Byes: why the huge discrepancy?
So is the SEC scheduling to blame for the astronomical discrepancy in opponents coming off of bye weeks that I discussed in this post? Or is it the individual schools themselves, who can ultimately choose when to schedule non-conference games and when to leave a slot open for byes? Well, statistically, the answer is both.
Total opponents coming off non-SEC-game weeks prior to matchup, 2007-2010
Alabama - 24
Auburn - 21
Florida - 16
Tennessee - 14
Georgia - 14
Mississippi - 13
LSU - 13
Vanderbilt - 12
South Carolina - 11
Mississippi State - 9
Kentucky - 8
Arkansas - 6
Yes, this means that for the 32 SEC games we will have played from 2007-2010, we will have faced only 8 teams that had an SEC game scheduled the week before, meaning that 75% of the time our SEC opponents will not have had a conference game the week before. Meaning that in addition to playing generally weaker opponents the week before, on 75% of the occasions, our SEC opponents have had the option, if they so choose, to schedule their bye week.
Looking at the stats above, you can see that the distribution is much more normal than that of the byes themselves, with most teams clustered in the middle of the range and 2-3 outliers on each end. However, a range of 6 to 24 is huge considering the SEC schedule should be completely impartial. So huge in fact that, without breaking out the really fancy statistics this time, I'll just say that it is virtually statistically impossible to see a distribution range this big if it was truly random. And since the SEC office makes the SEC schedule, it should be pretty random ideally.
Now, for about the 4th time, I will reiterate that I am not claiming conspiracy. That said, something about the way the SEC schedule is set up year in and year out (probably just by accident or failure to think ahead) clearly disadvantages Alabama (and Auburn for that matter) by giving their opponents so many more chances to schedule bye weeks ahead of their matchup, not to mention giving them generally easier opponents even when they don't have byes.
Now, the SEC schedule favoring certain schools only tells half the story. As has been discussed earlier, individual ADs are generally free to schedule non-conference opponents in the weeks for which there are no scheduled conference games. I know that often major non-conference matchups are set years in advance (such as Alabama already having lined up Michigan State in 2016, etc.), but by and large, schools do get ultimate say about what date they leave open for a bye.
And yes, I do realize that it is perfectly natural for schools to choose to leave open weeks prior to matchups against teams they would like more time to rest and prepare for (read: powerhouse teams). So no, this part of the puzzle is not expected to be random. Take us for example. Of course it's no coincidence that our bye week this year and next is before facing our toughest division opponent, LSU. All of that said though, check out the rates at which schools, when given the opportunity (meaning no SEC game the week before) scheduled a bye before each opponent:
Percentage of opponents scheduling byes during non-SEC-game weeks prior to matchup, 2007-2010
Alabama - 70.8% - 17out of 24
LSU - 38.5% - 5 out of 13
Tennessee - 28.6% - 4 out of 14
Florida - 25.0% - 4 out of 16
Vanderbilt - 25.0% - 3 out of 12
Kentucky - 25.0% - 2out of 8
Georgia - 21.1% - 3 out of 14
Auburn - 14.3% - 3 out of 21
Mississippi State - 11.1% - 1out of 9
Mississippi - 7.7% - 1 out of 13
Arkansas - 0.0% - 0 out of 6
South Carolina - 0.0% - 0 out of 11
That's right. When given the chance by the conference schedule, just over 70% of our SEC opponents have arranged bye weeks before playing us. No one else is even above 40%. And let's not forget that our opponents are given tons of chances to schedule byes before meeting us thanks to the conference schedule. It's especially curious to see our rate so much higher than, say LSU or Florida, given that the schedules for these years were mostly finalized before our re-emergence as a powerhouse last season.
So there you have it. Both the format of the SEC conference schedule and the fact that our opponents choose byes before playing us at a much, much higher rate than even other powerhouses like Florida and LSU combines to give us the unbelievably large discrepancy in bye-week opponents.
I'd love to hear what everyone thinks about this, and I wonder what, if anything, can or should be done about it. I suppose we can't do much about what individual ADs choose to do, but surely the SEC schedule can be adapted to ensure a more equal distribution of bye week opportunities.
Finally, consider this stat: since 2006, BCS conference teams coming off bye weeks win 54.8% of the time, compared to 45.2% for their opponents. That may not seem like a lot when it comes to one particular game, but play those odds out in 17 different games and they're likely to become a factor in at least one or two.
FanPosts are just that; posts created by the fans. They are in no way indicative of the opinions of SBN and the authors of Roll Bama Roll.
2 recs |
21 comments
| Add comment
Comments
I'm glad you did this write up
I was pointing this out to my wife last year just from some rudimentary research and thought I saw it as you did.
Nice piece…BAMA’s back!
ROLL TIDE ROLL!!!!!
by alanbama12 on Nov 10, 2009 6:04 AM CST reply actions 0 recs
We have entirely
too many of these “open for scheduling” dates on our opponents’ schedules prior to their games with us. Next year there are 6 teams we play that are fresh off of a bye week. The SEC office has commented on it, calling it “odd.” Hey, it’s not that difficult. Stop leaving open dates in front of us for every single team we play. 17 SEC opponents off of byes in the past few years vs. only 5 for the second highest number in LSU shows a lot of disparity in managing the league.
Just even it up a little SEC. Make UT play SoCar the week before us every other year. Make AU play UGa the week before us every other year. Ta-da!
Roll Tide! Beat L-S-SHOOOOOOOO!
by BamaDixi on Nov 10, 2009 8:12 AM CST reply actions 0 recs
Again, the opponent bye breakdown
Total SEC opponents coming off bye weeks, 2007-2010
Alabama – 17
LSU – 5
Tennessee – 4
Florida – 4
Vanderbilt – 3
Georgia – 3
Auburn – 3
Kentucky – 2
Mississippi State – 1
Mississippi – 1
South Carolina – 0
Arkansas – 0
by Bama07 on Nov 10, 2009 8:15 AM CST reply actions 0 recs
Why the huge discrepancy?
I think the answer is two fold. 1. We are feared. People need two weeks to prepare for us because they know they will need any advantage that they can get. So then why dont teams do this to Florida as much you might ask? 2. We are feared and teams feel that they have a chance to beat us and hang their season on that one token win (hypothetical of couse as the last two seasons have shown) but I still feel teams dont want to waste their off week before Florida because they walk in there with no hope of winning, although this idea is changing rapidly, teams feel that we are a juggernaut but that they have a chance to beat us. But they sure have been wrong lately!
Great Game Hokies! What a battle!
by The Voice of Reason on Nov 10, 2009 8:43 AM CST reply actions 0 recs
Alabama even
when we were down from about 97-07 was still Alabama.
Who does LSU want to beat the most? Alabama.
Who does UT want to beat the most? Alabama.
Who do Ole Miss and MSU want to beat the most? Alabama.
And, we know who Auburn wants to beat the most.
The point is unless we go on about a 30 year nose dive we will always be just about everybody in the West and UT’s number one foe. I actually think Arky wants to beat LSU more than Bama. But I would say Ole Miss & MSU want to beat us at least as much as they want to beat each other. That comes from the fact that although they don’t like each other they also feel the other is actually not that good. Thus I think they would rather beat Bama than each other.
So think about it. If everybody wants to beat you more than any other team on their schedule you will face a lot of byes. Look at the teams at the bottom. Who is sitting around saying- Man I want to beat SC, Arky, Ole Miss or MSU. Now why Vandy gets a few byes is just dumb luck.
I hate the NCAA more than UT & AU combined. At least with UT & AU you got a fighting chance.
by 5026 on Nov 10, 2009 9:18 AM CST reply actions 0 recs
Your points are true and well taken. This explains the 70.8% statistic in my post above.
However, it does not explain why the SEC schedule (especially considering everything you said above) allows them so many opportunities to do so.
by Bama07 on Nov 10, 2009 12:19 PM CST up reply actions 0 recs
I’ll just say that it is virtually statistically impossible to see a distribution range this big if it was truly random
Not to nit-pick, but this really isn’t true. You don’t want the schedule to be random, you want it to be equal and those two are not the same.
So it would be equal and fair to give everyone 4 SEC games and off week, and then 4 more SEC games, alternating which team you got to play after your bye every year. This would be fair. it wouldn’t be random at all, though.
One problem is very likely to be TV. The above is a good idea from a football perspective, but a terrible idea from a TV perspective. You don’t want all of your teams off the same week because that’s basically throwing away free money.
Another is that you’re talking about scheduling for 12 teams and many of the teams have constraints, ’Bama/UT needs to be in late October, lots of games are scheduled for thanksgiving weekend, schools want weekends off where they can schedule non-conference homecoming games, etc.
This is interesting, though.
I'm wrong all the time.
by PeteHoliday on Nov 10, 2009 11:35 AM CST reply actions 0 recs
Not to nit-pick, but this really isn’t true. You don’t want the schedule to be random, you want it to be equal and those two are not the same.
So it would be equal and fair to give everyone 4 SEC games and off week, and then 4 more SEC games, alternating which team you got to play after your bye every year. This would be fair. it wouldn’t be random at all, though.
Well clearly there is no such system in place to make it all equal, so the only alternative that would make it fair would be for it to be random. No one said the SEC had a system to make it like that. And as you stated so well, there are a lot of considerations (namely television) why this is so.
Since I’m not claiming conspiracy and since there is no system in place to make it equal, we are left to assume (hope?) that the distribution is random. However the statistics show that is not the case. That is all I was trying to say.
by Bama07 on Nov 10, 2009 12:17 PM CST up reply actions 0 recs
since there is no system in place to make it equal
That’s a bold claim. What do you base that on?
However the statistics show that is not the case.
I don’t think you really have enough seasons included to make that claim. See below for 1992-2008 numbers.
I'm wrong all the time.
by PeteHoliday on Nov 10, 2009 1:39 PM CST up reply actions 0 recs
That’s a bold claim. What do you base that on?
There’s obviously some miscommunication here. My claim is one that can’t be disputed: the SEC doesn’t have a system in place to guarantee equal opponents coming off bye opportunities. If they did, we wouldn’t be having this discussion.
I don’t think you really have enough seasons included to make that claim. See below for 1992-2008 numbers.
The sample size is statistically big enough to prove beyond a reasonable doubt that the distribution is not random. All this means is that something besides pure randomness is causing so many of our opponents to have so many bye opportunities.
As for the numbers you posted below, good work. I responded to them separately.
by Bama07 on Nov 10, 2009 4:16 PM CST up reply actions 0 recs
the SEC doesn’t have a system in place to guarantee equal opponents coming off bye opportunities.
Do you even know if this is possible given the considerations that must be taken into account? (time-specific rivalry games, TV contracts, etc.) My point is that partial scheduling of 12 teams is not exactly easy, and I seriously doubt we have even mentioned all of the different constraints the SEC has to work around.
The sample size is statistically big enough to prove beyond a reasonable doubt that the distribution is not random. All this means is that something besides pure randomness is causing so many of our opponents to have so many bye opportunities.
Well, proof beyond a reasonable doubt is not really a statistical term. But the statistical argument aside (I’m actually curious to see the math on this — N=4 seems like an awfully small sample size to be extrapolating from) we already know that it’s not random because of the sheer number of constraints and things that are fixed, so the question isn’t whether or not it’s random (it’s not) the question is either whether or not it’s equal (obviously not in the short-term, but, then, schedules are never equal in the short term) or whether or not it’s fair (something that is probably up for debate and would require a little more work than we’ve done so far).
I'm wrong all the time.
by PeteHoliday on Nov 10, 2009 4:31 PM CST up reply actions 0 recs
In statistical terms, n=161, the number of open slots prior to an SEC game over 4 years.
We know the schedule format is not random, no one is suggesting that it is. However, it should ideally be set up so that these 161 open slots are distributed among the 12 teams randomly (which of couse would mean in the long run roughly equally).
by Bama07 on Nov 10, 2009 4:55 PM CST up reply actions 0 recs
I don’t think your N can be 161 here. 161 is a measurement, not a number of measurements taken. Your N could vary a bit depending on how you set up your statistics, but each potential bye isn’t a measurement itself . . . to say nothing of the fact that if you could find a way to do that, the measurements wouldn’t be independent.
It would make more sense to look at seasons as your N and then run the stats that way, since each season can be considered independent of the others in terms of location of potential bye week scheduling.
One thing I don’t know: does the SEC schedule around dates given by teams? In other words, can Alabama block off a date in 2015 and say “This is the weekend we’re playing OutOfConference U” and have the SEC respect that?
I'm wrong all the time.
by PeteHoliday on Nov 11, 2009 12:49 PM CST up reply actions 0 recs
Put another way: what you’re really trying to find out is whether or not the SEC’s scheduling algorithm is fair, and it seems like it would be easier to just email the SEC and ask how they compile the schedules than it would be to try to mathematically deduce it from four years of statistics.
I'm wrong all the time.
by PeteHoliday on Nov 10, 2009 4:32 PM CST up reply actions 0 recs
I went through the data from 1992-2008 and it doesn’t really bear out the trend that you’re describing.
.
Potential bye weeks 1992-2008:
- Vandy — 59
- Auburn — 56
- ’Bama — 55
- Miss — 51
- MSU — 50
- Ark — 46
- Tenn — 45
- LSU — 41
- UGA — 41
- SoCar — 38
- UK — 36
- Fla — 35
.
Opponents’ Off of Potential Bye Weeks 1992-2008
- UGA — 60
- Ark — 59
- LSU — 55
- Auburn — 50
- Fla — 50
- ’Bama — 48
- SoCar — 44
- Miss — 41
- MSU — 41
- Tenn — 41
- UK — 36
- Vandy — 28
.
If you look at the net difference, Alabama has had the opportunity to play 7 more SEC games coming off of a bye than their opponents had chances to play them after a bye week. The spread of those numbers is interesting, though, ranging from -19 (UGA) to +31 (Vandy).
You can see the charts I’m basing these off of here.
I'm wrong all the time.
by PeteHoliday on Nov 10, 2009 1:37 PM CST reply actions 0 recs
A couple of important things to note. First, the SEC changed it’s schedule format in 2001 or 2002 (sorry I can’t remember exactly which) to allow for two rotating out-of-division opponents instead of just one. Second, the NCAA permanently changed to a 12-game season beginning in 2006, which in turn changed the format of the SEC schedule. Therefore the most telling seasons to look at in terms of current trends influencing future seasons would be 2006 onward. Unfortunately I did not include 2006 in my stats, but that would be the only other previous year completely valid to compare against.
All of that said, this is great work you did looking at the longer term picture. What’s telling about these statistics isn’t what’s going on now, but rather what went on in the years prior to my little informal study. Obviously some other schools, namely Georgia and Arkansas, were disproportionately disadvantaged during those years. So again, this just goes to show the system is broken, not that the SEC is out to disadvantage any one school.
That doesn’t change the fact that the current post-2006 trend is putting us at a big disadvantage now and in the immediate future.
What we should do is add 2006 to my study (since that was the first year permanently moving to the current format) and also I just discovered we can add 2011 (since the SEC schedule has been made already for that season). Those years might give us a better idea about the current trend, but anything before that just shows that different schools were being disadvantaged under the previous formats.
by Bama07 on Nov 10, 2009 4:11 PM CST up reply actions 0 recs
I’m not ready to say the system is “broken” quite yet.
The schools have no reason to allow a broken system to continue, so the fact that it persists suggests that there’s either a) a good reason the scheduling fluctuates or b) something about the way the SEC schedules games that makes the athletic departments okay with it.
I'm wrong all the time.
by PeteHoliday on Nov 10, 2009 4:35 PM CST up reply actions 0 recs
The schools have no reason to allow a broken system to continue, so the fact that it persists suggests that there’s either a) a good reason the scheduling fluctuates or b) something about the way the SEC schedules games that makes the athletic departments okay with it.
You’re probably right here, and I would agree to a certain extent. Schools will put up with a little discrepancy, but not the kind that has been shown in recent years if it continues.
I would simply argue that when they make the schedule (mostly around tv and other $ considerations, which of course makes all the ADs happy) they don’t really even look at who is being left open with byes. This would be fine (and random I might add) except the SEC since 07 has stuck with virtually the same schedule. More on this later.
by Bama07 on Nov 10, 2009 5:21 PM CST up reply actions 0 recs
Also, as for the 2002 change, that didn’t change the number of conference games, just who they were against, so it’s really irrelevant. We’ve played 3 out-of-division regular-season games ever since 1992. 2003 was the first season we didn’t play Vandy since the expansion.
Still, these numbers are all relevant, because we’re not talking about something random, we’re talking about the algorithm the SEC uses, so unless there’s some fundamental reason that we know of for it to have to change in 2006, I don’t see any reason to ignore the numbers.
I'm wrong all the time.
by PeteHoliday on Nov 10, 2009 4:50 PM CST up reply actions 0 recs
If I remember correctly, the 2006 conference schedule was already made prior to the NCAA’s decision to go to a 12-game season, meaning the first conference schedule designed for a 12-game season would have been 2007, which is convenient for my numbers. I could be wrong on this though so feel free to correct me. And our SEC schedule did change beginning that year.
The current format (beginning in 2007) has remained virtually the same. I don’t have time to do this for all 12 teams, but from 2007 to 2011, Alabama has played its 6 permanent opponents in the exact same slot each season, with Arkansas getting bumped up one week in 07 and OM getting bumped up one week in 09. With the exception of 07 (perhaps because of the FSU game?), the two rotating East teams have filled the same two slots as well. This means that after the NCAA went to a 12-game season, the SEC has used, and appears to continue to use, virtually the same schedule format which as has already been established disadvantages Alabama as well as Auburn.
so for 2007 – 2011
Arkansas – week 4 (4 of 5 years)
rotating East teams – filling weeks 5 & 6 (4 of 5 years)
Mississippi – week 7 (4 of 5 years)
Tennessee – week 8 (all 5 years)
LSU – week 10 (all 5 years)
Miss State – week 11 (all 5 years)
Auburn – week 13 (all 5 years)
So again, at least from our perspective, and I suspect this will mostly hold across most teams, the schedule is essentially remaining the same each year post-2006 change. The only thing that causes even the rare deviation from the set schedule appears to be big-time non-conference games which are scheduled years in advance.
by Bama07 on Nov 10, 2009 5:38 PM CST up reply actions 0 recs
If I were...
…a coach for one of these teams, I think I’d rather have a bye week after playing Bama….
"High standards come from passion within...." --Coach Nick Saban
by NiceLittleSaturday on Nov 10, 2009 2:14 PM CST reply actions 2 recs

by 




















