<?xml version="1.0" encoding="UTF-8"?>
<blog-post>
  <updated-at type="datetime">2009-01-07T16:24:00-05:00</updated-at>
  <title>Pollspeak: Humans vs. computers</title>
  <published-at type="datetime">2008-12-08T22:49:43-05:00</published-at>
  <comments-count type="integer">31</comments-count>
  <created-at type="datetime">2008-12-08T22:49:43-05:00</created-at>
  <comments-page>
    <current-page type="integer">1</current-page>
    <total-pages type="integer">4</total-pages>
    <per-page type="integer">10</per-page>
    <comments type="array">
      <comment>
        <quotable>
          <created-at>2008-12-10T19:36:42-05:00</created-at>
          <user>
            <image nil="true"></image>
            <comments-count type="integer">265</comments-count>
            <state>MO</state>
            <display-name>Mc Lovin</display-name>
            <city>Marshall                    </city>
            <id type="integer">631362</id>
          </user>
          <quoted-text>I work with computers all day long as a software engineer and can tell that there is nothing inherently accurate or reasonable about the so-called objective outbput from computers. I am sure you have heard the expression garbage in, garbage out. That is what the BCS computer algorithm is. As far as ranking teams, partly based on the rankings of the teams they beat, there are two problems. The biggest problem is that the quality of your opposition has no bearing per se on how good you are as a team. You could play the weakest schedule in the nation and still be the best team in the nation. The second problem is that the rankings themselves are based to a huge degree on the perception that people have about the overall quality of a conference. If people believed that the Big 12 was one of the weakest conferences in the nation, it is unlikely that OU would be in the BCS championship games. If you do not believe this, just ask 12-0 Utah. There is very little basis for doing an objective comparison between the Mountain West and Big 12 in calendar year 2008, yet it is clear that people rank 12-0 Utah lower than 11-1 Oklahoma largely on the basis of human bias and the biases builit into BCS computers about the relative strength of these teams conferences. Furthermore, even if it were true that Utah plays in the weakest conference in the nation, that sheds no light on the objective fact of whether or not they happen to be the best team in the nation.</quoted-text>
          <commentable-sequence type="integer">29</commentable-sequence>
          <body>the computers also rank Utah behind Oklahoma dude. They take all those factors into account including strength of schedule. So dude, like, what you said makes no sense.</body>
          <id type="integer">3612694</id>
        </quotable>
        <created-at>2009-01-07T16:24:00-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">390</comments-count>
          <state>MN</state>
          <display-name>SteveLD</display-name>
          <city>Saint Paul                  </city>
          <id type="integer">185352</id>
        </user>
        <quoted-text>the computers also rank Utah behind Oklahoma dude. They take all those factors into account including strength of schedule. So dude, like, what you said makes no sense.</quoted-text>
        <commentable-sequence type="integer">31</commentable-sequence>
        <body>It makes all the sense in the world.  My main point was that even if Utah played in the worst conference in the nation, that has absolutely no bearing on whether or not they are the best team in the nation.  Second, because conference primarily play games against one another, there is too little data on which to base meaningful comparisons between one conference and another.</body>
        <id type="integer">3897568</id>
      </comment>
      <comment>
        <quotable>
        </quotable>
        <created-at>2008-12-11T03:56:59-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">7</comments-count>
          <state>PA</state>
          <display-name>balok</display-name>
          <city>Clairton                    </city>
          <id type="integer">576615</id>
        </user>
        <quoted-text nil="true"></quoted-text>
        <commentable-sequence type="integer">30</commentable-sequence>
        <body>My question for all of the fans of computer rankings:  How many of you honestly believe that James Madison would beat Iowa by a touchdown (the Sagarin ELO-CHESS prediction)?  While it may be admirable that the BCS rules require the computer rankings to ignore margin of victory, does anyone honestly believe that a one-point victory margin indicates the same difference in quality between two teams as a three-touchdown victory margin?</body>
        <id type="integer">3615686</id>
      </comment>
      <comment>
        <quotable>
          <created-at>2008-12-10T18:32:52-05:00</created-at>
          <user>
            <image nil="true"></image>
            <comments-count type="integer">390</comments-count>
            <state>MN</state>
            <display-name>SteveLD</display-name>
            <city>Saint Paul                  </city>
            <id type="integer">185352</id>
          </user>
          <quoted-text>To counter that you NEED an objective system of measurement that won't be subject to gerryrigging the polls and trying to CREATE matchups in the BCS which is now what the polls are being used for. So, while the computers are not perfect and always need to be tweaked, if you use multiple computer formulas, the average of those should be pretty accurate. The computers this year accurately showed that Texas and Oklahoma played very tough schedules, while Alabama and USC did not. That kind of factor needs more emphasis and needs to counter balance human bias. Alabama got a huge bump up simply from beating an overranked (by the media) Clemson team and rode that false sense of a quality win to the number one ranking all year. The computers were drown out by the subjective biased media, which weren't able to filter out the false bump of the Clemson victory, as just one example.</quoted-text>
          <commentable-sequence type="integer">28</commentable-sequence>
          <body>I work with computers all day long as a software engineer and can tell that there is nothing inherently accurate or reasonable about the so-called objective outbput from computers.  I am sure you have heard the expression garbage in, garbage out.  That is what the BCS computer algorithm is.  

As far as ranking teams, partly based on the rankings of the teams they beat, there are two problems.  The biggest problem is that the quality of your opposition has no bearing per se on how good you are as a team.  You could play the weakest schedule in the nation and still be the best team in the nation.  The second problem is that the rankings themselves are based to a huge degree on the perception that people have about the overall quality of a conference.  If people believed that the Big 12 was one of the weakest conferences in the nation, it is unlikely that OU would be in the BCS championship games.  If you do not believe this, just ask 12-0 Utah.  There is very little basis for doing an objective comparison between the Mountain West and Big 12 in calendar year 2008, yet it is clear that people rank 12-0 Utah lower than 11-1 Oklahoma largely on the basis of human bias and the biases builit into BCS computers about the relative strength of these teams conferences.  Furthermore, even if it were true that Utah plays in the weakest conference in the nation, that sheds no light on the objective fact of whether or not they happen to be the best team in the nation.</body>
          <id type="integer">3612164</id>
        </quotable>
        <created-at>2008-12-10T19:36:42-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">265</comments-count>
          <state>MO</state>
          <display-name>Mc Lovin</display-name>
          <city>Marshall                    </city>
          <id type="integer">631362</id>
        </user>
        <quoted-text>I work with computers all day long as a software engineer and can tell that there is nothing inherently accurate or reasonable about the so-called objective outbput from computers. I am sure you have heard the expression garbage in, garbage out. That is what the BCS computer algorithm is. As far as ranking teams, partly based on the rankings of the teams they beat, there are two problems. The biggest problem is that the quality of your opposition has no bearing per se on how good you are as a team. You could play the weakest schedule in the nation and still be the best team in the nation. The second problem is that the rankings themselves are based to a huge degree on the perception that people have about the overall quality of a conference. If people believed that the Big 12 was one of the weakest conferences in the nation, it is unlikely that OU would be in the BCS championship games. If you do not believe this, just ask 12-0 Utah. There is very little basis for doing an objective comparison between the Mountain West and Big 12 in calendar year 2008, yet it is clear that people rank 12-0 Utah lower than 11-1 Oklahoma largely on the basis of human bias and the biases builit into BCS computers about the relative strength of these teams conferences. Furthermore, even if it were true that Utah plays in the weakest conference in the nation, that sheds no light on the objective fact of whether or not they happen to be the best team in the nation.</quoted-text>
        <commentable-sequence type="integer">29</commentable-sequence>
        <body>the computers also rank Utah behind Oklahoma dude. They take all those factors into account including strength of schedule. So dude, like, what you said makes no sense.</body>
        <id type="integer">3612694</id>
      </comment>
      <comment>
        <quotable>
          <created-at>2008-12-09T14:05:39-05:00</created-at>
          <user>
            <image nil="true"></image>
            <comments-count type="integer">350</comments-count>
            <state>EG</state>
            <display-name>Stuntman Mike</display-name>
            <city>Casablanca</city>
            <id type="integer">624219</id>
          </user>
          <quoted-text>They should weed out the computers entirely. The computers would be fine if the objective data on which they base rankings was capable of evaluating the relative strength of teams that play different schedules in different conferences. The problem is that there are far too many variables that computers cannot evaluate and this overwhelms the limited objective data available for accurately comparing teams that play in different conferences. To use an analogy from another field there is far too much noise and far too little signal.Under these circumstances, polls work better than computers because the best estimates that people have about the relative strength of teams, whatever subjective bias people may have, is often better than the so-called objectivity built into computers. If we eliminated the computers and told pollsters to rank teams based on their best estimates about who is best, we might avoid ridiculous situations such as occurred in the BCS title game last year where the BCS ranks Ohio State #1, but most people pick Florida to beat them in the championship game, and then Ohio State ends the season ranked 4th even though they lost to the #2 team in the nation. Similarly, we wouldn???t have Alabama ranked #1 with most people favoring Florida to beat them in the BCS championship game. The other point about the objectivity of computers is they are only as objective as the algorithm that human beings have built into them. The facts are that the bowls included in the BCS system were actually more competitive before the BCS system was in place than they are now. The margin of victory in the BCS bowls has been greater since they started the system than it was before they started the system. For all the claims made that the BCS system would me college football more competitive, the truth is it has made it less competitive. My personal preference would be to abandon the idea of a college football championship, including one based purely on polls, until such time as it is possible to have a playoff among all worthy contenders.Now, Stuntman Mike, you have accused me and others of fabricating data and, as you put it, throwing it out without being able to back it up. Show me one comment I have made concerning a matter of fact that is in any way inaccurate or which cannot be substantiated. In the absence of that you owe me an apology.</quoted-text>
          <commentable-sequence type="integer">15</commentable-sequence>
          <body>you are contradicting yourself. On the one hand you denigrate the BCS system for the bowl matchups, state they are less competitive (still without proof), and state that the human pollsters don't know what they are doing. Yet you then turn around and state that they should eliminate computers altogether.

The problem is that human pollsters DON'T rank the teams on who did best on the field. They have built in biases and prejudices. They start with preconcieved notions left over from last year (Which is why I am for NO polls that are included in the BCS formula to do ANY ranking before the 4th week of the season). THey don't have time to look at schedules and strength of schedule. They don't watch all the games. THe coache's poll is  a joke, since coach's only really see teams they played and teams they are going to play the next week. Most media people get most of their knowledge of what happened by watching College Gamedays highlight show.

To counter that you NEED an objective system of measurement that won't be subject to gerryrigging the polls and trying to CREATE matchups in the BCS which is now what the polls are being used for. So, while the computers are not perfect and always need to be tweaked, if you use multiple computer formulas, the average of those should be pretty accurate. The computers this year accurately showed that Texas and Oklahoma played very tough schedules, while Alabama and USC did not. That kind of factor needs more emphasis and needs to counter balance human bias. Alabama got a huge bump up simply from beating an overranked (by the media) Clemson team and rode that false sense of a quality win to the number one ranking all year. The computers were drown out by the subjective biased media, which weren't able to filter out the false bump of the Clemson victory, as just one example.

So you are completely off the mark and contradictory at the same time in your assessment of the value of the computers in the formula.</body>
          <id type="integer">3594816</id>
        </quotable>
        <created-at>2008-12-10T18:32:52-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">390</comments-count>
          <state>MN</state>
          <display-name>SteveLD</display-name>
          <city>Saint Paul                  </city>
          <id type="integer">185352</id>
        </user>
        <quoted-text>To counter that you NEED an objective system of measurement that won't be subject to gerryrigging the polls and trying to CREATE matchups in the BCS which is now what the polls are being used for. So, while the computers are not perfect and always need to be tweaked, if you use multiple computer formulas, the average of those should be pretty accurate. The computers this year accurately showed that Texas and Oklahoma played very tough schedules, while Alabama and USC did not. That kind of factor needs more emphasis and needs to counter balance human bias. Alabama got a huge bump up simply from beating an overranked (by the media) Clemson team and rode that false sense of a quality win to the number one ranking all year. The computers were drown out by the subjective biased media, which weren't able to filter out the false bump of the Clemson victory, as just one example.</quoted-text>
        <commentable-sequence type="integer">28</commentable-sequence>
        <body>I work with computers all day long as a software engineer and can tell that there is nothing inherently accurate or reasonable about the so-called objective outbput from computers.  I am sure you have heard the expression garbage in, garbage out.  That is what the BCS computer algorithm is.  

As far as ranking teams, partly based on the rankings of the teams they beat, there are two problems.  The biggest problem is that the quality of your opposition has no bearing per se on how good you are as a team.  You could play the weakest schedule in the nation and still be the best team in the nation.  The second problem is that the rankings themselves are based to a huge degree on the perception that people have about the overall quality of a conference.  If people believed that the Big 12 was one of the weakest conferences in the nation, it is unlikely that OU would be in the BCS championship games.  If you do not believe this, just ask 12-0 Utah.  There is very little basis for doing an objective comparison between the Mountain West and Big 12 in calendar year 2008, yet it is clear that people rank 12-0 Utah lower than 11-1 Oklahoma largely on the basis of human bias and the biases builit into BCS computers about the relative strength of these teams conferences.  Furthermore, even if it were true that Utah plays in the weakest conference in the nation, that sheds no light on the objective fact of whether or not they happen to be the best team in the nation.</body>
        <id type="integer">3612164</id>
      </comment>
      <comment>
        <quotable>
        </quotable>
        <created-at>2008-12-10T08:43:39-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">3</comments-count>
          <state>XX</state>
          <display-name>Zachi</display-name>
          <city>No City</city>
          <id type="integer">630033</id>
        </user>
        <quoted-text nil="true"></quoted-text>
        <commentable-sequence type="integer">27</commentable-sequence>
        <body>i don't know if anybody has already gotten this mail, i hope i'm not boring you to death with this, but i CAN see the point here regarding the BCS discussion.

BCS Declares Germany Winner of WWII

After determining the Big-12 championship game participants the BCS
computers were put to work on other major contests and today the BCS
declared Germany to be the winner of World War II.

&amp;quot;Germany put together an incredible number of victories beginning with the
annexation of Austria and the Sudetenland and continuing on into conference
play with defeats of Poland, France, Norway, Sweden, Denmark, Belgium and
the Netherlands&amp;quot; stated Goober Von Nuesbaum, chief architect of the BCS
computer software. &amp;quot;Their only losses came against the US and Russia;
however considering their entire body of work--including an incredibly
tough Strength of Schedule and significant margin of victories in most
battles--our computers deemed them worthy of the #1 ranking.&amp;quot;
Questioned about the #2 ranking of the United States the BCS stated &amp;quot;The US
only had two major victories--Japan and Germany. The computer models,
unlike humans, aren't influenced by head-to-head contests--they consider
each contest to be only a single, equally-weighted event....</body>
        <id type="integer">3602849</id>
      </comment>
      <comment>
        <quotable>
        </quotable>
        <created-at>2008-12-10T00:15:54-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">236</comments-count>
          <state>NC</state>
          <display-name>BlakCass</display-name>
          <city>Charlotte</city>
          <id type="integer">377354</id>
        </user>
        <quoted-text nil="true"></quoted-text>
        <commentable-sequence type="integer">26</commentable-sequence>
        <body>Finally,  someone with some sense! themathnerd hit it on the head. As much as I don't like (not hate) the BCS, it is not their fault that Ok is in the title game as much as it is the Big 12's fault. To Texas fans, you guys have a legitimate gripe; however, directing that gripe at the BCS or Florida is not the way to go. It is your own league that failed you.</body>
        <id type="integer">3601835</id>
      </comment>
      <comment>
        <quotable>
          <created-at>2008-12-09T17:10:47-05:00</created-at>
          <user>
            <image nil="true"></image>
            <comments-count type="integer">173</comments-count>
            <state>XX</state>
            <display-name>ronnieb789</display-name>
            <city>No City</city>
            <id type="integer">626859</id>
          </user>
          <quoted-text>As much as I hate the BCS, Texas not being in the title game is NOT the BCS's fault. BCS merely put the SEC champ vs the Big 12 champ in the Title game. The blame should go to the Big 12 for choosing Oklahoma over Texas for the Big 12 title game against Missouri.Big 12 is the ONLY BCS conference where head-to-head doesn't determine the tie breaker. They of course based their selection on the BCS rankings but the purpose of BCS was to determine who the #1 &amp;amp; #2 are, not the conference champions. The Big 12 deserves the blame for this mess, not the BCS.</quoted-text>
          <commentable-sequence type="integer">24</commentable-sequence>
          <body>You nimrod. How do you use head to head in determining a three way tie where the three teams beat eachother? Oh, it was a two way tie because the BCS ranked one of the teams 4 spots lower? So why doesn't the BCS also determine the winner? Grow a brain.   

If anything, the north vs. south format should be looked at. They should all play eachother and it will be a lot harder to wind up with a three way.  If it happens anyway, then use bcs or head to head in sending the 2 best teams from the conference regardless of north vs south.

But really, y'all need to let that head to head thing go. And I'm not even sure what &amp;quot;mess&amp;quot; you're talking about. Human and computers put OU in front.  The only mess left is in Texas it seems.</body>
          <id type="integer">3598151</id>
        </quotable>
        <created-at>2008-12-09T17:42:13-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">8</comments-count>
          <state>CA</state>
          <display-name>themathnerd</display-name>
          <city>Los Angeles                 </city>
          <id type="integer">610120</id>
        </user>
        <quoted-text>You nimrod. How do you use head to head in determining a three way tie where the three teams beat eachother? Oh, it was a two way tie because the BCS ranked one of the teams 4 spots lower? So why doesn't the BCS also determine the winner? Grow a brain.   If anything, the north vs. south format should be looked at. They should all play eachother and it will be a lot harder to wind up with a three way.  If it happens anyway, then use bcs or head to head in sending the 2 best teams from the conference regardless of north vs south.But really, y'all need to let that head to head thing go. And I'm not even sure what &quot;mess&quot; you're talking about. Human and computers put OU in front.  The only mess left is in Texas it seems.</quoted-text>
        <commentable-sequence type="integer">25</commentable-sequence>
        <body>Ronnieb789:

SEC West - If there's a 3 way tie, the lowest ranked team is eliminated. Head-to-head determines who goes among the Top 2.
SEC East - If there's a 3 way tie, the lowest ranked team is eliminated. Head-to-head determines who goes among the Top 2.
ACC Atlantic - If there's a 3 way tie, the lowest ranked team is eliminated. Head-to-head determines who goes among the Top 2.
ACC Coastal - If there's a 3 way tie, the lowest ranked team is eliminated. Head-to-head determines who goes among the Top 2. - If there's a 3 way tie, the lowest ranked team is eliminated. Head-to-head determines who goes among the Top 2.
Pac 10 - If there's a 3 way tie, the lowest ranked team is eliminated. Head-to-head determines who goes among the Top 2.

Maybe they're all nimrods but if Big 12 went by their criteria, Texas Tech (lowest ranked among 3) would be eliminated and Texas would win the head-to-head. Do some research next time.</body>
        <id type="integer">3598684</id>
      </comment>
      <comment>
        <quotable>
          <created-at>2008-12-09T16:28:07-05:00</created-at>
          <user>
            <image nil="true"></image>
            <comments-count type="integer">8</comments-count>
            <state>CA</state>
            <display-name>themathnerd</display-name>
            <city>Los Angeles                 </city>
            <id type="integer">610120</id>
          </user>
          <quoted-text nil="true"></quoted-text>
          <commentable-sequence type="integer">23</commentable-sequence>
          <body>As much as I hate the BCS, Texas not being in the title game is NOT the BCS's fault. BCS merely put the SEC champ vs the Big 12 champ in the Title game. The blame should go to the Big 12 for choosing Oklahoma over Texas for the Big 12 title game against Missouri.

Big 12 is the ONLY BCS conference where head-to-head doesn't determine the tie breaker. They of course based their selection on the BCS rankings but the purpose of BCS was to determine who the #1 &amp;amp; #2 are, not the conference champions.  The Big 12 deserves the blame for this mess, not the BCS.</body>
          <id type="integer">3597470</id>
        </quotable>
        <created-at>2008-12-09T17:10:47-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">173</comments-count>
          <state>XX</state>
          <display-name>ronnieb789</display-name>
          <city>No City</city>
          <id type="integer">626859</id>
        </user>
        <quoted-text>As much as I hate the BCS, Texas not being in the title game is NOT the BCS's fault. BCS merely put the SEC champ vs the Big 12 champ in the Title game. The blame should go to the Big 12 for choosing Oklahoma over Texas for the Big 12 title game against Missouri.Big 12 is the ONLY BCS conference where head-to-head doesn't determine the tie breaker. They of course based their selection on the BCS rankings but the purpose of BCS was to determine who the #1 &amp;amp; #2 are, not the conference champions. The Big 12 deserves the blame for this mess, not the BCS.</quoted-text>
        <commentable-sequence type="integer">24</commentable-sequence>
        <body>You nimrod. How do you use head to head in determining a three way tie where the three teams beat eachother? Oh, it was a two way tie because the BCS ranked one of the teams 4 spots lower? So why doesn't the BCS also determine the winner? Grow a brain.   

If anything, the north vs. south format should be looked at. They should all play eachother and it will be a lot harder to wind up with a three way.  If it happens anyway, then use bcs or head to head in sending the 2 best teams from the conference regardless of north vs south.

But really, y'all need to let that head to head thing go. And I'm not even sure what &amp;quot;mess&amp;quot; you're talking about. Human and computers put OU in front.  The only mess left is in Texas it seems.</body>
        <id type="integer">3598151</id>
      </comment>
      <comment>
        <quotable>
        </quotable>
        <created-at>2008-12-09T16:28:07-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">8</comments-count>
          <state>CA</state>
          <display-name>themathnerd</display-name>
          <city>Los Angeles                 </city>
          <id type="integer">610120</id>
        </user>
        <quoted-text nil="true"></quoted-text>
        <commentable-sequence type="integer">23</commentable-sequence>
        <body>As much as I hate the BCS, Texas not being in the title game is NOT the BCS's fault. BCS merely put the SEC champ vs the Big 12 champ in the Title game. The blame should go to the Big 12 for choosing Oklahoma over Texas for the Big 12 title game against Missouri.

Big 12 is the ONLY BCS conference where head-to-head doesn't determine the tie breaker. They of course based their selection on the BCS rankings but the purpose of BCS was to determine who the #1 &amp;amp; #2 are, not the conference champions.  The Big 12 deserves the blame for this mess, not the BCS.</body>
        <id type="integer">3597470</id>
      </comment>
      <comment>
        <quotable>
        </quotable>
        <created-at>2008-12-09T16:26:45-05:00</created-at>
        <user>
          <image nil="true"></image>
          <comments-count type="integer">254</comments-count>
          <state>OH</state>
          <display-name>UTAHFANNATIC</display-name>
          <city>Columbus                    </city>
          <id type="integer">569156</id>
        </user>
        <quoted-text nil="true"></quoted-text>
        <commentable-sequence type="integer">22</commentable-sequence>
        <body>Not a big fan of the SEC and I hope my team crushes Alabama, but I can't imagine any fan of the Big Ten talking smack to an SEC fan. Seriously I can't see it.</body>
        <id type="integer">3597468</id>
      </comment>
    </comments>
    <total-entries type="integer">31</total-entries>
  </comments-page>
  <body>&lt;div class=&quot;photo_container image_right&quot;&gt;
&lt;div class=&quot;photo&quot;&gt;&lt;img src=&quot;http://img.fannation.com/upload/si_blog_post_images/oklahoma-sooners.si.jpg&quot; alt=&quot;Oklahoma-sooners&quot; width=&quot;298&quot; height=&quot;291&quot; /&gt;&lt;/div&gt;
&lt;div class=&quot;photo_attributes&quot; style=&quot;text-align:right;&quot;&gt;The computers and coaches ranked Oklahoma No. 1.&lt;br /&gt;David E. Klutho/SI&lt;/div&gt;
&lt;/div&gt;
&lt;p class=&quot;MsoNormal&quot; style=&quot;margin: 0in 0in 10pt;&quot;&gt;&lt;span style=&quot;font-size: small; font-family: Calibri;&quot;&gt;&lt;em&gt;By Hugh Falk, &lt;/em&gt;&lt;/span&gt;&lt;a href=&quot;http://www.pollspeak.com/&quot;&gt;&lt;span style=&quot;font-size: small; color: #800080; font-family: Calibri;&quot;&gt;&lt;em&gt;Pollspeak.com&lt;/em&gt;&lt;/span&gt;&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;Oklahoma, Florida or Texas?&amp;nbsp; That was the question.&amp;nbsp;All three made very good cases for the BCS Championship Game, but only two could go, and it was up to the BCS formula to decide which two.&amp;nbsp;In reality, human voters have the most power in the BCS, and some might think too much.&amp;nbsp;&amp;nbsp;This year the &quot;some&quot; are Texas fans.&lt;/p&gt;
&lt;p&gt;The BCS may have put together the best 1 vs. 2 matchup that most people wanted to see.&amp;nbsp;Few wanted a rematch of Oklahoma and Texas (certainly most of the voters didn't).&amp;nbsp; So that means Florida had to get in, and Harris Interactive voters ensured that by ranking the Gators No. 1.&amp;nbsp;Many voters may think the Gators are the best team in the country, but a few might also know they needed to offset the computers who don't think as highly of Florida. Billingsley (as expected) is the only computer that &lt;a href=&quot;http://pollspeak.com/pollstalker/pollstalker.php?s=5&amp;amp;p=8&amp;amp;t1=27&amp;amp;t2=94&amp;amp;v=405&amp;amp;w=16&amp;amp;r=T&amp;amp;o1=diff&quot;&gt;ranked Florida over Texas&lt;/a&gt;.&amp;nbsp;Most computers rank Florida 4th, and Massey ranks the Gators 5th. The computers also don't think much of the SEC this year --Sagarin's ranks the SEC 3rd strongest and Anderson/Hester ranks them 4th behind the Big East.&amp;nbsp;However, if the computers had more power, it might be Oklahoma and Texas in the title game.&lt;/p&gt;
&lt;p&gt;So if the Gators had to be in the title game to prevent a rematch, it really came down to who would be the representative from the Big 12:&amp;nbsp;Oklahoma, Texas or Texas Tech.&amp;nbsp; Every computer and the Coaches' Poll ranks &lt;a href=&quot;http://pollspeak.com/pollstalker/pollstalker.php?s=5&amp;amp;p=8&amp;amp;t1=75&amp;amp;v=405&amp;amp;w=16&amp;amp;r=T&quot;&gt;Oklahoma No. 1&lt;/a&gt;.&amp;nbsp;So there was no hope for Texas or Texas Tech unless they were able to stay ahead of the Gators.&amp;nbsp; It was close, but the human voters (and likely the fans they represent) got their wish.&amp;nbsp;The BCS lowered the weight of the computer ratings in their formula for situations just like this.&amp;nbsp;So it worked out this year (except for Longhorn fans).&amp;nbsp;&lt;/p&gt;
&lt;p&gt;It's interesting to note that if the BCS allowed computers to account for scoring margins, it may have made the matchup easier.&amp;nbsp;For example, Sagarin ranks Florida 4th in its ELO_CHESS rankings, which are used by the BCS and mandated not to use margin of victory as a factor.&amp;nbsp;However, the Gators are ranked 2nd in Sagarin's standard ratings, which do account for margin of victory.&lt;/p&gt;
&lt;p&gt;So are the computers really adding anything to the formula?&amp;nbsp;Yes, but not much.&amp;nbsp;It would have to be a very, very tight race for the computers to make a difference at the top. There is an example of computers making a small difference at the bottom -- &lt;a href=&quot;http://pollspeak.com/pollstalker/pollstalker.php?r=T&amp;amp;s=5&amp;amp;p=8&amp;amp;w=16&amp;amp;t1=13&amp;amp;t2=0&amp;amp;v=405&quot;&gt;Boston College&lt;/a&gt;.&amp;nbsp;Ranked out of the top 25 in both human polls, BC was lifted to No. 24 on the strength of its computer ratings.&amp;nbsp;However, for every example of computer relevance, there are two examples of irrelevance.&amp;nbsp;This week it is &lt;a href=&quot;http://pollspeak.com/pollstalker/pollstalker.php?r=T&amp;amp;s=5&amp;amp;p=8&amp;amp;w=16&amp;amp;t1=71&amp;amp;t2=0&amp;amp;v=405&quot;&gt;Northwestern&lt;/a&gt;, which received no computer points, yet still managed to be ranked 23rd, and &lt;a href=&quot;http://pollspeak.com/pollstalker/pollstalker.php?r=T&amp;amp;s=5&amp;amp;p=8&amp;amp;w=16&amp;amp;t1=58&amp;amp;t2=0&amp;amp;v=405&quot;&gt;Mississippi&lt;/a&gt;, which only needed one computer ranking from the Billingsley Report to hang on to the 25th spot.&amp;nbsp;It's worth noting that Northwestern received its highest ranking (No. 20) from the Coaches' Poll, and as &lt;a href=&quot;http://pollspeak.com/football_2008_pre.htm#Big_10_Has_The_Edge_In_Coaches_Poll&quot;&gt;we pointed out&lt;/a&gt;, the Big Ten actually has a slight voting advantage in that poll.&lt;/p&gt;
&lt;p&gt;The big poll news this week is the release of the Coaches' and Harris Interactive ballots.&amp;nbsp;This is the one week of the year we get to see how every person voted.&amp;nbsp; I'll have some detailed analysis next week, but in the meantime, you can use Pollstalker to see the &lt;a href=&quot;http://pollspeak.com/pollstalker/pollstalker.php?r=F&amp;amp;s=5&amp;amp;p=15&amp;amp;w=16&amp;amp;t1=0&amp;amp;t2=0&amp;amp;v=0&quot;&gt;Harris Interactive ballots here &lt;/a&gt;and the Coaches' ballots will be added later tonight.&lt;/p&gt;
&lt;p&gt;For more poll analysis, go to &lt;a href=&quot;http://www.pollspeak.com/&quot;&gt;pollspeak.com.&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;&amp;nbsp;&lt;/p&gt;</body>
  <blogger>
    <image nil="true"></image>
    <comments-count type="integer">0</comments-count>
    <state>CT</state>
    <display-name>SI.com's Blog</display-name>
    <city>New Hartford                </city>
    <id type="integer">99587</id>
  </blogger>
  <id type="integer">30441</id>
  <intro>&lt;div class=&quot;photo_container image_right&quot;&gt;
&lt;div class=&quot;photo&quot;&gt;&lt;img src=&quot;http://img.fannation.com/upload/si_blog_post_images/oklahoma-sooners.si.jpg&quot; alt=&quot;Oklahoma-sooners&quot; width=&quot;298&quot; height=&quot;291&quot; /&gt;&lt;/div&gt;
&lt;div class=&quot;photo_attributes&quot; style=&quot;text-align:right;&quot;&gt;The computers and coaches ranked Oklahoma No. 1.&lt;br /&gt;David E. Klutho/SI&lt;/div&gt;
&lt;/div&gt;
&lt;p class=&quot;MsoNormal&quot; style=&quot;margin: 0in 0in 10pt;&quot;&gt;&lt;span style=&quot;font-size: small; font-family: Calibri;&quot;&gt;&lt;em&gt;By Hugh Falk, &lt;/em&gt;&lt;/span&gt;&lt;a href=&quot;http://www.pollspeak.com/&quot;&gt;&lt;span style=&quot;font-size: small; color: #800080; font-family: Calibri;&quot;&gt;&lt;em&gt;Pollspeak.com&lt;/em&gt;&lt;/span&gt;&lt;/a&gt;&lt;/p&gt;
&lt;p&gt;Oklahoma, Florida or Texas?&amp;nbsp; That was the question.&amp;nbsp;All three made very good cases for the BCS Championship Game, but only two could go, and it was up to the BCS formula to decide which two.&amp;nbsp;In reality, human voters have the most power in the BCS, and some might think too much.&amp;nbsp;&amp;nbsp;This year the &quot;some&quot; are Texas fans.&lt;/p&gt;
&lt;p&gt;The BCS may have put together the best 1 vs. 2 matchup that most people wanted to see.&amp;nbsp;Few wanted a rematch of Oklahoma and Texas (certainly most of the voters didn't).&amp;nbsp; So that means Florida had to get in, and Harris Interactive voters ensured that by ranking the Gators No. 1.&amp;nbsp;Many voters may think the Gators are the best team in the country, but a few might also know they needed to offset the computers who don't think as highly of Florida. Billingsley (as expected) is the only computer that &lt;a href=&quot;http://pollspeak.com/pollstalker/pollstalker.php?s=5&amp;amp;p=8&amp;amp;t1=27&amp;amp;t2=94&amp;amp;v=405&amp;amp;w=16&amp;amp;r=T&amp;amp;o1=diff&quot;&gt;ranked Florida over Texas&lt;/a&gt;.&amp;nbsp;Most computers rank Florida 4th, and Massey ranks the Gators 5th. The computers also don't think much of the SEC this year --Sagarin's ranks the SEC 3rd strongest and Anderson/Hester ranks them 4th behind the Big East.&amp;nbsp;However, if the computers had more power, it might be Oklahoma and Texas in the title game.&lt;/p&gt;
&lt;p&gt;So if the Gators had to be in the title game to prevent a rematch, it really came down to who would be the representative from the Big 12:&amp;nbsp;Oklahoma, Texas or Texas Tech.&amp;nbsp; Every computer and the Coaches' Poll ranks &lt;a href=&quot;http://pollspeak.com/pollstalker/pollstalker.php?s=5&amp;amp;p=8&amp;amp;t1=75&amp;amp;v=405&amp;amp;w=16&amp;amp;r=T&quot;&gt;Oklahoma No. 1&lt;/a&gt;.&amp;nbsp;So there was no hope for Texas or Texas Tech unless they were able to stay ahead of the Gators.&amp;nbsp; It was close, but the human voters (and likely the fans they represent) got their wish.&amp;nbsp;The BCS lowered the weight of the computer ratings in their formula for situations just like this.&amp;nbsp;So it worked out this year (except for Longhorn fans).&amp;nbsp;&lt;/p&gt;
&lt;p&gt;&amp;nbsp;&lt;/p&gt;</intro>
</blog-post>
