(01-16-2019 12:00 PM)usffan Wrote:  

(01-16-2019 11:38 AM)quo vadis Wrote:  

(01-16-2019 10:23 AM)usffan Wrote:  

(01-16-2019 10:06 AM)quo vadis Wrote:  

(01-16-2019 09:48 AM)usffan Wrote:  The problem with this is the whole "ranked" part. If it's based exclusively on computers, then people will whine about quirky teams that most would probably disagree with (for example, Mississippi State is the #8 Sagarin team - does anybody other than a Mississippi State fan and MAYBE a ridiculous SEC homer seriously believe Mississippi State should have been in the top 8?). But using human ratings opens up the potential for human biases. Remember a few years ago when Mac Brown rallied the Big XII writers and coaches to vote Cal down and Texas up in the final poll so that Texas would go to the Rose Bowl and Cal got relegated to the Holiday Bowl? You put a "top 8" situation in there and suddenly all of the west coast voters start rallying around moving Washington up and UCF down.

Ultimately, the games on the field have to matter. If a team loses a conference championship game to another team that's going to be in the playoff, why do we automatically assume that shouldn't count as a de facto playoff game? The whole notion that Georgia should have gone to the playoff this year was so silly, since that would have meant that the Alabama/Georgia SEC title game meant absolutely nothing when they played the next week.

A couple things come to mind:

1) Your example of Sagarin having Mississippi State #8 is a good reason why the previous systems that used computers, like the BCS, used multiple computers and also 'threw out' the highest and lowest rankings to control for those kinds of crazy outliers**. Also, because human beings do have biases, it makes sense to have multiple humans or a combination of computers and humans making these decisions, so that no one bias is likely to carry the day.

2) Concerning your Georgia/Alabama example, it doesn't resonate much with me. E.g., if #1 Duke and #6 North Carolina play in the finals of the ACC basketball tournament, does that game become a joke just because both teams are going to make the NCAA tournament no matter who wins? I don't think so and the ACC doesn't either.

In the end, though, whether humans or computers are involved, games on the field are ultimately what matter. How does a Clemson or a Notre Dame get ranked #2 or #3 in the CFP? By winning games against tough teams. There is no system that ignores games on the field.



** btw, while MSST at #8 does seem crazy, they did finish #19 in the Massey Composite while unranked in the final AP poll, so the human voters maybe underappreciated them.

I don't think it's valid to compare basketball to football for a couple of reasons. 68 teams are getting into the NCAA tournament compared to 4 (or 8 if expansion ever happens). If there were only 4 or 8 bids to the NCAA tournament, then you can bet your bottom dollar that the loser of that Duke/UNC game would be more likely to be left out for the champion of the Pac-12, for example.

I would agree with the "top 8 ranked teams" scenario if (and only if) there were also a completely independent commission (say, for example, the CFP committee) that were given free reign to 'schedule' the OOC games. Because that is the only way that non-blue bloods would ever have a realistic chance of gaining access to this. If the schedules were set up so that certain weeks were planned for these (let's say 2nd weekend in October and 2nd weekend in November), the committee could ensure that every team would have a chance at "winning games against tough teams." I realize that UCF's overblown sense of self-importance in turning down the chance to play Fl*rida shoots this argument in the foot, but if a CFP committee came in in October and forced UCF to play, say, LSU, it would have ended much of the whole debate. That, to me, is how you can justify going with a straight top 8 system, by inserting a chance to make sure whether a team is a paper tiger.

USFFan

Sure, if there were only 4 or 8 teams in the NCAA tournament, then the odds that any team would be left out if they lost a game late are much higher than if they won the game, merely because there are far fewer slots available.

But IMO, that doesn't mean such a system would favor the PAC champion. I think the PAC champion would be compared to the Duke/UNC loser and might get in or might not, depending on how good they are believed to be relative to each other.

The CFP, even with four teams, could favor conference champs moreso than it does. E.g., there could be a rule that says that in the final rankings, conference champs are to be ranked ahead of all non-champs, such that the only way a conference champ can be left out is in favor of a team that also won its conference. But there is no such rule, indicating that the conferences themselves don't think being a champ should be THAT important in determining playoff slots.

As for your qualification for a straight 8, i get what you are trying to do here. I just don't think it is possible to have a committee of some kind setting OOC schedules. OOC schedules are too valuable to the schools themselves, they provide the schools with the flexibility to craft schedules that meet their particular program-building needs, which may or may not include trying to look as good as possible to a ranking committee or anyone else. If school X wants to schedule light OOC because they believe that in their current situation they need easy wins more than tough losses, they shouldn't be forced to play a tougher OOC schedule just so a team like Memphis *can* play an LSU or Ohio State OOC to meet their playoff-seeking needs. So they just aren't going to give up that power.

Given that reality, I guess we have to disagree about the utility about straight 8. IMO, if the playoffs are just 8 teams, then one slot guaranteed to *anyone*, P5 or G5, is too high a price to pay to remove whatever doubt some might have about whether someone is a paper tiger or not.

In effect, a G5 autobid (or A5 autobid) rewards scheduling light, as a team that plays a light schedule and wins can say "hey, our light schedule didn't prove we aren't a Real Tiger, so you have to let us in the playoffs to prove we're not..."

The problem is that it's become the opposite - a built in crutch to ensure Alabama is able to get into the playoff. They play a tough conference schedule that they can't even win, but the purported schedule difficulty (I use purported because most SEC teams use the "SEC is so tough" argument for their scheduling difficulty rather than going on the road and playing tough opponents - there are occasional exceptions) gets them in like last year, rendering this a self-fulfilling prophecy. Hence my argument for an independent group making OOC schedules.

The thing is, using this year as an example, Georgia had a chance to demonstrate they were capable of beating Alabama and couldn't do it head to head. Ohio State didn't. So Ohio State was much more deserving of a playoff spot than Georgia. Thus, games on the field actually mattered. This is why conference champions need to be prioritized over an arbitrary assignment of 'best X number of teams."

USFFan

Well, neither Georgia nor Ohio State made the playoffs, so I'm not sure why they are being compared.

As for conference champions being prioritized, the CFP data suggests they are: Of the 20 teams that have made the playoffs so far, 17 of them have been conference champs, and one of the three that wasn't was Notre Dame, who you have to throw out because while they didn't win a conference they also didn't lose one either. Only Ohio State 2016 and Alabama 2017 have made the playoffs without being conference champs.

And as for Alabama specifically, they have made the playoff 5 times, and 4 of those times have been as conference champs, so it's not really fair to say that the playoffs have somehow become a crutch to get them in as not a champ.

As for SOS, I just looked at the final Sagarin rankings. Alabama's final SOS was #9, which was the highest of anyone in the final Sagarin top 10. Georgia SOS was #10, which was the second-best SOS among the top 10 teams. So the SEC doesn't seem to be skating by somehow with weaker SOS.