Back to Preseason Index Up to Feedback Index Chris' Home Page

Feedback - Football:Preseason Consensus


Why is Athlon a 59? It's boring, and it ranks my teams low. Oh well. Phil Steele's is awesome, so is Lindy's! You underate Lindy's!

Anonymous (6/08)

Response from the editor:

The big difference in score between Lindy's and Athlon was an error on my part, in adding up the points. The two lag Phil Steele mostly on quality of information (2,000 words per team vs. 500 to 800). I've fixed the calculations.

Your Web site is great but I was looking at your accuaracy rating and one thing trobules me. It's entirely skewed towards prognosticators who pick ties. Phil Steele and ATS Consulatants did nothing but pick loads of ties and were among your top four in prediction accuracy. Theoretically, what is to stop someone from picking a six-way tie for first in the ACC Coastal. From my understanding, they would score a "0" in your caculations. There needs to be a penalty for picking ties.

Anonymous (7/07)

Response from the editor:

You misunderstand the scoring system. If teams finished in a six-way tie for the ACC Coastal, then everyone would get a score of 0 no matter what they predicted. Ties in the conference standings are interpreted in the manner most charitable to each prediction.

The system is not symmetrical with respect to ties in magazines' predictions, however. If someone picked a six-way tie for the ACC Coastal, their prediction would be counted as 3½th place -- 3.5 is the average of 1, 2, 3, 4, 5, 6 -- for every one of the six teams. This would earn them a score of 2½ points for a team that finished alone in first place (3½ - 1), and a score of 2½ points for a team that finished alone in last place (6 - 3½). This asymmetry was placed in the scoring system as I originally devised it, specifically to provide an incentive for not picking ties.

A prediction of a six-way tie for the ACC Coastal for 2006 would have earned:
2½ (Georgia Tech, 1st)
+ 1½ (Virginia Tech, 2nd),
+ ½ (Virginia, 3rd)
+ ½ (Miami-Florida, 4th),
+ 1½ (North Carolina, 5th)
+ 2½ (Duke, 6th).
That score would be 9 points, which is worse than every single magazine actually scored. (The worst score for the ACC Coastal was 8 points last year.)

Actually, my system contains a built-in incentive against picking ties. Picking ties generally amounts to sacrificing a half-point per pick. If the teams actually end up tied, then magazines which picked the teams in either order would get a perfect score, anyway, so there is no benefit to picking ties even if the teams actually finish tied.

I think your review of Phil Steele's College Football Preview, was slanted do to the fact that he furnished you what several copies of the magazine.

The just the facts approach is pretty good, but I feel a preseason magazine needs the flowery color photos, and full color fluff pieces to make it a good magazine.

Anonymous (7/00)

Response from the editor:

Phil Steele's Magazine's rating isn't a lot different from last year's when I bought it on my own. His "prediction accuracy" rating was increased because he's a lot higher in the "last three years" standings today than he was a year ago -- he stood 13th out of 14 magazines for accuracy 1996-1998, but is 5th of 14 for 1997-1999. There was a small increase in "information content" rating (because there was more spring football info) and a small decrease in "price" rating because this year his magazine is more expensive.

Your tastes may be different from mine; that's fine. I'm more into "content" than "atmosphere." As I write in the "notes" page...

I grade the magazines on the features which are important to me. This does not mean that the same features will be important to you. Hopefully the reviews (especially the detailed information on each magazine) supply enough information that you can pick out the magazines that you will like best -- even if your priorities do not match mine.

I do have a question which puzzles me a litte. In your 1999 Rankings, you have a prediction accuracy column. In this column you rate a magazine like Jim Feist or ATS Consultants with 3 stars but they are #8 and #10 respectivly on your 97-99 accuracy list. They are rated 3 times as strong as Sporting News (#8) and Phil Steele (#5) who each received just one star. The numbers that I have listed are from your 97-99 accuracy list and why do those two publications which are equal to or higher in the ratings get less stars?

Mark Henderson (5/00)

Response from the editor:

The reason for this discrepancy is that the 1999 reviews are written prior the 1999 season -- at a time when the "last three years" of prediction accuracy was 1996-1998. The current "prediction accuracy" has been updated to include the results of the 1999 season, which were not available at the time the reviews were written. Currently the "last three years" table includes 1997-1999 data.

When the reviews were written, based on 1996-1998 data, the accuracy standings of the magazines that you mention were (out of 15 magazines): Jim Feist's 4th (***), ATS Consultants 5th (***), The Sporting News 11th (*), and Phil Steele tied for 13th/14th (*, but perhaps he really deserved zero stars!).

When the accuracy page was updated at the end of the 1999 season, the 1999 results were added in (and 1996 results subtracted out) of the "last three years" table. This caused Phil Steele (who had a very good 1999 finish -- tied for 3rd/4th out of 16 magazines) to move way up in the standings. At the same time, ATS Consultants (very bad 1999 accuracy -- tied for 14th/15th out of 16) and Jim Feist (even worse 1999 accuracy -- would have been dead last if he had predicted every conference) dropped quite a ways.

Don't worry, though. The current accuracy standings will be used when the year 2000 reviews are written, and you'll see the current standings reflected in the reviews that are written this summer. The magazines should start coming out in a month or so. (But when the 2000 season is over and I update the "accuracy" page again, it will once again be "out of sync" with the reviews for a couple of months before the 2001 reviews are written.)

Why do you have old feedback from prior to the 1998 football season up under the feedback section. You should update that sector or list that no one has submitted feedback for over a year.

I love your site and find it to be the very best for analyzing College Football Magazines.

Anonymous (1/00)

Response from the editor:

I don't get a whole lot of feedback. Most of what I do get wouldn't belong here for one of three reasons: (1) the author doesn't give permission to put it up on a public page, (2) it's almost identical to a question that is already answered here, or (3) it's not of general interest and best answered in a direct E-mail reply.

Phil Steele???

The under valuation of the PAC 10 is amazing. UCLA is ranked no lower than 11th by all other publications but unranked by P.S. Arizona State is from 5th to 8th by most , but 21th by P.S. which is interesting.
[...]
There seems to be a little regional bias here. Oh well its just on paper but talk about skewing the rankings.

Anonymous (7/98)

Response from the editor:

While Phil Steele does leave UCLA unranked, he is also the only magazine to have Southern Cal anywhere near the top ten. I think it is more a matter of him picking different teams near the top of the conference, than of "regional bias" or under-estimating the Pac-10. Phil Steele has four Pac-10 teams in his national top 25; no magazine I've found yet this year has more.

Even this early, when only about one-third of the preseason magazines have been collected, a single set of picks very different from the consensus don't really "skew the ratings" much. Had Phil Steele ranked UCLA #10 -- roughly where most other magazines have them -- it would only make a three-place difference in UCLA's consensus rank. And note that Phil Steele's "unique" picks for the Pac-10 don't impact the consensus ordering at all.

You did not include the BLITZ issue with Penn State number one. Granted, this magazine includes only about 10 pages of college football, but I believe that all preseason polls should be included.

Steve Good (8/97)

Response from the editor:

There are literally hundreds of national top 25 lists. Every sports magazine has one; many non-sports magazines (e.g., Playboy) have them; there are dozens on the internet, etc. It would be extremely difficult to do a thorough job of collecting every top 25 list.

Though I haven't always stuck to it perfectly, I try to include top 25 lists only from national sources that also predict the conferences. That is one way to keep the job (of collecting the preseason consensus) at a manageable level. It also ensures that it will at least come close to fitting on your screen width-wise.

The preseason ranking is underestimating Notre Dame. They have ridiculous amounts of talent for next year. Bob Davie has head coaching experience with them from the times Holtz was out. Regardless he is a good coach. Re-evaluate your placement of Notre Dame.

Brian Scnails (8/97)

Response from the editor:

The preseason consensus has nothing to do with "my placement" of any team. It is based solely on the predictions of national preseason magazines -- none of which are written or published by me.

The last three years Notre Dame has been fairly high up in the preseason top ten (#6, #7, #4) and not in the final AP top ten (#19, #11, unranked). If I were an Irish fan, I would be more concerned about seeing the top ten in the final polls, than worried about #9 preseason (their position as of this writing) being too low. [Hey to Keith H.]

I like that you showed the accuracy of all these Preseason mags that I drop money into every season just to see them make the popular picks. This shows that these guys for all of their knowledge and exprtise can't predict the future better than everyone else. I think that it is interesting to see that these guys can make the picks on the big name schools. They all seem to have trouble picking the smaller conferences.

This tells me that these guys aren't really experts at all and that I could do the same, if not better. I for one do not put much stock in these guys becuase they have continually picked Texas Tech near the middle or bottom of the conference and Tech continues to finish in the upper half.

They always seem to go with the fashionable pick or the pick that is easy to make. You rarely see these guys make a bold pick becuase they are afraid that one of their peers will laugh at them for going out on a limb.

Chris Snead (8/97)

Response from the editor:

Anyone can "play along at home" and try to beat the magazines' predictions. Just write down your predictions around spring football time (which is about when the magazines' predictions are made), and score your own entry when the season is over. It might be difficult to beat every magazine for total score (all I-A conferences)... but most fans can probably do better than the magazines' average score for their favorite conference.

By the way, one must be careful when comparing "accuracy scores" between conferences with different numbers of teams. It's easier to pick two six-team divisions (worst possible score: 36) than to pick a ten-team conference (worst possible score: 50). I'm not sure that the magazines have the most trouble with the minor conferences; for example, the average "accuracy scores" for the MAC and Pac-10 (same number of teams, no divisions) have been reasonably close in most years.


Back to Preseason Index Up to Feedback Index Chris' Home Page