I guess I have to respectfully disagree
gscr...@comcast.net
: I guess it depends what rankings you go by. I've always hated the whole
: "record vs. ranked opponents" stuff because most of those statistics
: only consider where the opponent was ranked at the time you beat them.
: The only rankings that have any meaning are those at the end of the
: season.
I guess I have to disagree. It seems to me that the ranking of a team when you play them is more relevant to how your team is than their eventual ranking (or their pre-season ranking, for that matter). This no-longer-used factor of adjusting your ranking based on your opponents' rankings at the end of the season seemed to ignore how that team might have changed (for better or worse) since you played them.
what if, for example, Rutgers had played and beaten Louisville their first game? what would that have done to the BCS ranking of both?
Glen in NJ
_______________________________________________
husker site list
hus...@tssi.com
http://romaine.tssi.com/mailman/listinfo/husker
Pat
gscr...@comcast.net wrote:
:>I guess it depends what rankings you go by. I've always hated the whole
:>"record vs. ranked opponents" stuff because most of those statistics
:>only consider where the opponent was ranked at the time you beat them.
:>The only rankings that have any meaning are those at the end of the
:>season.
:>
:>
:
:I guess I have to disagree. It seems to me that the ranking of a team when you play them is more relevant to how your team is than their eventual ranking (or their pre-season ranking, for that matter).
:
My point is that early season rankings are incredibly biased, and often
are almost entirely a function of how the previous season ended rather
than an accurate portrayal of how teams are currently ranked. With only
12-13 games per season, pollsters are basing their judgements off a
relatively small sample size as it is. In my opinion, you really need
10 or more games before you can accurately gauge how good teams are.
A perfect example is the 2001 Nebraska team, who for the most part, tore
through their first 11 regular season games before being obliterated in
their final two. The team that Colorado beat by 26 points was not the
2nd best team in the country, and I don't believe the team who beat
Nebraska was the 14th best.
If you want to get even more absurd, Penn State beat us 40-7 in 2002.
Because of how we fared in the previous season, we were ranked 8th in
the country at the time of that defeat. Did Penn State really beat one
of the ten best teams in the country that night; a team that went on to
a 7-7 record? The polls may have said otherwise, but that wasn't reality.
: This no-longer-used factor of adjusting your ranking based on your opponents' rankings at the end of the season seemed to ignore how that team might have changed (for better or worse) since you played them.
:
:
There is no longer a direct strength of schedule component built into
the BCS, and with the exception of *possibly* the Billingsley, I don't
think any of the computer rankings work like this. To the best of my
knowledge, most of them factor in strength of schedule and value
"quality wins" based on where opponents are currently ranked.
:what if, for example, Rutgers had played and beaten Louisville their first game? what would that have done to the BCS ranking of both?
:
:
:
I honestly don't think it would have had much of an effect. If
anything, it would have helped Rutgers break into the rankings earlier.
Andrew Smith
gscr...@comcast.net wrote:
:>I guess it depends what rankings you go by. I've always hated the whole
:>"record vs. ranked opponents" stuff because most of those statistics
:>only consider where the opponent was ranked at the time you beat them.
:>The only rankings that have any meaning are those at the end of the
:>season.
:>
:>
:
:I guess I have to disagree. It seems to me that the ranking of a team when you play them is more relevant to how your team is than their eventual ranking (or their pre-season ranking, for that matter). This no-longer-used factor of adjusting your ranking based on your opponents' rankings at the end of the season seemed to ignore how that team might have changed (for better or worse) since you played them.
:
:
It's a contradiction to advocate "ranking when you played them" over
eventual ranking or pre-season ranking since (1) for the first game of
the season, pre-season ranking = "ranking when you played them" and (2)
"ranking when you played them" is often a mixture of pre-season and
eventual ranking.
While it is true that teams can change for better or for worse during
the season (such as due to a key injury), the ratings for teams change
far more simply because voters and computers get more data. Ask
yourself, when do you have a higher confidence in your ability to judge
a team's ability, after 1 game or after 11?
Finally, the Billingsley computer ranking is one that uses "ranking when
you played them" and in some cases the results are almost comical due a
team either utterly failing pre-season expectations or being Cinderella
team.
:what if, for example, Rutgers had played and beaten Louisville their first game? what would that have done to the BCS ranking of both?
:
:Glen in NJ
:
:
:_______________________________________________
:husker site list
:hus...@tssi.com
:http://romaine.tssi.com/mailman/listinfo/husker
:
:
:
: