Dayton comes in at #10...
Posted: Tue Feb 05, 2008 10:56 am
In Ken Pomeroy's "luck" index. Three A10 teams (Richmond, STL and UD) are in the top 10. This basically is a conclusion that these A10 are "outperforming" themselves in terms of # of wins compared with the offensive/defensive efficiency indexes that are used to predict outcomes.
If you're looking for a good example of what an "unlucky" team looks like, consider Notre Dame's squad of 05-06.
http://kenpom.com/blog/index.php/weblog ... the_irish/
For the statistically curious/nerdy, KenPom calculates luck as, "the deviation in winning percentage between a team’s actual record and their expected record using the correlated gaussian method. The luck factor has nothing to do with the rating calculation, but a team that is very lucky (positive numbers) will tend to be rated lower by my system than their record would suggest."
http://kenpom.com/blog/index.php/weblog ... planation/
For an explanation of correlated gaussian method (just fancy word for a "bell curve") check out: http://www.rawbw.com/~deano/articles/BellCurve.html
Teams that are statistically "lucky" basically win more games than their offensive and defensive indexes predict. I've always found KenPom to be excruciatingly thorough and remarkably accurate in terms of his ratings system, so the fact that his "luck" calculation is not used in his rating calculation suggests that he doesn't let statistical analysis affect his ratings -- luck is an observation, but not a direct variable. To factor it in would be statistical double-jeopardy by using the same numbers twice in his calculation, similar to using overlapping factors like TO's and, say, shot clock violations. But if you consider it to be an accurate observation, you can apply it to teams that seem to be playing far outside their statistical averages -- you just can't apply it to the kenpom ratings.
In conclusion, dayton sucks and they're a bunch of lucky SOB's.
NB: Xavier is ranked 150th out of what, 340 teams? so we're almost perfectly average in terms of overall "luckiness."
If you're looking for a good example of what an "unlucky" team looks like, consider Notre Dame's squad of 05-06.
http://kenpom.com/blog/index.php/weblog ... the_irish/
For the statistically curious/nerdy, KenPom calculates luck as, "the deviation in winning percentage between a team’s actual record and their expected record using the correlated gaussian method. The luck factor has nothing to do with the rating calculation, but a team that is very lucky (positive numbers) will tend to be rated lower by my system than their record would suggest."
http://kenpom.com/blog/index.php/weblog ... planation/
For an explanation of correlated gaussian method (just fancy word for a "bell curve") check out: http://www.rawbw.com/~deano/articles/BellCurve.html
Teams that are statistically "lucky" basically win more games than their offensive and defensive indexes predict. I've always found KenPom to be excruciatingly thorough and remarkably accurate in terms of his ratings system, so the fact that his "luck" calculation is not used in his rating calculation suggests that he doesn't let statistical analysis affect his ratings -- luck is an observation, but not a direct variable. To factor it in would be statistical double-jeopardy by using the same numbers twice in his calculation, similar to using overlapping factors like TO's and, say, shot clock violations. But if you consider it to be an accurate observation, you can apply it to teams that seem to be playing far outside their statistical averages -- you just can't apply it to the kenpom ratings.
In conclusion, dayton sucks and they're a bunch of lucky SOB's.
NB: Xavier is ranked 150th out of what, 340 teams? so we're almost perfectly average in terms of overall "luckiness."