the regular season most games are played on one team's home court. The most accurate handicapping results in
such games are achieved by using the applicable home and road statistics for
March Madness, however, most games are played on a neutral court. When handicapping teams on a neutral court,
there are three possible sets of past performance statistics that can be used:
set of past performances will produce different results. The problem for the handicapper is deciding
which set of past performances will produce the most accurate results.
Rule: When handicapping a basketball game being
played on a neutral court, overall past performances -- home, road and neutral
court combined -- will produce the most accurate predictions.
neutral court tournament games, we are seeking to measure each team's abilities
without regard to any home court advantage.
Another name for the home advantage is the road disadvantage. A small part of the home edge is that the
home team is more comfortable. The road
team may have had a long and tiring trip, may be eating unfamiliar food, is
sleeping in unfamiliar beds, and may be in a different time zone. Most teams, however, travel enough that they
become acclimated quickly. The largest
part of the home advantage comes from the fact that the home team has more
familiarity with the court, has better access to regularly schedule practices,
and is playing in front of a crowd of fans that are cheering only for the home
team, and doing everything in their
power to negatively affect the visitors.
problem with using road game past performances only for both teams in the
tournaments is that road statistics do not measure a team's ability without
regard to home advantage or road disadvantage.
Road-game statistics are skewed by the fact that the advantages and
disadvantages for the teams in the game were unequal, with the road team always
taking the worst of it. The full road
disadvantage is automatically built into every road game performance. We can expect a team to play better on a
neutral court than on an opponent's home court.
teams play on a truly neutral court, the travel disadvantages are equal for
both, there is no fan advantage for either team, and both are equally
unfamiliar with the court. Logically,
using neutral court past performances only should produce an accurate
result. The problem with using only
neutral court games, however, is that there are rarely enough of them during
the season to produce a statistically reliable picture.
basketball teams change greatly from season to season. The only past performances that are relevant
to the current team are the current season games. In any given season, most teams play fewer than a half dozen
games on neutral courts, and, as discussed in my article about the hidden home
court edge, the courts listed as neutral may not be truly neutral. The larger our statistical sample, the more
precise our handicapping is likely is to be.
Conversely, the smaller our sample, the more likely it is that the
results will be skewed and will not measure the team's abilities accurately.
overall statistics solves both problems.
one is better than Rob when it comes to handicapping the March college
basketball tournaments. Rob?s rare
tournament Crowne Jewel play was an easy winner two days ago, and Rob?s Play of
the Day was winner last night.. Rob?s
March tournament record for the past two years is over 61%, and Rob?s record
here at Pregame so far this season is 67%.
To get today?s March Madness winners or Rob?s complete March Madness
Package, CLICK HERE.
Yes. See my blog article from today entitled "College Tournament Handicapping." UNLV is actually playing on its home court. St. Johns in the Big East was also playing on the court it uses as its home court for all season games against 1-A opponents -- Madison Square Garden.
Unlv over BYU last night would be example of Neutral court that was not a real neutral court, correct.
PS - If you are attempting to use neutral court records, you should also be sure those records are balanced between early season and late season, or better yet are all in the last part of the season. College teams can improve immensely during a season as the freshman get experience and the players mature physically. Unfortunately, most neutral games are clustered in the very early, preconference, "exhibition" portion of the season.
RJ- You are correct. The "overall" statistics are not perfectly balanced. If a team has a particularly strong home court advantage that may skew the overall numbers a bit. Some teams may have a particularly big road disadvantage.. (Hawaii is a good example in both categories. They have a strong home edge for the same reasons they have a big disadvantage on the mainland.) The "overall" numbers can also be out of balance because many teams in the NCAA play more home games than road games, and may play more weak teams in one place of the other.
There is often no perfect way to handicap. We are stuck with using the best method, rather than the perfect method. Using road numbers for both teams is the most inaccurate method because all the statistics in the sample would include the opponent's home advantage with absolutely nothing to balance that. Neutral court games would be the most valid if only there were enough of them for both teams to create a statistically valid sample. Unfortunately, in most season records there not a sufficient number of neutral court games to be a valid sample. That is particularly true after we eliminate those games that weren't truly on a neutral court. (If Fordham plays a game against UCLA at Madison Square Garden, the game will go into the neutral court record for both teams, even though there is nothing neutral about the court at all, and Fordham should be credited with almost a full home advantage.)
There is an exception to the rule in the article. If you can find two teams that have played at least 6 truly neutral court games against competition on the level they will meet today, you should use the neutral court records for both those teams to handicap. In most cases, however, using the "overall" - home, road, and neutral court combined - records will be the best measure of team ability.
This is a concept I've never run across. Interesting - with the possible exception of home teams with an especially big home court advantage.