B1G vs. SEC

[Moiraine]

Just read the ESPN recap of the OSU game, and they referred to the Buckeyes as the SEC North. Shamelessly stoking their business partner even when they lose.

I saw that in an ESPN article and searched it. They make it seem like Urban Meyer himself said he wanted to turn OSU into the SEC north but I googled it and couldn't find anything like it. It's just some crap they made up.

[Big Ten Fan]

 

 

Ok, so now that the Missouri game is over, I did a simple rudimentary exercise to see how well the SEC has performed compared to how the betting market expected them to perform prior to the bowls being played. Pretty cool the results that I found:

 

 

The SEC as a whole has underperformed by just 2 points per game. They're basically performing just like Vegas has expected them to perform. Almost exactly, really. (Note: Alabama and Tennessee have yet to play obviously). In fact, every team except the set of quadruplets from the SECW have actually outperformed their spread. Basically what this tells us is that any gloom and doom about the SEC is massively overstated. But, perception is a weird and fickle beast.

 

I'd like to do this for Sagarin's predictions as well, but I'll have to find an archive from early December, as his ratings update daily and will thus have some bowl outcomes already reflected in them, and I'd like to see the change that bowls may have had - can't use data that already has this factored in. I'll get around to that eventually.

 

This is not how you should be calculating the average point difference per game. You're letting negative and positives differentials offset each other. You should be using the absolute value of the differentials. For those ten games you get a sum of 144. So the average point difference was 14.4 points. This is only the average point differential and doesn't indicate who it was in favor of. But it does indicate the Vegas's spread was, on average, off by about 2 touchdowns. Which, in my opinion, seems pretty significant.

 

 

If that was my goal, that's what I would have done. But it wasn't. It was to get an idea of how over or under-rated (or over- or under-performing in the bowl games, as a whole) the SEC was. Yes, Vegas spreads are always off because of variance, that's how it works. My goal was to see if they were consistently off by a meaningful margin in one particular direction.

 

 

I just noticed you had 9 games and not 10, so the average was really 16. This indicates the magnitude of how far the vegas spread was off on average.

 

I don't believe you have achieved what you wanted with your calculations. A weighted average may be appropriate here to minimize the effect of outlier games. You have 5 positive differentials with a sum of 248 and 4 negative ones with a sum of -82. So 5/9 * 248 + 4/9 * -82 = (1240 -328) / 9 = 101.33 for the weighted point differential sum. The average would then be about 11.259.

 

 

Edited for clarity (misstated intent of weighted average initially)

Edited January 3, 2015 by Big Ten Fan

[tschu]

You might want to check that math.

[Coqui]

 

Just read the ESPN recap of the OSU game, and they referred to the Buckeyes as the SEC North. Shamelessly stoking their business partner even when they lose.

I saw that in an ESPN article and searched it. They make it seem like Urban Meyer himself said he wanted to turn OSU into the SEC north but I googled it and couldn't find anything like it. It's just some crap they made up.

 

SEC is just B1G South since all the successful SEC coaches started in the B1G

[Big Ten Fan]

 

 

 

Ok, so now that the Missouri game is over, I did a simple rudimentary exercise to see how well the SEC has performed compared to how the betting market expected them to perform prior to the bowls being played. Pretty cool the results that I found:

 

 

The SEC as a whole has underperformed by just 2 points per game. They're basically performing just like Vegas has expected them to perform. Almost exactly, really. (Note: Alabama and Tennessee have yet to play obviously). In fact, every team except the set of quadruplets from the SECW have actually outperformed their spread. Basically what this tells us is that any gloom and doom about the SEC is massively overstated. But, perception is a weird and fickle beast.

 

I'd like to do this for Sagarin's predictions as well, but I'll have to find an archive from early December, as his ratings update daily and will thus have some bowl outcomes already reflected in them, and I'd like to see the change that bowls may have had - can't use data that already has this factored in. I'll get around to that eventually.

 

This is not how you should be calculating the average point difference per game. You're letting negative and positives differentials offset each other. You should be using the absolute value of the differentials. For those ten games you get a sum of 144. So the average point difference was 14.4 points. This is only the average point differential and doesn't indicate who it was in favor of. But it does indicate the Vegas's spread was, on average, off by about 2 touchdowns. Which, in my opinion, seems pretty significant.

 

 

If that was my goal, that's what I would have done. But it wasn't. It was to get an idea of how over or under-rated (or over- or under-performing in the bowl games, as a whole) the SEC was. Yes, Vegas spreads are always off because of variance, that's how it works. My goal was to see if they were consistently off by a meaningful margin in one particular direction.

 

 

I just noticed you had 9 games and not 10, so the average was really 16. This indicates the magnitude of how far the vegas spread was off on average.

 

I don't believe you have achieved what you wanted with your calculations. A weighted average may be appropriate here to minimize the effect of outlier games. You have 5 positive differentials with a sum of 248 and 4 negative ones with a sum of -82. So 5/9 * 248 + 4/9 * -82 = (1240 -328) / 9 = 101.33 for the weighted point differential sum. The average would then be about 11.259.

 

 

Edited for clarity (misstated intent of weighted average initially)

 

 

Ugh. Have too much crap in my text editor scrap page. So 5/9 * 62 + 4/9 * -82 = (310 - 328) /9 = -18/9. So actually, an average of -2. So, yes, your calculation is close anyway in this scenario.

 

So, in terms of just over or under predicting, Vegas seems pretty fair. However, in terms of how accurate they were, they seem pretty off. On the one hand, I guess you could say that SEC represented themselves, fairly well, with respect to the spread. However, the spread had pretty low predictive power (seen by looking at the average magnitude of the differentials) so unless you were interested in just betting, and then only on the whole range of games, I don't think it was too indicative of the SEC's strength.

 

Edit: fixed minor typo

Edit 2: fixed redundancies in sentence

Edited January 3, 2015 by Big Ten Fan

[BigRedBuster]

Alright, let's revisit. Bama-tOSU isn't over yet, but I think that with 5 data points from the top of the SECW underperforming their expectations in bowls, we can knock the SECW down one notch. We have a solid trend among those teams. Note that TAMU and Arkansas both did well though. Overall, that division is still really good, and basically the top came down, but the bottom rose up. There are still 6 top-20 power teams, at least, in that division, without a doubt. Bama, Ole Miss, MSU, Arkansas, Auburn, LSU in no particular order (LSU probably last out of that group).

 

The SEC East on the other hand has done well in its bowls. Tennessee still has to play Iowa which should be a completely terrible game...I'm not holding my breath for this result. There was this narrative this season that the SECE was terrible...sure, UK and Vandy are bad, but that division largely was underrated badly. Georgia can easily make a case for being the best or 2nd best team in the SEC.

 

In terms of ranking the conferences? Won't change much. The Pac will close the gap a little bit, but that's all. Oregon, Stanford, Utah beat their expected spreads, Arizona, ArizonaSt, and USC did not. I haven't done the math but I think that they as a whole will pick up some ratings points, and the SEC will lose a few. It won't be nearly enough though.

 

Narratives are fun for some people, and I know many who don't like the conference are glad that the 5 SECW teams lost, and that's fine. Hate away, I won't stop you.

 

 

 

One conference that can make a little jump in the metrics would be the B1G... now the thing about the B1G is that the top teams have all done well in bowls, but the bottom teams have been pretty terrible. That's good, perception-wise, for the B1G since most casual fans don't care that Illinois got romped, they just see that Sparty and Wisconsin won. But top to bottom conference rankings don't care - each team is equal. Still, it's been a good bowl season for the Big Ten who was favored in 0 bowl games and could be sitting at 5-4. (Note that despite being 0 for 10 in being favored, probability dictates that the Big Ten would still go 3-7 or so, on average). Still, a good showing for the Big Ten.

 

Things look good for the Big Ten. In the next couple of years, I'm thinking that the Big Ten will be right there with the Big 12 for the 3rd best conference due to the Urban Meyer and Jim Harbaugh hires, Penn State getting back to full strength, etc etc.

I can't get past the parts of your post where you clearly still think the SEC is some amazingly dominating conference that is so much better than all the rest.

 

Sorry, there just isn't any facts that prove that.

[tschu]

 

I can't get past the parts of your post where you clearly still think the SEC is some amazingly dominating conference that is so much better than all the rest.

 

Sorry, there just isn't any facts that prove that.

 

 

I've never said that; in fact, I've posted probably 10 times in this thread that they're the best conference but that it's not by a big margin. (The margin over the B1G is pretty sizeable though.) But thanks for playing.

[BigRedBuster]

 

 

I can't get past the parts of your post where you clearly still think the SEC is some amazingly dominating conference that is so much better than all the rest.

 

Sorry, there just isn't any facts that prove that.

 

I've never said that; in fact, I've posted probably 10 times in this thread that they're the best conference but that it's not by a big margin. (The margin over the B1G is pretty sizeable though.) But thanks for playing.

You claim they should have 9 teams in the top 20. That would mean that almost 50% of the top 20 are from one conference. That would also mean that 64% of that conference is in the top 20.

 

By math you basically ares dying what I stated.

 

On the other hand, I don't see anything that says they are better than at her conferences.

[deedsker]

 

 

I can't get past the parts of your post where you clearly still think the SEC is some amazingly dominating conference that is so much better than all the rest.

Sorry, there just isn't any facts that prove that.

 

I've never said that; in fact, I've posted probably 10 times in this thread that they're the best conference but that it's not by a big margin. (The margin over the B1G is pretty sizeable though.) But thanks for playing.

You claim they should have 9 teams in the top 20. That would mean that almost 50% of the top 20 are from one conference. That would also mean that 64% of that conference is in the top 20.

By math you basically ares dying what I stated.

On the other hand, I don't see anything that says they are better than at her conferences.

This. You want to claim Arkansas is a top 20 team, but yet their victors can't beat #19 Wisconsin. Logic fail.

[tschu]

Buster it depends on whether you count sheer volume, or are using a weighted average or a central mean. Obviously a central mean is probably best. The Pac-12 has 6 teams roughly within that range, and if you stretch it a little bit you can include a seventh in Arizona, just like I stretched the SEC a bit to say you can include that 9th team in Texas A&M. But obviously there are only 12 teams in the Pac-12. That's why the margin isn't huge.

 

I'm obviously not gonna bother to reply to deedsker's stupidity

[deedsker]

The best team the SEC say they have beat this year is Minnesota? Clemson? That is how you build the best conference?

[deedsker]

I guess Wisconsin? The won the vaunted B1G West.

[Big Ten Fan]

 

 

I can't get past the parts of your post where you clearly still think the SEC is some amazingly dominating conference that is so much better than all the rest.

 

Sorry, there just isn't any facts that prove that.

 

 

I've never said that; in fact, I've posted probably 10 times in this thread that they're the best conference but that it's not by a big margin. (The margin over the B1G is pretty sizeable though.) But thanks for playing.

 

 

I think it is arguable who is the best conference this year. Over the past several years, I would agree, though, based on national championships, if nothing else(and there are plenty of other arguments that could be made for them being the best conference besides that). Are you saying that, even over just this year, the margin over the big ten is pretty sizable? If so, what leads you to believe this?

[deedsker]

i just don't get how anyone could say the PAC is not the best conference this year. Non conference to bowl games. Even if tOSU beats Oregon. Top to bottom they have out performed everyone.

[Mavric]

i just don't get how anyone could say the PAC is not the best conference this year. Non conference to bowl games. Even if tOSU beats Oregon. Top to bottom they have out performed everyone.

 

This is where I'm at.

 

Regular Season OOC vs. Power 5:

Pac-12: 8-3

ACC: 9-7

SEC: 5-6

Big XII: 4-6

B1G: 6-11

 

The regular season numbers can be skewed by mismatched opponents. That is, a good team from one conference playing a bad team from another conference. Still, the Pac-12 was the best by a wide margin.

 

Bowl Season (so far):

Pac-12: 6-1

SEC: 6-5

Big Ten: 5-5

ACC: 4-7

Big XII: 1-4

 

Bowl games are not perfect match-ups but they are generally fairly even. And again, the Pac-12 wins by a wide margin.

 

Combining the two, they are 14-4 (77%). No other conference is above .500.