2018-2019 initial season projection

Finally!  The Sagarin starting ratings are out.  Seems a little late this season.
And the asteroid has orbited back toward perihelion.

Potentially 13 games against the Sagarin Top 25 (Tennessee is not a given).
27 games against the Sagarin Top 100.
No opponent from the bottom half of Division I.

In other words, Kansas will once again be playing one of the toughest, if not
THE toughest, schedules in all of Division I.

Here is Kansas' season:

SITE   OPPONENT                         SCORE    PREDIC    ERROR    PROB.
----   -----------------------------   -------   ------    -----    -----
NEUT   #  4 Michigan State                        -0.05             0.498
HOME   # 95 Vermont                              +17.81             0.947
HOME   #156 Louisiana-Lafayette                  +22.48             0.980
NEUT   # 12 Marquette                             +3.72             0.632
NEUT   # 18 Tennessee (or next entry)             +4.99             0.675
NEUT   # 47 Louisville (or previous entry)        +8.21             0.772
HOME   # 90 Stanford                             +17.31             0.942
HOME   #126 Wofford                              +20.68             0.970
HOME   # 65 New Mexico State                     +14.23             0.902
HOME   #  2 Villanova                             +0.59             0.521
HOME   #116 South Dakota                         +20.17             0.967
AWAY   # 66 Arizona State                         +7.90             0.764
HOME   #124 Eastern Michigan                     +20.65             0.970
HOME   # 41 Oklahoma                             +10.87             0.838
AWAY   # 31 Iowa State                            +3.67             0.631
HOME   # 23 TCU                                   +8.76             0.787
AWAY   # 42 Baylor                                +4.54             0.660
HOME   # 24 Texas                                 +8.87             0.790
AWAY   #  9 West Virginia                         -0.77             0.472
HOME   # 31 Iowa State                           +10.07             0.820
AWAY   #  8 Kentucky                              -0.88             0.468
AWAY   # 24 Texas                                 +2.47             0.589
HOME   # 28 Texas Tech                            +9.50             0.806
AWAY   # 13 Kansas State                          +0.69             0.525
HOME   # 55 Oklahoma State                       +13.06             0.882
AWAY   # 23 TCU                                   +2.36             0.585
HOME   #  9 West Virginia                         +5.63             0.696
AWAY   # 28 Texas Tech                            +3.10             0.611
HOME   # 13 Kansas State                          +7.09             0.740
AWAY   # 55 Oklahoma State                        +6.66             0.728
AWAY   # 41 Oklahoma                              +4.47             0.658
HOME   # 42 Baylor                               +10.94             0.840

The cumulative probability depends on who we play in the NIT.  It's
22.22 plus either 0.675 (if we play Tennessee) or 0.772 (if we play
Louisville).  Either way, the fractional win will round up to a
projected regular season record of 23 wins and 8 losses.  Of course,
the Sagarin starting ratings are heavily biased by last season, so
the huge changes that have taken place aren't fully reflected in the
new ratings.  However, with so many of the preseason polls placing
the Jayhawks in the #1 position, I suspect that many readers will
expect a 23-8 record to be pessimistic.

As I have pointed out repeatedly over the many years of doing season
projections, you can't simply think that "Hey, we're ranked #1 in the
preseason polls, which means we should be favored in every game,
therefore our projected regular season record should be 31-0."  Such
a projection would be the case only if the probability of winning each
individual game is 100 percent, but it clearly is not.  Especially with
games against highly ranked Michigan State and Villanova, one must
recognize that the probability of winning those games is closer to 50
percent, so the odds favor winning only one of those two.  That's why
one should look at the cumulative probabilities, as shown in the table
above.

With random scheduling, it will take about 9 games before all 353
Division I teams are connected, and 10 to 11 games before one might
consider them "well connected", which won't happen until mid-December,
so the current Sagarin ratings should be taken with the proverbial
grain of salt.