Each week during the season I will pick out a few of the questions I receive and post my answers publicly. Send your questions here.
My apologies to everyone. The question and answer section just got completely away from me after Thanksgiving. I appreciate each and every email, and I still intend to answer, but it will be after the National Championship game before I have time. Thanks for your patience, and enjoy the Bowl Games.
Question- November 20, 2008
I have read that you give higher marks in your ratings for November wins than earlier wins. WOW!! What an intelligent way to put together a BCS Poll. If you were mis-quoted, please let me know.
I am talking Texas & Oklahoma. I was at the Texas/OU game & there was no doubt in any ones mind who was the better team. Did OU have to play Missouri ? NO!. Did OU have to play a 4 game back to back like Texas did? NO.
I think HEAD to HEAD tells the story. This stuff about what have you done lately completely ignores the over all picture.
Sorry Horace, but I must respectfully disagree. You seem strongly in favor of using head to head, and with that I 100% agree. My rankings reflect head to head more than ANY BCS computer. In a two way tie, HTH is an appropriate use of a tie breaker, but when it's a 3 way tie, it just becomes an endless loop. I'm sure you can see this. Assumming OU beats Tech (which is a mighty big if, lol) Texas should go because they beat OU, OU should go because they beat Tech....no, Tech should go because they beat Texas! There is no clear winner. If you move to strength of schedule, after this weekend's games Texas will have played the strongest SOS of the 3...but AFTER OU plays OSU, the Sooners will have played the hardest schedule.
My rankings are not a "what have you done for me lately" system at all....quite the opposite. It's just being true to head to head competition. Texas beat OU, Texas was ranked ahead of OU. Tech beat Texas, Tech was ranked ahead of Texas. If OU beats Tech, they will be ranked ahead of Tech. I think that is about as fair as you can get, don't you?
As far as being mis-quoted...no, it was a correct quote. Beating a #1 team in November is worth more than beating a #1team in October. Why? The point values are greater because teams have accomplished more. For instance Don't you think Texas was a more valuable opponent after they beat OU, Mizzou, and OSU? After all, they had accomplished more on the field than any other team. The Longhorns were on track to be my highest rated team in CF history. Don't you think they were worth more as an opponent on Nov 1then they were on September 1?
Question- November 13, 2008:
Hi Richard, Question about this week's ratings/rankings... Iowa defeated the #2 ranked team but only eight points were added to their rating. Other teams added double digit points to their respective ratings after a victorious Saturday, but their wins were not over squads in the Top 10 (e.g., TX Tech, Alabama, Utah). Why was Iowa's victory not valued as highly? Regards, Mark
Hello Mark.....that is a great observation. I certainly understand your confusion. In my system teams are penalized according to the # of losses on the season. So, even though Penn State had great value as an opponent, Iowa only receives a portion of that. They would have gained more if the Hawkeyes were 7-3, more if they were 8-2, and they would actually have jumped Penn State if both teams were now 9-1. I'm not sure which teams you are looking at, but my guess is that they had less losses on the season than Iowa. It's also possible they were allowed to jump the team they beat (by having an idential record). Iowa was blocked from doing that by this being November and having such a disparity in records (9-1 to 6-4). Hope this helps.
Question: November 12, 2008
First of all, I would like to thank for keeping up your website. I think it is a great resource for college football fans and I am a frequent visitor. I just have one question that I do not understand. Ohio State (8-2) is ahead of Penn State (9-1) in your Nov 9th poll. Yet Penn State beat Ohio State just two weeks ago at Ohio State. I realize that a loss to Iowa is certainly not impressive in the computers, but why wouldn't it be set it up to look at obvious things like that. If a team with the same record (or better in this case), beats the other team, common sense says that the team that won should be ranked higher. I was just wondering what your thoughts on why things like this happen in the computers (especially since this isn't the first time I've seen this - not necessarily in your poll, but the other computer rankings as well).
Thank you very much!
Hey Alex.....This is a perfect example of what I wrote about in my last Sporting News article. I don't think ANY computer poll (not even one as good as mine, lol), will ever be as good as an INFORMED UNBIASED human voter. Unfortunately, they are rare. I think computer polls should be used as a source of information, as a starting place for human voters to help them make sound choices. In this case, from a computer standpoint, my system is really good about honoring head to head competition UNTIL a team losses to another opponent. In this case the Nittany Lions were heavily penalized for losing to an Iowa team with 4 losses (and a #54 ranking), and conversely, the Hawkeyes didn't get a huge bump because they have 4 losses. With Ohio State beating a very good (#32) Northwestern, the #'s just were not supportive of keeping Penn State above Ohio State. If for instance Penn State had lost to Northwestern, they would not have fallen below Ohio State. Keep in mind that my system is designed to support one loss teams over two loss teams, so Penn State now has an advantage that will give them an opportunity to jump Ohio State given enough opponents to play. It may take thru the Bowl Games to do it, but perhaps even beating Michigan State might, it's too close to call at this point.
At any rate, the point I want to make is that situations like this are purely a numbers game. Penn State really suffered from that loss. The computer is basically saying Penn State's one loss to Iowa is worse than Ohio State's two losses to USC and Penn State. If I had a personal poll, Penn State would be right below Oklahoma, maybe even below Utah which would put them at #7 or #8. In this case the Coaches and sportswriters have Penn State about where I think they should be. I would also have Ohio State (and Georgia) lower, probably #11 and #12. But, my personal opinion is not what the BCS is interested in HA! Hope this helps.
November 4, 2008-
Very interesting week 10 poll results. A couple of observations:
1. Alabama wins there game 35-0, while Penn St doesn’t play. Doesn’t make sense that Alabama drops after a win while Penn St does not drop after a bye. Yet you dropped Alabama in week 7 from #1 to #2 when Alabama was in a bye week. Looks like a double-standard to me.
2. Alabama wins 35-0 and Texas loses, yet 9-0 Alabama remains below one-loss Texas. Again, doesn’t make sense.
I guess Alabama fans could see this coming after you dropped Alabama yet again in week 9 despite the fact that the team went to rival Tennessee and dominated. So my question is:
Have you lost your mind? Maybe your computer program is in an infinite do-loop. You mention SEC media bias in one of your Q&A postings. I can’t help but wonder if you might not be Big 12 biased (Oklahoma roots)?
By the way, you’ll have to email me if you want me to see your response. This was the first and probably last time I will go to your website. I was hoping to find some justification for your #4 ranking (impact on BCS numbers). Finding none, I am moving on to something with more reason.
Dare you to post this one.
Bob....I usually try to be supportive and understanding when responding to fans because I always remind myself about the passion people have for their favorite team, just as my passion is strong for college football overall. But, you dared me to post this one, so here goes, I'll give it to you just the way I see it. You are being quite unfair, and really, not very observant. If you're going to critcize, do your homework. Each week I supply more information than any other BCS computer website, including detailed comparisons of weekly point values, conference comparisons and a very time consumming master file listing EVERY GAME played alphabetical by team. I do all of that for the fans, as a convenience, to make it easier to understand the system and judge whether or not they they favor this type of format over another computer or a human poll. If they do, great, if not, I completely understand and respect that. My system is clearly not meant for everyone. I've said many times in my writings over the years, you'll either love it or hate it, there is usually no in between.
All you have to do each week is look at the POINT VALUE column and not just the RANK column of your favorite team and your questions will usually be answered. In this case you should be able to see that Penn State's point value did not change (as they had a bye week) and Alabama gained slightly from beating # 91 Arkansas State. Surely you are not telling me the Tide deserves more for beating a # 91 team are you? You should know scoring margin does not matter. 35-0 might as well be 200-0 or 21-20 as far as all of the BCS computers are concerned. But I'm sure you are a well enough educated college football fan to know that right?
You mentioned Alabama dropped in week 7 from #1 to #2. Again, if you had looked closer you would see Alabama's rating did not change due to the bye week (just like Penn State this week). The reason Alabama dropped was because Texas jumped ahead of them. That's what happens when point values change right? Are you telling me Texas beating Oklahoma did not deserve a change in their rating? I hope not. I'm sure you are smarter than that.
As far as Texas, no team in CF has accomplished as much as the Longhorns before their loss. Surely you recognize that has to count for something right? Texas lost a whopping 21.40 points last week. Alabama gained 1.65 for beating Arkansas State. The end result left Texas 1.42 points ahead of Alabama. Looks pretty logical and sounds pretty fair to me. If you have read the "Dynamics of the system" from the link on the front page you will already know that Alabama, as an undefeated team, now has a huge advantage over one loss teams the rest of the season. If you have not read it, then SHAME ON YOU because it just proves once again, all you want to do is complain without doing your homework.
So...you think I am biased towards the Big 12 huh? Geez.....you're killing me here!, lol! The Big 12 South has 4 teams in EVERYONE'S Top 10 and you think I'm biased because I have Oklahoma roots? Wow...you're really digging deep for that one! You should hear what the Oklahoma and Oklahoma State message boards are saying about me!!!!!!! , lol!
And finally.....as much as I hate to lose you as a visitor to my site...it's your loss. I have a very good informative, convenient site to read and navigate (even to the point of providing quick links to all the other computers). And honestly, since I get hundreds of thousands of visitors each year....I doubt you'll be missed.
OK....now that I have all that out of my system, I'll finish nice by saying this...you (and all CF fans) just need to let the season play out. Don't be so quick to judge before the product is finished at the end of the year. I didn't want to be so harsh, but you dared me, and you have to take responsibility for your actions. I'm ALWAYS up for a good dare. Don't worry about Alabama in my rankings. Trust me, all they have to do is win. With their schedule, the rankings will take care of themselves!
I noticed Florida received an increase in week 8 when they had a bye week. In casual glances, I have never seen an increase in a bye week for any other team. What was the reason for the increase? Jason
Jason..... Typically a team's rating does not change, but there are two exceptions. First, an undefeated team always remains ranked ahead of every team they have beaten (even in a bye week), and secondly, if a team (in this case Florida) experiences a bye week, and the team they just beat before the bye (in this case LSU) plays a high enough ranked team to jump Florida, the program recognizes this and prevents a leap that would jeopordize the integrity of the head to head rule which was instituted the week before. It is a rare occourrence, but it happens from time to time. I think last year it happened twice. Hope this helps.
First, let me state that I am an Ohio State fan. With that in mind, I have two questions for you:1) I see that you have Georgia as your national champion for 1942, even though Ohio State won the AP title that year. Georgia's rating for 1942 was 284.490; what was Ohio State's rating in 1942? Was Georgia's rating elevated because the Bulldogs won (and played) two more games that season?
2) In 1944, Ohio State (311.796 rating) finished just behind Army (312.074). Similarly, the Buckeyes barely edged out Penn State in 1968 by a score of 322.908 to 322.808. To me, the scores seem so close that a "tie" might be warranted (and some computer pollsters like FACT do award ties when the top two teams are very close). Do you ever award ties, and if so, what margin would indicate a statistical tie?Thanks in advance,
Answer: Hello CW....sorry for the delay. Very busy time as you can imagine. Ohio State’s rating was 281.290. The situation in 1942 when Georgia beat out the #1 AP team Ohio State was due more to strength of schedule than the extra two games the Bulldogs played. Georgia played and beat two #1 teams during the course of the season (#1 Alabama on October 31 and a very, very good #1 Georgia Tech team in the last regular season game).
As far as declaring ties, the system says any two teams within 3 points of one another playing on a neutral field would be declared a tossup. In that sense I suppose declaring joint national champions would be appropriate, however, technically, mathematically, I don't believe in ties. Even if one team is a fraction of a point better, I believe they should be declared #1. Just my personal philosophy. Truth is unless they play, we really never know the better team.
Question 9-27-08: Why is it when a lower tier Pac-10 team (Oregon St.) beats a top Pac-10 team (USC) it is described in the media as showing that the Pac 10 is a weak conference. Yet when the same thing happens in the SEC (Ole Miss beats Florida), the media, including the broadcasters of the game today, describes it as demonstrating the strength of the conference? How can the exact same thing happening mean one thing for one conference and the opposite for another conference? Kevin in Washington
Answer:: Excellent question! I think this would be described as a clear case of bias. Unfortunately the media is entitled to opinions of this nature and once a “buzz” is created, it’s near impossible to stop the blitz. I’m not saying the SEC is not the strongest conference, it is in my opinion, But Ole Miss beating Florida does not show a “strength”, it diminished the power rating of the conference overall. If there is any defense for the media it would be there is no guage for the strength of Oregon State versus Ole Miss as neither was a Top 25 team and that’s all the media uses as a judge. However, if you look closely at my rankings, you’ll notice Oregon State was a higher ranked opponent. As a result, the penalty for a loss to Florida was greater than that of USC and effectively did not hurt the PAC 10 as much as the SEC. Unfortunately the media does not have the tools available to make that type of distinction, unless of course they access my rankings before they vote. (highly recommended, lol).
Question: 9-16-08 You’re system seems flawed. It penalizes teams for bye weeks! Griffin in Missouri
Answer: Yes, it does create a temporary setback, but only temporary. Later in the season, it creates a boost to play when other teams have a bye. In the end, it all works equally with all the teams who play the same number of games. The only real advantage is playing in a championship game which creates an opportunity for an extra game that other teams do not have.
Question: 9-9-08 What’s up with your #’s? Your rankings are always way different than the other guys! Terry in Ohio
Answer: Yes, it is true, my #’s are different, but that is a good thing, not a bad. The reason is my formula takes into account the most recent game more heavily. As a result my rankings have more of a fluctation from one week to the next. Other systems tend to create more of an “average” rank on the season. As a result, the other BCS computers have more in common with one another, and I have more in common with the Coaches and Harris voters which would be the proper place to compare my numbers.