Mark Farley Thinks His 5-Loss Team Should Finish Top-5

[Mvemjsunpx]

http://www.sportsnetwork.com/merge/tsnf ... 1_2014.htm

[HI54UNI]

Mark Farley is an idiot. And not because he thinks UNI should be a top 5 team. He's just an idiot.

[JohnStOnge]

If you go by power ratings, he's right.

Sagarin right now:

1. North Dakota State
2. Illinois State
3. Northern Iowa
4. South Dakota State
5. Coastal Carolina

Massey right now:

1. North Dakota State
2. Illinois State
3. Northern Iowa
4. South Dakota State
5. New Hampshire

They were the #3 team in a conference that was WAY heads and shoulders above any other FCS conference this season. They lost by 8 at Iowa and by 3 at Hawaii. Then during the regular season they lost by 1 on the road and by 3 at home to two MVFC playoff teams. Oh and by the way they beat both of the participants in the upcoming national championship game. Then they lost a rematch to one of those participants in the playoffs.

If you consider their schedule difficulty, who they beat, and who they lost to I don't think putting them in the top five is unreasonable at all. You have to ask yourself how other teams you would put in the top five instead of them would've done against that schedule. Again: Their slate included three games agains the two teams in the national championship game, a game at Iowa, a game at Hawaii, and three other games against FCS playoff teams.

[D1B]

If you go by power ratings, he's right.

Sagarin right now:

1. North Dakota State
2. Illinois State
3. Northern Iowa
4. South Dakota State
5. Coastal Carolina

Massey right now:

1. North Dakota State
2. Illinois State
3. Northern Iowa
4. South Dakota State
5. New Hampshire

They were the #3 team in a conference that was WAY heads and shoulders above any other FCS conference this season. They lost by 8 at Iowa and by 3 at Hawaii. Then during the regular season they lost by 1 on the road and by 3 at home to two MVFC playoff teams. Oh and by the way they beat both of the participants in the upcoming national championship game. Then they lost a rematch to one of those participants in the playoffs.

If you consider their schedule difficulty, who they beat, and who they lost to I don't think putting them in the top five is unreasonable at all. You have to ask yourself how other teams you would put in the top five instead of them would've done against that schedule. Again: Their slate included three games agains the two teams in the national championship game, a game at Iowa, a game at Hawaii, and three other games against FCS playoff teams.

The author of the article agreed too.

[Mvemjsunpx]

UNI should be a top-5 team… if they only played all their games in the UNI-Dome.

I also like how the article mentions the two home wins over the finalists, but kinda glosses over the fact they got slaughtered when they played one of them again on the road.

[D1B]

UNI should be a top-5 team… if they only played all their games in the UNI-Dome.

I also like how the article mentions the two home wins over the finalists, but kinda glosses over the fact they got slaughtered when they played one of them again on the road.

Easy fellah. UNI sucks. Ok?

Send your excellent thoughts to the author and the ratings services. I'm sure they'll appreciate your input.

[Mvemjsunpx]

UNI should be a top-5 team… if they only played all their games in the UNI-Dome.

I also like how the article mentions the two home wins over the finalists, but kinda glosses over the fact they got slaughtered when they played one of them again on the road.

Easy fellah. UNI sucks. Ok?

Send your excellent thoughts to the author and the ratings services. I'm sure they'll appreciate your input.

Sucks? No. Top-5? Definitely not. And I've already given my thoughts on computer rankings for football elsewhere so I'm not going to do it here.

[SuperHornet]

UNI doesn't suck. Massey sucks, and always has. UNI is PLAYOFF material, but NOT Top-5 material. Unless you're talking about the bogus thing starting today, those are two different things....

[JohnStOnge]

And I've already given my thoughts on computer rankings for football elsewhere so I'm not going to do it here.

Yes and I've successfully rebutted you countless times and can't get you to see that you're wrong! And you're not the only one.

NEXT year I've got to remember to plan ahead and test you against the Sagarin system in predicting FCS games. Only thing is I'd like to wait unti enough games to be played for Sagarin's system to get to the point where everything is "connected." He used to announce that on his site when it happened but I didn't see it this year.

I think it's reasonable to say that one good way to test how "accurate" a system is at ranking teams is to look at how well it does in predicting outcomes. So we can let you predict either all FCS vs. FCS games each week after we start or a random sample of all FCS games each week.

Then after the Sagarin system outperforms you you'll change your mind, right?

Probably not. I've done that before and even after the Sagarin system beats them they won't change their mind. Still fun though.

[JohnStOnge]

BTW, just so you guys know, I consider Sagarin to be a better system for ranking teams than Massey's. The reason is that Sagarin's system is designed to as good a job as possible of predicting outcomes while Massey's system is designed to do as good a job as possible of explaining past outcomes. That is Massey's philosophy. I actually had an e mail exchange with him about it. He believes that the best and fairest way to rank teams is to look at what has happened in the past and mathematically "explain" it. What actually HAS happened.

I believe that the best way to rank teams from best to worst is to look at who you would EXPECT to win in each matchup.

However the distinction is moot in this case because both systems have UNI rated in the top 5.

And sure there's plenty of error. But there is among human beings too. Like in 2003 you had a pretty strong media consensus that Oklahoma may have had the best college football team ever until they got to the Big 12 Championship game and Kansas State clubbed the Sooners 35-7. Then all of a sudden it was like they NEVER said all that stuff about how great Oklahoma was. They immediately went to what an outrage it was that they were in the BCS championship game.

[Gil Dobie]

My top 5 would include ISUR, UNH, NDSU, and a choice between SDSU, UNI, CCU and EWU. Those are the best 7 teams I've seen this year.

[JohnStOnge]

To address one more assumption I heard expressed in the past:

There is not sufficient evidence to conclude that Sagarin is less accurate when predicting I-AA/FCS games than it is when predicting I-A/FBS games. That's based on the period 2004 through 2006 when Todd Beck was keeping up with accuracy for I-AA games. The site is at http://tbeck.freeshell.org" onclick="window.open(this.href);return false;. He actually started at the end of 2003 but since that was not a full year I didn't include it. Accuracy changes slightly as the year goes on so to do a fair comparison you need full year vs. full year.

Anyway, for the three full years Sagarin's system predicted winners correctly for 73.7 percent (1498 of 2032) of I-A games and for 71.3 percent (1242 of of 1743) of I-AA games. A SLIGHTLY higher percent correct rate for the I-A games. But the difference is not "significant" at the confidence level typically used by convention (95%). You'd conclude that the rate coming out higher for I-A games could very easily have been due to chance.

In any case, as a practical matter, Sagarin was about as accurate for I-AA games as it was for I-A games while Todd Beck was following performance with respect to I-AA games for three years and those are large sample sizes. So if you were thinking there is a difference between Sagarin's predictive accuracy for I-A games and it's predictive accuracy for I-AA games that is of any practical significance you probably ought to stop thinking that. That outlook was pretty much debunked by the data Todd Beck reported for those three years.

[AZGrizFan]

Big deal. I can predict the winners 75% of the time too. It's not really all that difficult.

[D1B]

Cool, Z's here to lose another debate!

[clenz]

Big deal. I can predict the winners 75% of the time too. It's not really all that difficult.

I've run a top 25 puck game for 5 years now. Ive had literally tens of thousands ballots submitted. The average of all id them is 69 percent.

75% by a human would be/is pretty high.

It happens. Im at roughly 82 over the last 2 years but 75 is tough.

[JohnStOnge]

Big deal. I can predict the winners 75% of the time too. It's not really all that difficult.

Ok. You can try to do that next year too. I'm not talking about cherry picking games. You have to pick every game or a "representative sample" of every game.

The favorite by the updated line during 2004 - 2006 won 73.0 percent of the time. To me the updated line represents kind of a consensus of who people who are putting their money on the line following football think are going to win. I've heard people say that if you really want a ranking of teams from best to worst you should look at who would be favored to win each possible game according to the line. And Sagarin's system is about as accurate in terms of picking favorites as the line is.

[AZGrizFan]

Big deal. I can predict the winners 75% of the time too. It's not really all that difficult.

I've run a top 25 puck game for 5 years now. Ive had literally tens of thousands ballots submitted. The average of all id them is 69 percent.

75% by a human would be/is pretty high.

It happens. Im at roughly 82 over the last 2 years but 75 is tough.

This is picking games straight up, right? I've played the BSC challenge every year for the past 6 years and rarely get below 80% in a given week. Most conferences aren't that difficult....out of 5-6 games a week there's usually only one that's a toss up. Over a season I'm positive I could get 75%...and even at 70%, you're telling me the computer, in all it's genius, only adds 5%?

[AZGrizFan]

Cool, Z's here to lose another debate!

I'm not "debating" anybody, dipshit. Merely making a statement.

[D1B]

Cool, Z's here to lose another debate!

I'm not "debating" anybody, dipshit. Merely making a statement.

Sure about that?

[clenz]

I've run a top 25 puck game for 5 years now. Ive had literally tens of thousands ballots submitted. The average of all id them is 69 percent.

75% by a human would be/is pretty high.

It happens. Im at roughly 82 over the last 2 years but 75 is tough.

This is picking games straight up, right? I've played the BSC challenge every year for the past 6 years and rarely get below 80% in a given week. Most conferences aren't that difficult....out of 5-6 games a week there's usually only one that's a toss up. Over a season I'm positive I could get 75%...and even at 70%, you're telling me the computer, in all it's genius, only adds 5%?

Top 25...not just your home conference.

If you can't get at least 80% of your own conference you're a moran.

Also ran a conference pick em this year in the mvfc and average was 85%.

Much easier to pick what you're familiar with.

A computer can't take into account injuries, field conditions, team momentum, match up issues, etc...

The fact is that for a computer to have the same, or better %, as humans tells me quite a bit about human picking. Humans should have an advantage that can't evev come close to being matched by a computer.


5% over the course of thousands of games makes a HUGE difference

[AZGrizFan]

I'm not "debating" anybody, dipshit. Merely making a statement.

Sure about that?

Yes, I'm sure.

[AZGrizFan]

This is picking games straight up, right? I've played the BSC challenge every year for the past 6 years and rarely get below 80% in a given week. Most conferences aren't that difficult....out of 5-6 games a week there's usually only one that's a toss up. Over a season I'm positive I could get 75%...and even at 70%, you're telling me the computer, in all it's genius, only adds 5%?

Top 25...not just your home conference.

If you can't get at least 80% of your own conference you're a moran.

Also ran a conference pick em this year in the mvfc and average was 85%.

Much easier to pick what you're familiar with.

A computer can't take into account injuries, field conditions, team momentum, match up issues, etc...

The fact is that for a computer to have the same, or better %, as humans tells me quite a bit about human picking. Humans should have an advantage that can't evev come close to being matched by a computer.


5% over the course of thousands of games makes a HUGE difference

The computer also doesn't have inherent biases. Allegedly.

[D1B]

This is picking games straight up, right? I've played the BSC challenge every year for the past 6 years and rarely get below 80% in a given week. Most conferences aren't that difficult....out of 5-6 games a week there's usually only one that's a toss up. Over a season I'm positive I could get 75%...and even at 70%, you're telling me the computer, in all it's genius, only adds 5%?

Top 25...not just your home conference.

If you can't get at least 80% of your own conference you're a moran.

Also ran a conference pick em this year in the mvfc and average was 85%.

Much easier to pick what you're familiar with.

A computer can't take into account injuries, field conditions, team momentum, match up issues, etc...

The fact is that for a computer to have the same, or better %, as humans tells me quite a bit about human picking. Humans should have an advantage that can't evev come close to being matched by a computer.


5% over the course of thousands of games makes a HUGE difference

Nice work, Clenz. If UM spent any money on academics, Z might just get this.

[clenz]

Top 25...not just your home conference.

If you can't get at least 80% of your own conference you're a moran.

Also ran a conference pick em this year in the mvfc and average was 85%.

Much easier to pick what you're familiar with.

A computer can't take into account injuries, field conditions, team momentum, match up issues, etc...

The fact is that for a computer to have the same, or better %, as humans tells me quite a bit about human picking. Humans should have an advantage that can't evev come close to being matched by a computer.


5% over the course of thousands of games makes a HUGE difference

The computer also doesn't have inherent biases. Allegedly.

You think you could pick 75% of MAC, Sun Belt, AAC, PFL, SoCon, Big South, etc... games correctly?

What about getting into D2 games? Could you pick 75% of them correctly?

[Mvemjsunpx]

I'm not sure how this thread got on this subject, but I've picked every game involving a DI team (including games against non-DIs) straight-up since 2004. Here are the figures for those that want some sort of idea how these numbers look:

Disclaimer - Somewhere in there (2008 on, I believe), I started counting postseason games (incl. conference championship games) as worth double.


2004 - 1000-396 (71.6%)
2005 - 990-409 (70.8%)
2006 - 1072-412 (72.2%)
2007 - 1068-384 (73.6%) I apparently forgot to tally up the bowl games here, so it's not quite complete
2008 - 1178-412 (74.1%)
2009 - 1133-416 (73.1%)
2010 - 1142-420 (73.1%)
2011 - 1164-408 (74.0%)
2012 - 1158-403 (74.2%)
2013 - 1247-412 (75.2%)
2014 - 1258-396 (76.1%) So far…