OU vs KU 8pm CST ESPN2

[WichitaSooner]

That doesn't really address my point. In baseball the coach might decide to make a change and bring in a pinch hitter. The coach had 5 players to choose from. The player selected ends up striking out and I'm saying that if the other possible choices were relied on there would be that player who got the job done. My point is changing the selection process does not change this from happening one iota.

I'm not trying to be dense, I guess I just honestly don't understand your point. Of course no decision by a coach will work out 100% of the time. That doesn't mean it was the wrong choice. If anyone could guarantee a successful result, that person should be gambling on games, not coaching.

And of course there is no way to know that one of the other four guys would have gotten a hit. Just like there is no way to know whether a team that just missed the tourney field would have won a game (or more) had they made it. They may, they may not. If your point is that the committee isn't perfect, I absolutely agree. There will never be a perfect system because unlike in pro sports, teams play widely divergent schedules, so the committee just has to do its best to come up with factors that they think are most relevant, and then try to apply those factors to the teams they are choosing between.

 

[sheepdogs1]

Really, from a historical standpoint, the committee's seeding and selection is pretty good. Here's the historical data:

http://mcubed.net/ncaab/seeds.shtml

So you can see, 1 seeds win almost 80% of all the games they play in the tournament, 2 seeds 70%, 3 seeds 64%, 4 seeds a around 60%, and so on, and so forth as you go down each seed progressively wins a smaller % of games.

And it's even more remarkable as you get into the seed vs. seed matchups in the first round:

-8/9 is almost exactly 50/50 with 8 seeds winning slightly more often
-7/10 is right around 60/40
-6/11 is around 65/35.

Same at the top:

-Obviously the 1/16 over 99%.
-The 2/15 is 95/5.
-The 3/14 is about 85/15.
-The 4/13 is 80/20.
-Even the infamous 5/12 game, the 5 wins a little over 67% of the time.

Taken together, the symmetry is actually quite impressive and speaks to how well the committee actually seeds a competitive tournament.

You're using old selection methodology to make your point. If it is as good as you say then why make the change? Do you believe the change will improve things even better?

In addition, my focus was the at large bids. How many of the top seeds are at large?

 

[thebigabd]

Nice variable there, sheepdogs... Lets see how he responds.

 

[sheepdogs1]

I'm not trying to be dense, I guess I just honestly don't understand your point. Of course no decision by a coach will work out 100% of the time. That doesn't mean it was the wrong choice. If anyone could guarantee a successful result, that person should be gambling on games, not coaching.

And of course there is no way to know that one of the other four guys would have gotten a hit. Just like there is no way to know whether a team that just missed the tourney field would have won a game (or more) had they made it. They may, they may not. If your point is that the committee isn't perfect, I absolutely agree. There will never be a perfect system because unlike in pro sports, teams play widely divergent schedules, so the committee just has to do its best to come up with factors that they think are most relevant, and then try to apply those factors to the teams they are choosing between.

My point is do you expect the new process to offer up different results.

 

[WichitaSooner]

So if OU went 13-0 in the non-con, and 6-12 in the league... Losing your last 12 games in a row, that does not matter to the committee?

Right or wrong, this is the current system, yes. The timing of wins and losses doesn't matter. Who you beat/lose to matters, but when those games were played doesn't. Last year's OU squad probably proves more than any other that the committee isn't lying when they say that games at the end of the season no longer get any more weight than games at the start. There are a few teams this year who will also make the tourney despite stumbling down the stretch.

 

[Max Power]

So if OU went 13-0 in the non-con, and 6-12 in the league... Losing your last 12 games in a row, that does not matter to the committee?

correct. The committee decided to treat every game as equal as far as timing is concerned. Only looking at who you beat and who beat you. doesn't matter if it was march 12th or November 5th

 

[Zimwillett]

You're using old selection methodology to make your point. If it is as good as you say then why make the change? Do you believe the change will improve things even better?

In addition, my focus was the at large bids. How many of the top seeds are at large?

If anything, I think the new criteria will improve the overall selection/seeding process. The NET is a far better metric than the RPI. The RPI did not really have a component to account for actual on court result and focused too much on Opponents' Winning Percentage and Opponents' Opponents' Winning Percentage.

The NET is cleaner. It accounts for real world factors that measure how good/bad a team is like scoring margin (although I don't think they should cap it at 10), game location, and offensive and defensive efficiency.

I also like placing the value of wins into "Quadrants" that actually account for where the game is played. As Big 12 fans, I think we all know how tough it can be to go win in places like Stillwater and Morgantown, no matter how good or bad OSU or WVU are. The previous system just didn't really account for that.

That's not to say there won't be individual anecdotal examples of the committee "getting it wrong." There will always be some 10/11 seeds that go in and play terribly, while some good team that got left out goes and wins the NIT. That'll happen regardless of what criteria are used.

 

[BoulderSooner]

So if OU went 13-0 in the non-con, and 6-12 in the league... Losing your last 12 games in a row, that does not matter to the committee?

no

not under the current criteria

 

[thebigabd]

correct. The committee decided to treat every game as equal as far as timing is concerned. Only looking at who you beat and who beat you. doesn't matter if it was march 12th or November 5th

Interesting.

How do you feel about that? If you lose your last 12 games in a row should you ever make the NCAA Tournament? Seems unlikely that would happen, but it makes you question the methodology.

 

[Zimwillett]

In addition, my focus was the at large bids. How many of the top seeds are at large?

I'm not sure I understand. Sometimes 1 seeds are "at large." Sometimes they are not.

I haven't studied it closely, but my feeling is the only seeds that are universally never "at large" are the 14/15/16 seeds. With the "First Four" I think it's also less common for a 13 to be an "at large," but I don't know that for sure. What I do know for sure is that in 1999 OU got in as an at large 13 seed and made a pretty memorable run.

 

[BoulderSooner]

In addition, my focus was the at large bids. How many of the top seeds are at large?

I used top 4 seeds as the limit for "top" and the ESPN projection

1 seeds - 1 at large
2 seeds - 3 at large
3 seeds - 2 at large
4 seeds - 4 at large

so 10 of the top 16 seeds are at large

 

[WichitaSooner]

My point is do you expect the new process to offer up different results.

Ok, thanks for clarifying since I am apparently dumber than usual today! Not necessarily different results in any meaningful way as far as which teams advance deep in the tournament, but maybe it will make a slight difference in ensuring that teams who have the best overall bodies of work make it, as opposed to the past system that rewarded teams who simply got hot down the stretch.

 

[Zimwillett]

Ok, thanks for clarifying since I am apparently dumber than usual today! Not necessarily different results in any meaningful way as far as which teams advance deep in the tournament, but maybe it will make a slight difference in ensuring that teams who have the best overall bodies of work make it, as opposed to the past system that rewarded teams who simply got hot down the stretch.

Perfectly stated. The new metrics aren't really going to change much. My guess is the field and the seeding this season will look virtually identical to what it would have been last season. Not sure how you'd do it, but I'd actually be interested to see someone try to game that out.

If such a thing did exist, I would expect to maybe see a few minor variations in seeding. I really have no way of knowing whether or how it would impact the bubble. This year is a little odd given how weak the bubble is, but my guess is it wouldn't be all that much different between the old and new systems.

 

[Max Power]

Interesting.

How do you feel about that? If you lose your last 12 games in a row should you ever make the NCAA Tournament? Seems unlikely that would happen, but it makes you question the methodology.

I'm indifferent, its the methodology they want to use as long as they are upfront and its not some secret formula for you to guess. If games in November and December don't matter, why should you play in holiday tournaments or showcase games?

 

[stormspencer]

Ok, thanks for clarifying since I am apparently dumber than usual today! Not necessarily different results in any meaningful way as far as which teams advance deep in the tournament, but maybe it will make a slight difference in ensuring that teams who have the best overall bodies of work make it, as opposed to the past system that rewarded teams who simply got hot down the stretch.

I also believe that it will further encourage teams to schedule a more difficult non-conference schedule.....even for the P-5 teams. Case in point would be North Carolina State this year. Their non-con SOS is 352nd out of 353 teams in the country. Their metrics aren't terrible, but if they don't win @BC and then lose their opening round game in the ACC tourney (likely against another bubble team - Clemson), they aren't going to get the benefit of the doubt from the committee.

No matter what metrics the committee has used in the past or how they have evolved to this day, one of the few constants that the committee rewards is having a tough schedule. The committee would much rather you schedule some tough games and even lose a few rather than inflate your record on a bunch of pansies.

 

[sheepdogs1]

I'm indifferent, its the methodology they want to use as long as they are upfront and its not some secret formula for you to guess. If games in November and December don't matter, why should you play in holiday tournaments or showcase games?

It is all reward for the team, it's fans, and it is a tradition. The sponsorship probably brings money into the coffers of the NCAA as well.

 

[Zimwillett]

It is all reward for the team, it's fans, and it is a tradition. The sponsorship probably brings money into the coffers of the NCAA as well.

Yes. But I could argue that those early tournaments and non conference games more closely approximate what you will see in the Big Dance than you will in conference play---especially in the Big 12. Neutral sites, unfamiliar opponents, etc. Not saying those are like the Big Dance---they aren't. Just that they're more similar than conference games.

I actually agree that it would be good if the committee could figure out a way to somehow factor in how a team is playing late. The problem is I don't think they've found an objective quantifiable way to do it. How do you quantify whether or not a team is "hot" or "playing well"? The last 10 games thing was kind of arbitrary and unfair. Why cut it off at 10? Worse still, how do you differentiate between Team A who played 6 Q1 games in their last 10 vs. Team B who played none?

I think the committee defaults to "entire body of work" because they don't want to get into this quagmire, and I'm fine with that.

 

[WichitaSooner]

Yes. But I could argue that those early tournaments and non conference games more closely approximate what you will see in the Big Dance than you will in conference play---especially in the Big 12. Neutral sites, unfamiliar opponents, etc. Not saying those are like the Big Dance---they aren't. Just that they're more similar than conference games.

I actually agree that it would be good if the committee could figure out a way to somehow factor in how a team is playing late. The problem is I don't think they've found an objective quantifiable way to do it. How do you quantify whether or not a team is "hot" or "playing well"? The last 10 games thing was kind of arbitrary and unfair. Why cut it off at 10? Worse still, how do you differentiate between Team A who played 6 Q1 games in their last 10 vs. Team B who played none?

I think the committee defaults to "entire body of work" because they don't want to get into this quagmire, and I'm fine with that.

Best explanation I heard the committee give on this is the golf analogy. At the end of a round of golf, the low man wins. Doesn't matter whether he played better on the front nine or back, or which holes he scored better or worse than his competitors. Add them all up and that tells you who played the better round. That is how the committee views the season.