Computers, Humans, and Polling Disparities

Well, I for one welcome our new computer overlords. The BCS in its current form – ridiculous - is hitting the home stretch. At this point, no self-respecting fan of college football should be able to sing any praises of this system which proved to be experimental at best and lacking, limited, and lousy at worst. Systematic and complex ranking algorithms, whether human or CPU-based, have been both the calling card and scourge of college football for quite some time.

In a grand sense, there is no reason to believe computers over people, or people over computers, when people are really the cause of each measurement methodology. But one thing is for sure, it’s pretty tough to program a computer with regional bias, rose-coloring, ignorance, or slight. As the graph shows, someone always stands to gain from human perception in voting, and someone is usually getting punished.

So here is your BCS ranking red-meat hate served rare, Irish fans. In the current BCS rankings, Notre Dame has the biggest gap (shown by distance from zero upward) between their BCS computer ranking average (15), and their ultimate BCS ranking (23), a disparity of eight places. Louisville is getting the most warm, fleshy, human love: seven places better than the bots in the overall BCS. Oregon, Ohio State, Stanford, and Fresno State are each dialed in and falling on the equator with a zero, indicating their ranking is the same via human and robot.

BCS Computer Dispairty

Click on image for full-size view

In short, if you believe the (human) world is against ND and by the numbers they should be ranked much higher, and you love movies such as I, RobotTerminator 2, and War Games - well, this might be your week. Notre Dame received the least human love this week and it’s not even close.

-Irish Twin 2 (@HLS_Podcast)

About The Irish Twins
  • jimmyg41

    Computers can only spit out the information that HUMANS have programmed it to spit out, you’re the idiot!

    • http://twitter.com/BrianJMcKeown Brian McKeown (@BrianJMcKeown)

      No shit? I was told computers were magic instruments that are powered by hamsters and God-given to man so that we might better be able to calculate complex problems, streamline communication and watch porn. I wonder what other lies have been told to me?

    • NDtex

      Trying to find the part where one of our Irish Twins called anyone an idiot.

      /looks

      /keeps looking

      /crtl+f “idiot”

      Huh…just you. Weird. Thanks for reading! Let’s try comprehension next time!

    • http://rantingmayhem.wordpress.com Mayhem

      *YOUR

  • http://www.herloyalsons.com ID_10_T

    jimmy – from the article: “there is no reason to believe computers over people, or people over computers, when people are really the cause of each measurement methodology.”

  • http://gravatar.com/fjc1109 fjc1109

    Jimmy doesn’t get it. Must have graduated from a SEC school.

  • http://twitter.com/stogiesnbeer Mark (@stogiesnbeer)

    I feel like this has been the case for years, and especially this year. Maybe it’s just from the Mothership of ESPN, but I feel like there has been a huge backlash against ND this year for *daring* to be crushed by Bama in the title game last year. It’s definitely interesting to see the disparity between human and computer rankings, is this going to be a regular thing?

  • http://www.twitter.com/michianacomedy michiana comedy

    Before we start raging against a Notre Dame bias, let’s keep a few things in mind:

    1. Notre Dame has what is being more and more revealed to be a bad loss against Michigan, and by 11 points. “But but but! Tommy got picked off on in the end zone on a touchdown drive to make it a four-point game!” Glad you mentioned that:

    2. Pollsters don’t pay attention to games, as stupid as that sounds. So ND gets no credit for beating Navy with some guys on defense who weren’t on the two-deep, because it was “only by four points.” (I notice some ostensible ND fans, mostly on other sites, have also not noticed this plainly obvious fact.) Which brings me to:

    3. Style points matter to the shiny-object-desiring drooling imbeciles that vote in polls, and ND is just not going to be the team that runs it up on anybody. That’s not the way a Tommy Rees offense and a bend-but-don’t-break defense works against any team with a pulse, and our opponents are, for the most part, too good for that.

    4. That being said, BCS computer rankings are FORBIDDEN to factor in margin of victory AT ALL. Which is overcorrectingly stupid. If you look at other computer rankings that are allowed to consider MoV, ND looks worse: Sagarin, for example, has ND in the low 30s vs. #23 for the BCS version (and he’s the low guy, so that score got thrown out).

    5. And also, of course some of it is bias. They know just as we do that ND needs to be #14 to make a BCS bowl, and I’m quite certain that many pollsters don’t feel that ND “deserves” to be in one based on their last two performances, which by the way were so long ago that the only people associated with the program who are still there are Mike Collins and Touchdown Jesus. So they vote accordingly.