Still, there is much to be learned from point spread data. One simple thing to look at is home court advantage. It’s been a personal obsession of mine for a few years now. Over a decade-plus, I have written pieces on HCA ranging from\u00a0horrible<\/a>\u00a0to poor<\/a>\u00a0to\u00a0lame<\/a>\u00a0to finally, reasonably useful<\/a>.1<\/sup><\/a><\/span><\/span> The slog culminated in my ever-updating\u00a0home-court ratings<\/a> for every team in college basketball based on certain game statistics that have been shown to be predictive of future home-court advantage.<\/p>\n Coming up with some sort of regression equation to identify home-court advantage is neat for me but not the most transparent thing from the reader’s perspective. One can fairly easily dismiss the results if they conflict with their own thinking because Ken gave a lot of disclaimers and “regression” might be a code word for “guessing”. The words do sound the same and share a frightening number of letters.<\/p>\n But some skepticism is expected and the conscientious analyst should be poking holes in one’s work, anyway. In that spirit, I was curious to see what the market thought of home-court advantage on a team level. So using all of the point spreads I could get my hands on for the past five seasons, I looked at every case where two teams played in each other’s home venue during the regular season. I took the home-court advantage for each team to be half the difference in the point spread in those two games.<\/p>\n For example on January 8, 2014, Akron was favored by 6.32 points2<\/sup><\/a><\/span><\/span> at Ball State. In the return game on January 29, the Zips were favored by 13.96. Take the difference of 7.64, cut it in half, and the 3.82 is the assumed home court advantage in the pair of games. (Overly precise, for sure.) That difference is mostly home-court advantage but there were 21 days between games. In that time the market might have taken a different view of one or both teams, or some degenerate with deep pockets could have developed a sudden interest in betting MAC games.<\/p>\n But get enough games for each team and a signal starts to emerge. In the era of unbalanced conference schedules there aren’t that<\/em> many pairs of games in a given season, and some teams don’t have any games to use because their conferences aren’t worth the time of oddsmakers. But over the past five seasons there were 258 teams that have at least 15 pairs of games to use.<\/p>\n My main interest here is how these numbers compared to my statistical HCA estimates. The team with the largest spread-derived home-court advantage is Hawaii with a value of 4.5. In my system the Rainbow Warriors are ranked 238th with a value of 2.8. That’s not a good start! It turns out that’s the biggest outlier in the comparison, though. Good news for your guy: The market tends to give a larger home-court advantage to the teams that also have a large statistically-derived home court advantage. Here’s a comparison in scatterplot form:<\/p>\n All in all, I’m pretty happy with that. There is some noise for sure, but there will always be. Home court advantage is a moving target, but because there’s no way to precisely measure it, we need years of data using the methods shown here to get a reasonably stable number.<\/p>\n The noise goes both ways, too: There’s noise in my method and noise in the point spread. But given the general agreement in the two methods, there’s additional confidence that we can distinguish between good, average, and bad home courts. As the games threshold is raised to limit the teams included in the market-based method, the correlation between the two methods increases. A games threshold of 40 only leaves me with just 81 teams, but the correlation jumps to 0.72. This suggests that with more data, the two methods would produce converging estimates for each team.3<\/sup><\/a><\/span><\/span><\/p>\n Circling back, the Hawaii discrepancy is interesting since there isn’t a travel component in my method. Obviously, a 6-hour plane trip should take a lot out of the traveling team. Boston College’s home court is rated poorly by both methods, but it would suddenly be a tough place to win in if they moved to the Pac-12.<\/p>\n It’s also interesting since over the past five seasons, Hawaii’s home\/road difference in scoring margin in conference games has been a mere 4.13 points, suggesting a home-court advantage of around 2. The team has gone 25-15 at home while going 21-19 on the road. According to the market, Hawaii has been either underachieving at home or overachieving on the road for an extended period of time. And based on its performance against the spread, it is entirely the former.<\/p>\n As I said, I have a lot of confidence in the crowd, so I fully expect Hawaii does have a strong home court. But if we can’t see that in the results over a five-year period, it should give you some appreciation for (a) how difficult it is to measure home-court on a team level and (b) how little difference there is between the best and worst home-courts in the country.<\/p>\n Teams that run up a long home-court winning streak are doing it mostly because they are better than their opponents and they just happened to save their worst performances for either poor opponents at home or games on the road. Sure, the home court contributes to a long winning streak, too, but any special<\/em> advantage a team has in their building is a very small factor.<\/p>\n One other interesting thing to look at is the season average for home-court. It’s been documented that home-site advantage is decreasing<\/a> across different<\/a> sports<\/a> and leagues in recent years and college basketball is no different<\/a>. The market has followed this trend, except for last season.<\/p>\n The odd thing is that in real life, home-court advantage did make a resurgence last season, with conference winning percentage increasing to 61.0% for home teams after reaching a historic low of 59.0% in 2017. It’s an unexpected change especially if one believes fouls are an important player in home-court advantage, since fouls called per possession plummeted to a level not seen in at least two decades.<\/p>\n Next, let’s take a look at the average HCA by conference.<\/p>\n And finally, the market-based home-court values for each team with at least 15 game pairs to use.<\/p>\n <\/span>![](\"\/images\/hca.png\")
<\/p>\nSeason HCA Games<\/strong>\r\n 2014 3.68 1626\r\n 2015 3.46 1730\r\n 2016 3.35 1858\r\n 2017 3.23 1824\r\n 2018 3.30 1668\r\n<\/pre>\n
rk Conf HCA<\/strong>\r\n 1 B12 4.0 \r\n 2 SEC 3.8 \r\n 3 B10 3.8 \r\n 4 MWC 3.7 \r\n 5 P12 3.7 \r\n 6 Sum 3.7 \r\n 7 BE 3.5 \r\n 8 BSky 3.5 \r\n 9 BW 3.4 \r\n 10 MVC 3.4 \r\n 11 ACC 3.4 \r\n 12 MAC 3.4 \r\n 13 SB 3.4 \r\n 14 A10 3.4 \r\n 15 Amer 3.3 \r\n 16 WCC 3.2 \r\n 17 CUSA 3.2 \r\n 18 CAA 3.2 \r\n 19 Horz 3.2 \r\n 20 SC 3.1 \r\n 21 MAAC 3.0 \r\n 22 OVC 2.9 \r\n 23 Ivy 2.9 \r\n<\/pre>\n
H\/A\r\n rk Team HCA Pairs<\/strong>\r\n 1 Hawaii 4.5 26\r\n 2 Denver 4.4 37\r\n 3 LSU 4.4 24\r\n 4 West Virginia 4.3 42\r\n 5 Arkansas 4.2 23\r\n 6 Missouri 4.2 25\r\n 7 Iowa 4.2 27\r\n 8 Alabama 4.2 23\r\n 9 Baylor 4.2 43\r\n 10 Oklahoma 4.1 45\r\n 11 Washington 4.1 35\r\n 12 BYU 4.0 45\r\n 13 New Mexico 4.0 37\r\n 14 Kansas 4.0 44\r\n 15 Utah 4.0 35\r\n 16 Kansas St. 4.0 43\r\n 17 Iowa St. 4.0 41\r\n 18 Texas 4.0 41\r\n 19 Indiana 4.0 27\r\n 20 Boise St. 4.0 38\r\n 21 Old Dominion 4.0 20\r\n 22 Nebraska Omaha 4.0 37\r\n 23 Michigan 3.9 26\r\n 24 Washington St. 3.9 34\r\n 25 Texas Tech 3.9 43\r\n 26 Fresno St. 3.9 37\r\n 27 Oklahoma St. 3.9 44\r\n 28 Arizona 3.9 34\r\n 29 Texas A&M 3.9 24\r\n 30 Utah St. 3.9 36\r\n 31 Colorado 3.9 34\r\n 32 Georgia St. 3.9 40\r\n 33 Mississippi St. 3.9 24\r\n 34 Montana St. 3.8 36\r\n 35 Buffalo 3.8 33\r\n 36 Ohio St. 3.8 26\r\n 37 Georgia 3.8 25\r\n 38 Marshall 3.8 21\r\n 39 Penn St. 3.8 25\r\n 40 North Carolina St. 3.8 20\r\n 41 Purdue 3.8 26\r\n 42 Oral Roberts 3.8 30\r\n 43 Nebraska 3.8 27\r\n 44 South Carolina 3.8 24\r\n 45 Maryland 3.8 23\r\n 46 Air Force 3.7 38\r\n 47 Florida St. 3.7 20\r\n 48 Oregon St. 3.7 34\r\n 49 Rutgers 3.7 28\r\n 50 Wyoming 3.7 38\r\n 51 Notre Dame 3.7 18\r\n 52 San Diego St. 3.7 37\r\n 53 Mississippi 3.7 25\r\n 54 Oregon 3.7 34\r\n 55 Illinois 3.7 27\r\n 56 Minnesota 3.7 27\r\n 57 UNLV 3.7 38\r\n 58 Arizona St. 3.7 34\r\n 59 Duke 3.7 20\r\n 60 Xavier 3.7 44\r\n 61 TCU 3.7 44\r\n 62 Sacramento St. 3.7 37\r\n 63 Providence 3.7 44\r\n 64 Massachusetts 3.6 22\r\n 65 Tennessee 3.6 25\r\n 66 Southern Utah 3.6 37\r\n 67 South Dakota St. 3.6 37\r\n 68 Colorado St. 3.6 37\r\n 69 Arkansas St. 3.6 42\r\n 70 SMU 3.6 39\r\n 71 Seton Hall 3.6 43\r\n 72 Marquette 3.6 43\r\n 73 Nevada 3.6 38\r\n 74 Montana 3.6 38\r\n 75 Creighton 3.6 44\r\n 76 Florida 3.6 24\r\n 77 Eastern Michigan 3.6 35\r\n 78 Davidson 3.6 24\r\n 79 Louisville 3.6 23\r\n 80 North Dakota St. 3.6 37\r\n 81 Georgia Tech 3.6 19\r\n 82 Georgetown 3.6 43\r\n 83 Northern Arizona 3.6 37\r\n 84 Cal St. Northridge 3.6 31\r\n 85 Butler 3.6 44\r\n 86 Portland St. 3.5 37\r\n 87 Ball St. 3.5 35\r\n 88 Louisiana Lafayette 3.5 42\r\n 89 Weber St. 3.5 36\r\n 90 Memphis 3.5 38\r\n 91 Michigan St. 3.5 27\r\n 92 UCLA 3.5 34\r\n 93 Missouri St. 3.5 40\r\n 94 Kentucky 3.5 24\r\n 95 Valparaiso 3.5 41\r\n 96 Indiana St. 3.5 41\r\n 97 Wake Forest 3.5 20\r\n 98 Portland 3.5 43\r\n 99 Louisiana Tech 3.5 21\r\n100 VCU 3.5 22\r\n101 Illinois St. 3.5 43\r\n102 Toledo 3.5 35\r\n103 Loyola Chicago 3.5 42\r\n104 Central Michigan 3.5 33\r\n105 California 3.5 34\r\n106 Cincinnati 3.5 39\r\n107 Northwestern 3.5 27\r\n108 South Dakota 3.5 36\r\n109 Miami FL 3.5 19\r\n110 Northern Colorado 3.5 35\r\n111 Richmond 3.5 23\r\n112 George Washington 3.5 22\r\n113 Wisconsin 3.5 26\r\n114 Georgia Southern 3.5 37\r\n115 Auburn 3.5 23\r\n116 IUPUI 3.5 39\r\n117 UT Arlington 3.5 41\r\n118 Vanderbilt 3.4 25\r\n119 Western Kentucky 3.4 28\r\n120 William & Mary 3.4 43\r\n121 Bradley 3.4 41\r\n122 Saint Louis 3.4 22\r\n123 UC Davis 3.4 31\r\n124 Charlotte 3.4 19\r\n125 St. John's 3.4 44\r\n126 Ohio 3.4 35\r\n127 Southern Illinois 3.4 41\r\n128 Chattanooga 3.4 40\r\n129 The Citadel 3.4 39\r\n130 Stanford 3.4 34\r\n131 UCF 3.4 38\r\n132 Duquesne 3.4 21\r\n133 Pittsburgh 3.4 19\r\n134 Eastern Washington 3.4 36\r\n135 East Tennessee St. 3.4 34\r\n136 Dayton 3.4 24\r\n137 UNC Wilmington 3.4 43\r\n138 UC Riverside 3.4 32\r\n139 Connecticut 3.4 39\r\n140 Little Rock 3.4 39\r\n141 Fort Wayne 3.4 36\r\n142 Miami OH 3.4 34\r\n143 Cal Poly 3.4 32\r\n144 North Carolina 3.3 20\r\n145 Idaho St. 3.3 36\r\n146 La Salle 3.3 23\r\n147 Gonzaga 3.3 42\r\n148 Evansville 3.3 42\r\n149 Akron 3.3 35\r\n150 East Carolina 3.3 29\r\n151 North Dakota 3.3 36\r\n152 Towson 3.3 43\r\n153 San Diego 3.3 45\r\n154 Fordham 3.3 22\r\n155 Pacific 3.3 44\r\n156 Hofstra 3.3 42\r\n157 Virginia 3.3 19\r\n158 UNC Greensboro 3.3 40\r\n159 Wichita St. 3.3 41\r\n160 Cleveland St. 3.3 42\r\n161 Kent St. 3.3 35\r\n162 Saint Joseph's 3.3 23\r\n163 USC 3.3 34\r\n164 South Florida 3.3 38\r\n165 Troy 3.3 41\r\n166 Northern Kentucky 3.2 27\r\n167 Drake 3.2 43\r\n168 Elon 3.2 40\r\n169 James Madison 3.2 42\r\n170 Green Bay 3.2 41\r\n171 Detroit 3.2 40\r\n172 Villanova 3.2 44\r\n173 Louisiana Monroe 3.2 41\r\n174 Rice 3.2 20\r\n175 Western Illinois 3.2 36\r\n176 Southern Miss 3.2 21\r\n177 Houston 3.2 37\r\n178 Northern Iowa 3.2 43\r\n179 Northeastern 3.2 43\r\n180 Western Michigan 3.2 33\r\n181 Appalachian St. 3.2 38\r\n182 Wright St. 3.2 42\r\n183 Texas St. 3.2 41\r\n184 Siena 3.2 42\r\n185 San Jose St. 3.2 36\r\n186 UC Santa Barbara 3.1 31\r\n187 College of Charleston 3.1 42\r\n188 UTSA 3.1 20\r\n189 Eastern Illinois 3.1 27\r\n190 Rhode Island 3.1 23\r\n191 Pepperdine 3.1 44\r\n192 Samford 3.1 38\r\n193 North Texas 3.1 20\r\n194 Temple 3.1 38\r\n195 Middle Tennessee 3.1 21\r\n196 Bowling Green 3.1 34\r\n197 George Mason 3.1 23\r\n198 South Alabama 3.1 40\r\n199 Oakland 3.1 42\r\n200 Austin Peay 3.1 26\r\n201 Boston College 3.1 18\r\n202 Clemson 3.1 20\r\n203 Northern Illinois 3.1 33\r\n204 Wofford 3.1 40\r\n205 Milwaukee 3.1 41\r\n206 Morehead St. 3.1 26\r\n207 Long Beach St. 3.1 31\r\n208 Niagara 3.1 43\r\n209 Canisius 3.1 45\r\n210 Cal St. Fullerton 3.0 32\r\n211 Brown 3.0 35\r\n212 Tennessee Tech 3.0 26\r\n213 Iona 3.0 41\r\n214 St. Bonaventure 3.0 22\r\n215 Furman 3.0 39\r\n216 DePaul 3.0 45\r\n217 Eastern Kentucky 3.0 26\r\n218 UAB 3.0 21\r\n219 Marist 3.0 44\r\n220 Santa Clara 3.0 44\r\n221 Quinnipiac 3.0 42\r\n222 Tulane 3.0 30\r\n223 Harvard 3.0 35\r\n224 Drexel 3.0 43\r\n225 Mercer 3.0 35\r\n226 Tennessee Martin 3.0 26\r\n227 Saint Mary's 3.0 44\r\n228 Dartmouth 3.0 35\r\n229 Belmont 3.0 27\r\n230 Idaho 3.0 27\r\n231 Youngstown St. 3.0 42\r\n232 Illinois Chicago 3.0 42\r\n233 Virginia Tech 2.9 19\r\n234 Fairfield 2.9 44\r\n235 Syracuse 2.9 20\r\n236 Tennessee St. 2.9 26\r\n237 FIU 2.9 21\r\n238 Western Carolina 2.9 38\r\n239 UTEP 2.9 21\r\n240 VMI 2.9 34\r\n241 Monmouth 2.9 46\r\n242 San Francisco 2.9 43\r\n243 Manhattan 2.9 45\r\n244 Rider 2.9 45\r\n245 Southeast Missouri St. 2.9 26\r\n246 Yale 2.9 34\r\n247 Saint Peter's 2.9 47\r\n248 Delaware 2.9 41\r\n249 Loyola Marymount 2.8 44\r\n250 Princeton 2.8 35\r\n251 Tulsa 2.8 31\r\n252 Penn 2.8 35\r\n253 Columbia 2.8 35\r\n254 Murray St. 2.8 26\r\n255 Cornell 2.7 34\r\n256 SIU Edwardsville 2.7 27\r\n257 Jacksonville St. 2.7 27\r\n258 Florida Atlantic 2.5 21\r\n<\/pre>\n