Twenty years ago this week, Kentucky beat Syracuse in the men’s NCAA Tournament title game in front of a crowd of 19,229 at Continental Airlines Arena in New Jersey. It was the last time the Final Four was held in a basketball arena.reports a few years ago that the NCAA would consider returning the Final Four to arenas in the future.
">1From 1997 through 2008, the Final Four was played in domed stadiums — all but one in football stadiums — with the court in one corner or end zone. This “traditional dome” configuration enabled in excess of 50,000 fans to witness the NCAA Tournament’s biggest games in person. Beginning with two regionals in 2008,Ford Field) and 2008 South Regional in Houston (Reliant Stadium, now NRG Stadium). The setup was previously used in some regular season games, and a variation of it without seats surrounding the court was used for UCLA-Houston at the Astrodome in 1968 (the “Game of the Century”).
">2 the NCAA moved the court to the 50-yard line, allowing for crowds of 70,000 or more. Traditional domes continued to be used in the occasional regional up through last year, but the so-called “open dome” setup has been used at every Final Four and four more regionals since 2009.
Since then, many have speculated that these open domes spawned a “dome effect,” diminishing shooting performance — particularly from long range — by forcing players to adjust to distant and unfamiliar backdrops behind each hoop. The most infamous example is the 2011 national championship game at Reliant Stadium — now NRG Stadium, poster child for the dome effect and site of this weekend’s Final Four — in which Butler shot 12-of-64 (18.8 percent) and fell to Connecticut 53-41. It’s a common enough assumption that there’s now an annual procession of articles hinting at a phenomenon. But is the dome effect real?
Teams in open domes shoot slightly worse overall — but they shoot threes just fine
We looked back at every Sweet 16, Elite Eight, Final Four and national championship game played over the past 10 years — a total of 300 team performances in 150 games. Thirty-nine were played in “open” domes, 48 in “traditional” domes, and 63 in regular old basketball arenas. Here’s how teams have performed by venue and shot type in those games:3
Teams playing late-round NCAA Tournament games in open domes score fewer points4 and shoot worse on 2-point field goals and free throws than do teams playing in arenas or traditional domes. But 3-pointers — often at the heart of the dome effect critique — are converted at a lower rate in traditional domes. In fact, teams shoot threes better in open domes than they do anywhere else, including standard arenas. Overall, teams playing in open domes have an effective field goal percentage of 47.4 percent, 0.5 percentage points worse than teams in standard arenas and traditional domes.
These themes hold up when we compare teams’ performances to their season averages — an important consideration given that certain venue types host later rounds and better teams. Free throws aside, teams don’t shoot as well or score as much in the regionals and Final Four as they do over the course of the entire season — which makes sense, considering that they’re playing against top competition. Here, we’re interested in how much worse they shoot and how that varies by venue and shot type.
In line with our previous observations, points, 2-point field goal percentage and free throw percentage suffer most in open domes, while 3-point field goal percentage suffers most in traditional domes. Relative to their full-season performance, teams playing in open domes see their effective field goal percentage decline by an average of 4.8 percentage points, 0.5 percentage points worse than teams playing in arenas or traditional domes. For a typical late-round NCAA Tournament team, that differential equates to about two-thirds of a point per game.5
Team offense and opponent defense matter — venue type does not
So we’ve determined that teams shoot a little worse and score a little less in open domes — but they also play against better defenses. This makes sense — open domes host later-round games.
The defenses that play in these open dome games allow opponents to shoot 40.2 percent from the field during the season; they give up 62.7 points per game, and have an adjusted defensive efficiency of 91.7 — all weighted by number of appearances. This compares to season-long averages of 40.4 percent, 63.2 and 92.1 for teams that play in traditional domes and 40.7 percent, 63.5 and 92.0 for teams that play in standard arenas. So defenses do get better as venues progress from regular arenas to traditional domes to open domes.
In fact, how well a team shoots in a late-round NCAA Tournament game is most related to its full-season effective field goal percentage and the other team’s full-season field goal percentage allowed.6 Likewise, how much a team scores in a late-round NCAA Tournament game is most related to its full-season points per game average and the other team’s full-season points per game allowed.7 These relationships are noisy but meaningful. (That is, there is a statistically significant relationship between the full-season offensive and defensive stats and offense in all venue types, even though the individual data points are a little all over the place.)
On the contrary, there is no statistical relationship linking open domes as a venue type to poor shooting or to decreases in shooting performance — including, importantly, when you control for the quality and efficiency of the offenses and defenses playing in each. NRG Stadium on its own is another story.
NRG Stadium is unique — open domes are not
In line with its reputation, NRG Stadium has produced lower-than-average scoring and worse-than-average shooting, both overall and among open domes. Defenses there have been a mixed bag — tough against the shot, middle of the pack in points allowed, and toward the bottom in defensive efficiency.8 (Other open domes — Lucas Oil Stadium in Indianapolis, AT&T Stadium in Arlington, Texas, and the Georgia Dome in Atlanta — have been the site of good or very good shooting performances, on average.)
| OFF. PERFORMANCE (SWEET 16 AND BEYOND, 2006-15) | OPPONENTS’ FULL-SEASON DEF. PERFORMANCE | |||||||
|---|---|---|---|---|---|---|---|---|
| VENUE | NO. OF GAMES | AVG. POINTS | EFF. FG% | 2P% | 3P% | FT% | ALLOWED FG% | ADJ. EFFICIENCY |
| Georgia Dome | 3 | 69.2 | 50.6% | 48.8% | 35.9% | 68.9% | 40.0% | 90.3 |
| AT&T Stadium | 6 | 64.9 | 50.4 | 49.8 | 34.7 | 68.3 | 40.1 | 92.7 |
| Lucas Oil Stadium | 9 | 65.4 | 49.6 | 46.9 | 37.0 | 73.4 | 40.8 | 91.4 |
| Ford Field | 6 | 70.2 | 47.1 | 45.0 | 34.3 | 66.9 | 40.4 | 90.7 |
| NRG Stadium | 12 | 65.4 | 44.5 | 43.8 | 30.7 | 72.7 | 40.2 | 93.1 |
| Superdome | 3 | 63.7 | 44.4 | 41.5 | 36.8 | 71.4 | 38.3 | 88.2 |
Comparing how teams shot at NRG with how they shot during the regular season, the stadium’s negative impact on effective field goal percentage is bad. The effect is the worst among open domes and the fourth-worst among the 29 venues9 that hosted the regionals or Final Four in the past 10 years. Its impact on 3-point shooting is worst among open domes and fifth-worst overall, behind two arenas and two traditional domes:10
Daunting as it may seem, the NRG effect is not so cut and dried. Across venues, NRG stadium has a meaningful effect on 2-point field goals and effective field goal percentage, but not on threes.11 NRG’s impact on threes is only meaningful when compared with other open domes — the rest of which turn out to be pretty good places to shoot from long range.
So where does all of this leave us? First, we can pretty safely dispel the dome effect as it relates to these newfangled open domes. It fails both statistical tests and the eye test, what with five of the six open domes having less of an effect on threes than the average regional or Final Four venue in the past 10 years.
As for NRG Stadium, well, that’s more complicated. Across venues, the NRG effect is (arguably) real, but not for the reasons we might have thought (threes). Its impact on threes is real, but only among open domes. And while that might not be super compelling, Final Fours are, for now, played exclusively in open domes, so being unique among them matters for something. We’ll find out tonight.
CORRECTION (April 11, 4:15 p.m.): The original version of this article misspelled the author’s name. It is Greg Guglielmo, not Gugliemlo.