home court advantages based on their 9-year historical average home / road scoring margin differential.

The key trend to notice here is the consistent increase in the oddsmakers’ average home court values from 2019 to 2020 (second and third columns from the right). We said before how oddsmakers are effectively adapting to the decline of home court advantage on a national scale, but this is proof that they are not making blanket adjustments. At this extreme end of the dial, the books appear to be responding to multi-year trends of certain teams consistently performing better at home than on the road (relative to expectations).

VCU is the marquee example. The Rams home court value based on actual home / road scoring margin differential over the past nine years has been consistently around 5 points (5.62 9-year average), a trend the oddsmakers finally seemed to pick up on last season. In 2019, the oddsmakers were assigning an average home court value of 2.95 to VCU in conference home games. Last year, that figured jumped all the way to 4.75, the highest in the country.

Michigan, Nebraska, Rhode Island, Louisiana Tech, Notre Dame, Arkansas, Arizona State, Purdue and Air Force are just some of the additional examples highlighted in the chart below in which oddsmakers seem to be responding to the historical data trends in scoring margin differential between conference home and road games.

The framing and interpretation of this data is critical. It’s important to remember that home court advantage in this context is a byproduct of home AND road performance. Teams who consistently play well on the road will likely be falsely represented as having a *small* home court advantage...

• Last year, San Diego State was a prime example. The Aztecs actual home / road scoring margin differential equated to a slight negative home court advantage. That’s because San Diego State was lights out away from home, exemplified by a sterling 10-1 against the spread on the road. Viejas Arena is universally regarded as one of the premier venues in all of college basketball and the multi-year averages in both actual results-based HCA and oddsmakers spread-based HCA tell the true story.
  

On the low end of the home court advantage spectrum, we can see the inverse of the trend above, which aligns with the macro trend of diminishing home court value on the national scale. Notice how the majority of the oddsmakers’ estimated home court values in the ‘2020’ column are less than the 2019 averages (‘2019' column). Per the 9-year home / road scoring margin differential averages, many of these teams have a home court value of two points or less. Again, oddsmakers appear to be adjusting accordingly.

In the last two rows lie Wagner and Maine, both of whom checked in with an average implied home court value of less than 2 points last season (Wagner = 1.89; Maine = 1.78). Anecdotally, their tiny home court advantages might be explained by geographic location and conference affiliation. Maine’s remote location in Orono, ME seems to do more harm than good, as the Black Bears must embark on a grueling travel expedition for every single conference road game. Wagner, on the other hand, has the opposite issue (which results in the same adverse effect). The Seahawks play on Staten Island near the heart of New York City, easily accessible for almost all Northeast Conference opponents. I alluded to this in last year’s ‘V1’ of ‘Handicapping Home Court’, but most of the NEC is clustered in and around the NYC metro area, which inherently reduces the travel tension for most programs. This dynamic also holds true for the America East, the Patriot League and the Ivy League, all of whom are concentrated in the northeast part of the country.

We’ll wrap this home court value analysis refresh with a zoomed-out view of the largest home court advantages in college basketball based on the 9-year oddsmaker multi-year average. In terms of which teams truly have the ‘strongest’ home court advantages in the country, this picture is the closest approximation we can derive from this conference home / road differential approach. I should note that I intend to refresh this with a moving-average based calculation next year, which will assign higher weights to recent years:

Three of the usual suspects reappear in this slice, Denver, Colorado and Utah, who will forever enjoy a superior home court edge due to altitude – that is, unless college basketball players start training in Colorado Springs with the Olympians. Hawaii still boasts one of the strongest home court advantages in the eyes of oddsmakers for obvious reasons, though a slight decline in recent years indicate that edge may be dwindling.

In summary, this data dive verifies that bookmakers are making quick and targeted home court value adjustments based on new information. However, extrapolating oddsmakers’ home court value calculations by team and by conference with this aggregated approach must be done with caution. This approach fails to take into account game-to-game situational factors and also ignores the natural ebbs and flows of team performance throughout conference play. Still, even in the face of those limitations, this analysis delivers a useful snapshot of home court advantage trends, from the macro level all the way down to the individual school level, specifically in how closely predictive-based spreads are tracking with results-based scores.