Forget Polls, ND rated 20-32 by...

[Easton Pa ND Fan]

several power rating services. The average was # 42 for
35 services. The best ratings are in the group below.
Links to these individual sites are in the top of the
mratings Linked page...:alcoholic:

# 20 - Solesismic
# 20 - Tuba
# 23 - Massey MOV
# 24 - Keke-Moon
# 25 - Junky's
# 28 - Doktor
# 32 - DeSimone
# 32 - Laz Index

http://www.mratings.com/cf/compare.htm

[rontdtarchala]

we smack sparty we will jump into the national polls

[Easton Pa ND Fan]

we smack sparty we will jump into the national polls

As a subway alum, I will feel...:rofl::alcoholic::nanana: and finally
:dead:...Tom

[Flyin_Irish]

And MSU's avg on that site is 44, two spots lower than ND. Spit out some numbers, Easton, what's your prediction?

Also notable, teams ranked ahead of us are 1: USC (duh), 39: BC and 41 NC. Again, suggesting that USC is the only team on our schedule who would be expected to beat us convincingly.

[SanDiegoIrishFan]

I've never heard of those other polls, burt I agree that we will be ranked after we spank Sparty!

[Easton Pa ND Fan]

I've never heard of those other polls, burt I agree that we will be ranked after we spank Sparty!

Power rating services assign mathematical and statistical
formulas to assign a 'power' number to a team. Comparing
these numbers allow win predictions and point spread esti-
mates. You can even compare teams that do not schedule
each other.

Polls are opinion or voting services that rank their guess-
work in some order. They cater to popular opinion and have
some predictive value.

The problem is that when you see teams ranked, it has
a stronger effect on judgement than a power number. The
popular Pigskin site is a good example. Their 15th rank-
ed team (Oregon) is 15 spots from the 30th spot (South
Carolina). Yet the power rating difference is only
6% or 6 score points...Tom

http://www.powerratings.com/cfootball.htm

[rontdtarchala]

we will have to dominate but I think we do just that... and I will go on record right now saying that usc is a close game... and I think we win if its close in the fourth...

[BGIF]

...

Polls are opinion or voting services that rank their guess-
work in some order. They cater to popular opinion and have
some predictive value.

The problem is that when you see teams ranked, it has
a stronger effect on judgement than a power number. ...Tom


And the voters in those polls don't use similar criteria for ranking. Some coach's vote the Homer Voter, themselves and/or fellow conference members (Conference Venue Sharing), or opponents (Boost Strength of Schedule). Some local writers do the same.

Some voters rank whose best each week while others vote from the preseason based on which team they think will win the National Championship.

Back in '92 a writer from Arizona voted Alabama #1 each week from Preseason through the Final Poll. In the Preseason he was the butt of jokes as many voters had Bama lowly ranked or not at all. Each week he continued to vote Bama #1 even though SEC AND T'Town writers weren't.

His rational was simple. "Look at the schedules." Bama had a weak OOC schedule. A weak SEC schedule. They had timely byes before tough SEC foes or the tough foes played Bama the week after playing a tough foe. Giving Bama the perceived edge in less potential injuries at least the bruises due to heavy combat.

That writer voted not who was best for a given week or who would be best at the end. He voted primarily on who had the easiest road to take.

And Bama, they actually won. Not in their minds but actually on the field when they thrashed Miami and in the final polls.

[BGIF]

several power rating services. The average was # 42 for
35 services. The best ratings are in the group below.
Links to these individual sites are in the top of the
mratings Linked page...:alcoholic:

# 20 - Solesismic
# 20 - Tuba
# 23 - Massey MOV
# 24 - Keke-Moon
# 25 - Junky's
# 28 - Doktor
# 32 - DeSimone
# 32 - Laz Index

http://www.mratings.com/cf/compare.htm




Tom, you and I have been discussing rankings/ratings for several seasons. I've been following polls and rankings for a quite awhile but haven't heard of many of those services. Massey is the only one on your list above I recognize. BTW, Sagarin is listed for each of his three variations.

What is the credibility (reliability?) of these services?

[BGIF]

BTW, back in that '92 season Bama won ugly weekly. They would have beaten SDS by a field blocked or made at the final gun. They would have beaten Michigan recovering 6 wolverweasel fumbles but they would have given 5 of them back. They barely won week after week. But they won them all while the better teams each found a way to lose a game or two and slide in the rankings.

[Easton Pa ND Fan]

Tom, you and I have been discussing rankings/ratings for several seasons. I've been following polls and rankings for a quite awhile but haven't heard of many of those services. Massey is the only one on your list above I recognize. BTW, Sagarin is listed for each of his three variations.

What is the credibility (reliability?) of these services?

BGIF,

Keke-Moon, Junky's and Tuba are new to me. I saw a re-
ference to one of them somewhere as being a resurrection
of an earlier site.

Solesismic, Massey MOV, Doktor, DeSimone and the Laz In-
dex appear yearly on Massey's site. They are also listed
under various categories ('retro', etc.) in Beck's track-
ing site where the prediction accuracy is given. Beck's
shows an average accuracy of 78% +/-.

Not all start at the same time as several require a
"wiring" of all D1 teams playing D1 opponents. Last
year, I used Keeper's Combined and Predictive indexes
for my 'head on the chopping block' predictions here.
His site has some interesting articles as well.

I read you other posts in this thread with interest.
You have a real handle on insights and background
information - keep up the good work...Tom

http://www.mratings.com/cf/compare.htm

http://www.thepredictiontracker.com/ncaaresults.php

http://home.att.net/~keepers.football/wsb/html/view.cgi-home.html-.html

[deesmygal]

It's a shame with the polls.....but we will have to be 4-0 to get any consideration.

They are scared of the back-lash from 2007.


All this will take care of itself in time.

[notredomer23]

after a spartan massacre we shall be ranked 24th

[rontdtarchala]

again your most likely correct 23

[SanDiegoIrishFan]

Tom,

Do any of the ratings services rate by position and therefore look at matchups or do they just calculate one overall power rating per team?

[Easton Pa ND Fan]

Tom,

Do any of the ratings services rate by position and therefore look at matchups or do they just calculate one overall power rating per team?

offense and defense. Thompson SPRS does the best, by
far, on breakouts. I do not see 2008 updates on their
website, it may be that they need 3-4 weeks of data or
they went kaput. If kaput, it is a real shame because
they showed how certain offense and defense statistics
contributed to scoring and wins.

The stats that they measured can be assigned to indi-
vidual positions.

The two sections inside the introduction (left side of
link page below) are definitely worth the read...Tom

http://webpages.charter.net/ted_thompson/

[SanDiegoIrishFan]

Interesting. I wonder how good their predictions are only three weeks into the season. Does anyone track the season-long trend in accuracy for these predictions?

[Irish#1]

we will have to dominate but I think we do just that... and I will go on record right now saying that usc is a close game... and I think we win if its close in the fourth...

Let's not get ahead of ourselves. As much as I hate USC, I have to admit they looked very very very good against OSU. I'm just hoping for a win this week, before I start to think about Purdue. I'll get to USC later.

Keep in mind, the road team has been the winner of this series in recent years, so that is a plus on our side.

[Easton Pa ND Fan]

Interesting. I wonder how good their predictions are only three weeks into the season. Does anyone track the season-long trend in accuracy for these predictions?

http://www.thepredictiontracker.com/ncaaresults.php

I'm down 3 power rating groups for predictions.
Marsee and Thompson provided 3 sets of seasonal
summary (9 of 27 columns with 120 rows) from
which the excel program calculated formulas.

These formulas were applied to current game data
in the next 121 to 135 rows. Now I have 9 blank
columns in these rows becuase Marsee's house was
in a hurricane and Thompson seems to be offline.

A hurricane is no excuse :guns: