If you can read this, either the style sheet didn't load or you have an older browser that doesn't support style sheets. Try clearing your browser cache and refreshing the page.

(Yahoo)   As we prepare for one more year of the BCS, it decides to go out in style as one of its computers still ranks Notre Dame at number 1 this week   (sports.yahoo.com) divider line 33
    More: Fail, Notre Dame, BCS, AJ McCarron, strength of schedule, margin of victory, BCS championship, ColleyMatrix  
•       •       •

1806 clicks; posted to Sports » on 11 Jan 2013 at 5:47 AM (1 year ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



33 Comments   (+0 »)
   
View Voting Results: Smartest and Funniest

Archived thread
 
2013-01-11 06:09:34 AM
Well, notre dame was clearly the better team. Shame that it takes a computer to recognize it
 
2013-01-11 06:16:15 AM
To be fair, it's the computer on Lou Holtz' desk.
 
2013-01-11 06:52:41 AM
Done in two.
 
2013-01-11 06:56:05 AM
The the ranking also not take into account win-loss record? That would kind of seem like a fairly significant variable in this case.
 
2013-01-11 07:03:17 AM
To be fair, that computer's full top ten goes Notre Dame, Boston College, Georgetown, Holy Cross, St. John's, Gonzaga, Marquette, Providence, Villanova, and Siena.
 
2013-01-11 07:07:20 AM
It's because the computer isn't allowed to know the final scores, just who won. As far as the computer knows, maybe Notre Dame beat USC, Stanford, Oklahoma, Michigan and Michigan State by 50 points each, and then lost to Alabama by only 1 point. It's just being asked to rank a Notre Dame team that has only 1 loss (to Alabama) against an Alabama team that has only 1 loss (to Texas A&M). Notre Dame's one loss came against a tougher opponent than Alabama's one loss.
 
2013-01-11 08:21:40 AM

skazzytl: It's because the computer isn't allowed to know the final scores, just who won. As far as the computer knows, maybe Notre Dame beat USC, Stanford, Oklahoma, Michigan and Michigan State by 50 points each, and then lost to Alabama by only 1 point. It's just being asked to rank a Notre Dame team that has only 1 loss (to Alabama) against an Alabama team that has only 1 loss (to Texas A&M). Notre Dame's one loss came against a tougher opponent than Alabama's one loss.


but the rankings are farked because of the human element. Michigan was in the top 25 because they are michigan(only team whose prennial overrating rivals ND), their starting QB got moved to RB halfway through the year. USC was a joke. michigan state bad, wake terrible, bc terrible.

Say what you will, before the title game, ND's SOS was 24, Bama 3. After the game, Bama had the toughest SOS, Notre Dame 6th.
 
2013-01-11 08:50:56 AM

funktilious_j: skazzytl: It's because the computer isn't allowed to know the final scores, just who won. As far as the computer knows, maybe Notre Dame beat USC, Stanford, Oklahoma, Michigan and Michigan State by 50 points each, and then lost to Alabama by only 1 point. It's just being asked to rank a Notre Dame team that has only 1 loss (to Alabama) against an Alabama team that has only 1 loss (to Texas A&M). Notre Dame's one loss came against a tougher opponent than Alabama's one loss.

but the rankings are farked because of the human element. Michigan was in the top 25 because they are michigan(only team whose prennial overrating rivals ND), their starting QB got moved to RB halfway through the year. USC was a joke. michigan state bad, wake terrible, bc terrible.

Say what you will, before the title game, ND's SOS was 24, Bama 3. After the game, Bama had the toughest SOS, Notre Dame 6th.


Hey now let's not forget my Hokies in the 'perennially over ranked' dept.
 
2013-01-11 09:26:52 AM
USA!  USA!  USA!  That's what I'm talking about when I talk about America!

funktilious_j: Michigan was in the top 25 because they are michigan(only team whose prennial overrating rivals ND), their starting QB got moved to RB halfway through the year.


To be fair, he did have a nerve injury that rendered him even MORE incapable of throwing the ball, so it wasn't for sheer ineffectiveness.

kudayta: Hey now let's not forget my Hokies in the 'perennially over ranked' dept.


Oh yeah, they're perennial bed-shiatters, too.
 
2013-01-11 09:31:35 AM

skazzytl: It's because the computer isn't allowed to know the final scores, just who won. As far as the computer knows, maybe Notre Dame beat USC, Stanford, Oklahoma, Michigan and Michigan State by 50 points each, and then lost to Alabama by only 1 point. It's just being asked to rank a Notre Dame team that has only 1 loss (to Alabama) against an Alabama team that has only 1 loss (to Texas A&M). Notre Dame's one loss came against a tougher opponent than Alabama's one loss.


Yep, seems about right. A loss didn't really hurt Alabama either. Plus, like you mention, ND's loss came against a much higher ranked opponent.
 
2013-01-11 09:35:24 AM

funktilious_j: skazzytl: It's because the computer isn't allowed to know the final scores, just who won. As far as the computer knows, maybe Notre Dame beat USC, Stanford, Oklahoma, Michigan and Michigan State by 50 points each, and then lost to Alabama by only 1 point. It's just being asked to rank a Notre Dame team that has only 1 loss (to Alabama) against an Alabama team that has only 1 loss (to Texas A&M). Notre Dame's one loss came against a tougher opponent than Alabama's one loss.

but the rankings are farked because of the human element. Michigan was in the top 25 because they are michigan(only team whose prennial overrating rivals ND), their starting QB got moved to RB halfway through the year. USC was a joke. michigan state bad, wake terrible, bc terrible.

Say what you will, before the title game, ND's SOS was 24, Bama 3. After the game, Bama had the toughest SOS, Notre Dame 6th.


I think you forgot about some small school in Tallahassee.

/and Norman
 
2013-01-11 09:52:28 AM
So ND, which would be the 7th best team in the SEC, is better than Alabama? Thanks, BCS.
 
2013-01-11 10:26:22 AM
Can one of you nerds explain to me how a computer determines strength of schedule? It seems to me that at some point this becomes subjective; that a human must tell the computer that, e.g., SEC teams are on average better than MAC teams.

Are there enough interconference games during the first five or six weeks of the season for a computer to do a meaningful six-degrees-of-separation comparison of all Division I schools, and if so, is it economically feasible, or is it too tempting to simply plug in a set of weights based on conference, previous year's ranking, etc.?
 
2013-01-11 10:35:03 AM

Cornelius Dribble: Are there enough interconference games during the first five or six weeks of the season for a computer to do a meaningful six-degrees-of-separation comparison of all Division I schools


No, which is why there aren't BCS rankings for the first couple months of the season.
 
2013-01-11 11:03:13 AM
And the stupid asshole at the end still thinks margin of victory should still be used.

Because beating the shiat out of a D2 team 210-0 means you're the best team in the world.

farking idiots.
 
2013-01-11 11:08:55 AM

Cornelius Dribble: Can one of you nerds explain to me how a computer determines strength of schedule? It seems to me that at some point this becomes subjective; that a human must tell the computer that, e.g., SEC teams are on average better than MAC teams.


Okay, I am going to let you in on a little secret: HUMANS TELL THE COMPUTER EVERYTHING! There are not sentient AIs writing their own code. A person programmed the computer to calculate what it does, and a person fed it data. It has no volition. It simply churns through numbers. When someone says "the computers say X", what they ought to really be saying is "some humans used a really complex calculator to do some math, and the humans came up with X" Now, what the computer doesn't do is fudge the results to account for the "intangibles", which translates directly as "my favorite team is composed of infallible ubermenchen, and every other team is composed of syphilitic quintuple amputees that only score because the referees are paid to undermine my beloved clade of Aryan god-warriors". So the formula that is used by the computers keeps changing because stark numbers don't jibe with what we want to believe. In this case, they manipulated the formula so much that it spat out a totally spurious result because the eternal hunt to make a computer as stone-stupid as a hick in a sports bar worked.

All of this is still orders of magnitude better than basing rankings solely on the beliefs of people with a vested interest in the outcome, ie coaches, and the most panty-wetting of all fangirls, ie sports writers
 
2013-01-11 11:12:05 AM

Cornelius Dribble: Can one of you nerds explain to me how a computer determines strength of schedule? It seems to me that at some point this becomes subjective; that a human must tell the computer that, e.g., SEC teams are on average better than MAC teams.

Are there enough interconference games during the first five or six weeks of the season for a computer to do a meaningful six-degrees-of-separation comparison of all Division I schools, and if so, is it economically feasible, or is it too tempting to simply plug in a set of weights based on conference, previous year's ranking, etc.?


It's the Colley Matrix, so the entire methodology is laid out in full and there are a number of comparisons that can be done.
Compare top 5
Compare conferences
Methodology

I actually like the mathematical purity of the CM. It has its drawbacks that make it an imperfect measurement of "best team"--e.g., no accounting for home/road results, no margin-of-victory, etc.--but it's an elegant way of measuring "best schedule results" based on only win/loss data. It's completely objective, which is different than some computer rankings that have programmer "tweaks." Naturally, the almost completely subjective world of sports fandom tends to revile this sort of thing...

ND is #1 over Bama simply because overall ND played a slightly harder schedule and lost to a better team. This is hardly a crime against nature, just a calculation that doesn't feel right based on the Tide's pummeling of the Irish.

/fark the Irish
//hell, fark the Tide, too
 
2013-01-11 11:41:42 AM
Wait I thought Ohio St. was undefeated....
 
2013-01-11 11:48:40 AM

Igotnothing2: Wait I thought Ohio St. was undefeated....


They were. They also played one of the weakest schedules in the country.
 
2013-01-11 11:59:40 AM
"one of its computers still ranks Notre Dame at number 1 this week "

Deep Lou?
 
2013-01-11 12:15:36 PM
c-product.images.fansedge.com

Book it. Done.
 
2013-01-11 01:16:29 PM
Its one of the many computer polls that make up the BCS, which is why they use more than one, to correct fark ups like this one.

Oh and NDs schedule was not that hard no matter what people think, they just got lucky that their higher profile named teams had really shiatt years. Stanford was probably the best team they played and just got them early in the year, kind of like how Florida got lucky to play A&M early, play those games later in the season and both lose those games.
 
2013-01-11 01:17:23 PM

Fade2black: Igotnothing2: Wait I thought Ohio St. was undefeated....

They were. They also played one of the weakest schedules in the country.


And are cheaters so fark OSU
 
2013-01-11 03:17:11 PM

Fade2black: Igotnothing2: Wait I thought Ohio St. was undefeated....

They were. They also played one of the weakest schedules in the country.


What's interesting is this entire conversation may be valid in a year.
 
2013-01-11 04:30:32 PM

Fade2black: And the stupid asshole at the end still thinks margin of victory should still be used.

Because beating the shiat out of a D2 team 210-0 means you're the best team in the world.

farking idiots.


If you did take margin of victory into account, you'd probably graduate it in X number of layers, probably capped at 3 touchdowns. I think that would be fair... beat the #6 team in the middle of the season by 3 touchdowns? You are probably pretty damn good. Only beat them by a field goal or less? Pretty evenly matched. Struggle against a non-ranked team? Probably not as good as your ranking, and so forth...

Beyond three touchdowns, though, I think it's safe to say you are leagues better than the other team, so that would probably be the "max" bias applied to such a ranking.
 
2013-01-11 05:03:54 PM

Tell Me How My Blog Tastes: USA!  USA!  USA!  That's what I'm talking about when I talk about America!

funktilious_j: Michigan was in the top 25 because they are michigan(only team whose prennial overrating rivals ND), their starting QB got moved to RB halfway through the year.

To be fair, he did have a nerve injury that rendered him even MORE incapable of throwing the ball, so it wasn't for sheer ineffectiveness.

kudayta: Hey now let's not forget my Hokies in the 'perennially over ranked' dept.

Oh yeah, they're perennial bed-shiatters, too.


On the other hand, he hasn't thrown into double coverage since the injury, so he's got that working for him.
 
2013-01-11 05:49:56 PM

TabASlotB: Cornelius Dribble: Can one of you nerds explain to me how a computer determines strength of schedule? It seems to me that at some point this becomes subjective; that a human must tell the computer that, e.g., SEC teams are on average better than MAC teams.

Are there enough interconference games during the first five or six weeks of the season for a computer to do a meaningful six-degrees-of-separation comparison of all Division I schools, and if so, is it economically feasible, or is it too tempting to simply plug in a set of weights based on conference, previous year's ranking, etc.?

It's the Colley Matrix, so the entire methodology is laid out in full and there are a number of comparisons that can be done.
Compare top 5
Compare conferences
Methodology

I actually like the mathematical purity of the CM. It has its drawbacks that make it an imperfect measurement of "best team"--e.g., no accounting for home/road results, no margin-of-victory, etc.--but it's an elegant way of measuring "best schedule results" based on only win/loss data. It's completely objective, which is different than some computer rankings that have programmer "tweaks." Naturally, the almost completely subjective world of sports fandom tends to revile this sort of thing...

ND is #1 over Bama simply because overall ND played a slightly harder schedule and lost to a better team. This is hardly a crime against nature, just a calculation that doesn't feel right based on the Tide's pummeling of the Irish.

/fark the Irish
//hell, fark the Tide, too


Whoa, whoa. This is no place for logic and reason. Next you'll be telling me that some computer has Alabama ranked ahead of Texas A&M.
 
2013-01-11 06:58:48 PM

velvet_fog: So ND, which would be the 7th best team in the SEC, is better than Alabama? Thanks, BCS.


Herpa derp. The SEC is top heavy. It had five schools with a combined five league wins and the other nine went to bowls.

Not only that, but Notre Dame would get SEC-level recruits. Sure, on paper, they're getting good recruits now (after three years under Brian Kelly). I could only imagine what he'd do in an SEC school.

Disclaimer: No, I'm not a Notre Dame fan, nor am I anti-SEC. It is, without question, the best conference in college football. But it's very top heavy. It's not up to the same standards it has set, say, the last four years.

Even then, Florida was a paper tiger who was lucky as hell to play Texas A&M and got rolled in their bowl. LSU got Les Miles'd. Georgia and South Carolina both struggled against inferior opponents.

I'll give you Alabama, Georgia, and Texas A&M for sure as teams who could compete with anyone anywhere. South Carolina maybe. But, after that, the rest of the conference can be had.
 
2013-01-11 07:03:49 PM
I meant to say "lucky as hell to play Texas A&M the first week".
 
2013-01-11 07:12:13 PM
This... actually doesn't surprise me. One of the biggest weaknesses in the Colley method is that it doesn't take into account the results of individual games. In other words, it doesn't know that Alabama beat Notre Dame. All it knows is that Alabama played Notre Dame and that it ended the season 13-1. As far as it's concerned, that one loss was just as likely to have been against Western Carolina as Texas A&M. So while it may be a useful tool to compare regular seasons, it breaks down as soon as playoff games, which deliberately pit the best teams against each other to see who wins, are introduced -- and to be sure, the BCS title game is a playoff, albeit a small and crappy one.

(On the other hand, even with that glaring flaw, it hasn't been so bad. For all the years the BCS has been in effect, this season and the last are the only ones where the BCS champion wasn't Colley's number one. Incidentally, in both cases the champion had played an FCS übercupcake that year and the team ranked #1 hadn't. The lesson should be clear enough.)
 
2013-01-12 02:34:05 AM
So, for no reason aside from human bias in favor of the SEC, Alabama should have had all of the #1 rankings. Got it.
 
2013-01-12 08:02:55 AM
well no shiat, they each had one loss, and notre dame lost to a much higher ranked team!

/notre dame was definitely a top five team, they had a great year, but i don't know why they even bothered playing the bcs foregone conclusions championship game
 
2013-01-12 01:26:45 PM

skazzytl: It's because the computer isn't allowed to know the final scores, just who won. As far as the computer knows, maybe Notre Dame beat USC, Stanford, Oklahoma, Michigan and Michigan State by 50 points each, and then lost to Alabama by only 1 point. It's just being asked to rank a Notre Dame team that has only 1 loss (to Alabama) against an Alabama team that has only 1 loss (to Texas A&M). Notre Dame's one loss came against a tougher opponent than Alabama's one loss.


I'd argue that Alabama's one loss came against a tougher opponent than Notre Dame's only loss. Texas A&M beat Alabama. Notre Dame did not.

Logically expressed, it would be Texas A&M > Alabama > Notre Dame.
 
Displayed 33 of 33 comments

View Voting Results: Smartest and Funniest


This thread is closed to new comments.

Continue Farking
Submit a Link »






Report