clock menu more-arrow no yes

Filed under:

An Analysis of Power Conference NCAA Invitations, 1999-2007

New, comments

First things first.  99% of this analysis was done by one of our members here, johnr.  John did an outstanding job, and because this requires considerable analysis of a significant dataset, John put in a huge amount of work to get this done.  I did nothing but compile it and put it up here.  This is interesting stuff, so take a long look and thank John for all his hard work and effort.  He has most definitely earned it.

John was wondering what the sort of "minimum standard" was for a Big 6 (BCS) team to get into the NCAA tournament in terms of out of conference record.  What he did was compile a list of all the BCS team's records over the last 9 years and create a scatterplot chart of the spreadsheet.  He used Ken Pomeroy's site (thanks, Ken) to obtain the won-loss data, and utilized an online statistical tool called Stata to produce some of the graphics.  For the sake of clarity, I am going to post both of them with attendant explanations.  Some of them will show the same thing, but because of technical issues, it will be easier for me simply to show them all and explain them rather than just post one and try to explain it.

So without further ado, here we are:

The second graph, which has "jitter" applied to show more of the teams (obviously, many teams are at the same data points and overwrite each other) is below.  Note that the legend is reversed -- red X is in and blue diamond is out in this graph:


  1. The non-conference record for all teams includes NCAA and conference tournament games.
  2. The conference record does not include the SEC tournament, EXCEPT in the case of the 2008 Kentucky data point indicated above.  The line I have drawn at .75 indicates Kentucky's regular season record so that you can properly compare it with the rest of the teams.

John's Conclusions:

  • Courier-Journal online (in Feb) said that since the SEC went to 12 teams and two divisions in 1991-92, every eligible team with a conference record of 10-6 or better has reached the NCAA Tournament.

  • The 2007 Final Four records book indicates that since 1985: - the most losses of any team admitted at-large is 14 - the fewest wins of any team admitted at-large is 16 - the lowest win pct of any team admitted at-large is .533 (16-14)

  • I analyzed the win-loss and tourney invitation records of the 620 power conference teams from the 9 seasons ending 1999 to 2007. (A serious weakness of my data is that the season win-loss records actually include post-season play. The strength of the data is the size and comprehensiveness of the data set.) 

  • The most striking thing in the data graphic is how rarely a power conference team finishes with a winning conference record and a losing non-conference record...and even more rarely with a conference win pct better than 70%.

  • I found that:
    • Only two teams (out of 620) from the 6 power conferences both finished below .500 in non-conference play, and finished above .700 in conference play
    • They were Texas 1999 & Michigan State 2004. Both got into the tourney. 1999 Texas 19-13 13-3 (6-10) [made tourney, AT LARGE 7 SEED] 2004 Mich St 18-12 12-4 (6-8) [made tourney, AT LARGE 7 SEED] 
    • Texas finished 1st in their conference play that season. Michigan State was tied for second.
    • 1999 Texas went 18-11 for the season, 19-12 after conf tourney, and 19-13 overall
    • 2004 Michigan State went 17-10 for the season, 18-11 after conf tourney, and 18-12 overall - note: thus, both lost in conf tourney and received at-large bids
    • Another team finished exactly .500 in conference, and .69 in non-conference play: 2002 Wisconsin 19-13 11-5 (8-8) [MADE TOURNEY, AT LARGE 8 SEED ]
    • Wisconsin finished tied for 1st in conf play
    • Wisconsin 2002 went 18-11 for the season, 18-12 after conf tourney, and 19-13 overall

  • There are 3 other teams who had winning conf records (>.500 & .<700) and losing non-conf records: 2006 Stanford 16-14 11-7 (5-7) [DID NOT MAKE TOURNEY] 2005 Indiana 15-14 10-6 (5-8) [did not make tourney] 2001 Georgia [made tourney] Georgia 2001 finished season at 16-13, 16-14 after conf tourney, 16-15 overall.

  • In those 9 seasons, there were 167 teams that finished with 11 or more wins in conference. Of these, only two failed to get a bid: 2003 Georgia 19-8 11-5 (8-3) [DID NOT MAKE TOURNEY -- ineligible!] 2006 Stanford 16-14 11-7 (5-7) [DID NOT MAKE TOURNEY] --> NIT? Stanford 2006 finished the season 15-12, was 15-13 after conf tourney, 16-14 overall.

  • I recently posted to ASeaofBlue regarding the similarities to 2008 Kentucky of the 2004 Michigan State and 1999 Texas squads, including early season performance and other weaknesses.

  • BOTTOM LINE: - No power conference team since at least 1999 has finished with a conference win pct greater than .700 and failed to make the NCAA tournament, regardless of non-conference performance. (Actually, the threshold is .650, not .700. No team with a conf pct >.650 has failed to make the tourney, except 2003 Georgia, who self-imposed a "no post season play" penalty for NCAA violations. That's a 100% record of inclusion for eligible teams, based on 154 teams total.

  • What about prior to 1999? (retrieved several weeks ago) says: - since 1995, 91% of teams that finished above .500 (262 of 289) in one of the six top conferences (excluding automatic qualifiers) were invited to the tourney. 

  • Between 1995 and 2006, only 6 teams from one of these major conferences finished 4 games above .500 in conference and were not invited. West Virginia (11-7 in 1997) Nebraska (10-6 in 1999) Boston College (10-6 in 2003) Seton Hall (10-6 in 2003) Colorado (10-6 in 2004) Indiana (10-6 in 2005).

  • The data seem to be about the weight of conference performance being greater than non-conference. However, it also is a proxy for closing the season strong. It shows that winning after New Years is important.


    • I used data from Ken Pomeroy.
    • I did not control/clarify for automatic-bids (unfortunately).
    • W-L records may include post-season play (unfortunately).
    • I excluded current Big East teams for the seasons prior to their entry into the Big East.

Additional Comments:

Another thing I found remarkable is how few teams in this 9-year period have actually found themselves above .600 in conference and below .500 out of conference.  Obviously, that is tilted quite a bit by the inclusion of post-season tournament and NCAA tournament games, but even if you move the X-axis up to .600, the number is still very small.

Kentucky finds itself in a rather unique situation.  but if history since 1999 holds, it would be extraordinary in the extreme for Kentucky not to be invited to the tournament.  It just doesn't happen at .700 (let alone .75) in a power conference.

Update [2008-3-16 12:10:58 by Truzenzuzex]:  From ESPN Bubble Watch, we have this.  A lot of hedging there, but they are coming out in favor of UK getting in.  But there sure isn't a lot of conviction there.