Sunday, December 6, 2015

"The Ivy League ---Just as dumb as everyone else" -- thoughts on the proposed post-season basketball tourney

I read a couple days ago that the Ivy League is on the verge of implementing an annual post-season basketball tournament

After 100 years or so, you think you know a conference...

Every Division I conference is granted single automatic bid for the conference champion to participate in the annual NCAA Men's basketball Tournament ("March Madness").

Conferences are allowed to designate the terms by which their champion is determined.  Originally, it was pretty sensible. A conference champion was the team with the best record in conference play.  In the event of a tie, tie breakers or some other means determined the champion.

The Ivy League has been smart and awarded their bid to their regular season champ.

The Ivy League system for breaking ties was exceptionally well designed --- the teams tied for the league would play each other in a single game ---winner takes the bid;  Both teams get a SOS bump.  The team that makes the tourney might end up one seed higher than they would have been ranked before the game; The loser would stand a better chance of getting into the CBI or the NIT or Bob's Fishmarket's Off-brand Post-season Tournament with a higher seed than they would have had without the game.

It made sense.  It was smart.  It was what you expect from smart people. 

Now it appears they are just as dumb as the rest of us.

It appears they will be using the new post-season tournament to determine which member of their conference will use the conference's automatic bid to the NCAA Tournament,  just like every other NCAA Division I basketball conferences.

Why is this dumb you may ask? 

 Let me tell you...

What you are giving up when you let a post-season tourney chose your champion

In college basketball, when you send your best regular season team,  you are usually sending a team that has lost 3 to 11 games out of 27-31 games.  So that team will have won roughly 59 to 90% of their games over the course of the season.  (Usually far closer to the 80-90% range...)

A team that wins more than the other teams in their conference is not necessarily more talented...although they can be... but they always are more skilled at winning.  

They have figured out how to go on runs at the right time.  They have figured out when a burst of hustle can create a turnover that destroy's an opponent's confidence.  They can smell blood in the water when a game is on the verge of turning.

Conference tournaments allow a twisted ankle or flu outbreak to flush a season of hard work down the toilet.  It invalidates all the hard work your student athletes put in learning to give consistent, smart, focused effort.  The very things college is supposed to encourage and develop are not rewarded by giving an automatic bid to the NCAA tournament to a team that sucked all season, only made the conference tournament because everyone was invited, and had 2-3 good games.

So it screws the students who have practiced the lesson of competing every day. It steals from them.

Those kids have earned that berth.  They have earned the right to compete against the best of the collegiate ranks. They have earned the right to potentially have a bad game on the biggest of stages, not have the right stolen to prop up a small-time tournament.

We wouldn't steal opportunities from are own kids, so why is this acceptable?

I would hope that thought would be enough to get school presidents and athletic directors rethinking that policy, but I am jaded enough to see that as smart as all of those people are, their attentions are on more compelling matters than sports minutia.  So they are given to groupthink and rely on conventional wisdom instead of actively thinking through peripheral sports matters such as these.

So let me give them a reason that touches closer to home.  Collegiate sports are just advertisements for a university.  A conference membership is losing money and  losing exposure when they allow their champion to be replaced by one of their dog teams.

It is one thing for a conference above the mid-major level to have a "winner take all" post-season tournament that reassigns their bid to one of their bad teams who got hot.  

There are 347 Division I teams.  There are 65 teams in the 5 football power conferences (The ACC, SEC, PAC-12, Big 12, and Big Ten).  Those 65 dominate college recruiting. 

The worst team in a power conference is more physically talented than most of the other 282 other schools at the DI level.  If that power 5 dog of a team finally figures out how to win, they will be dangerous in the NCAA tournament because they got hot and (for the NCAA Tournament) have really good talent.

I would argue even from the next conference ---Big East --- downwards, the worst teams in these non-power conferences do not have the talent to go on deep runs in the NCAA tournament.

Further, the NCAA Tournament selection committee seeds the teams a conference send to the tournament based largely on their resume.

Let me put the implications of this in perspective.

RPI =  a bunch of crap used to prop up the power conferences

RPI is a statistic used to compute the relative strength of teams and conferences vs. each other.  It is a very limited and frankly, stilted measure, but it is well-established as numbers the NCAA selection committee uses to evaluate teams for the tournament and widely accepted. 

Let's take a look at the last 5 years.

RPI 2012-16
Conference 2015-6 (so far) 2014-5 2013-4 2012-3 2011-2 Average
Big 12 1 1 1 5 3 2.2
Big East 2 2 4 3 2 2.6
Big Ten 6 4 2 2 1 3.0
ACC 4 3 5 4 6 4.4
PAC-12 3 5 3 6 10 5.4
Atlantic 10 5 7 6 7 7 6.4
SEC 9 6 7 8 4 6.8
MWC 9 11 10 1 5 7.2
AAC 7 8 8 7.7
MVC 15 12 11 9 8 11.0
WCC 17 9 9 10 11 11.2
CUSA 18 16 13 11 9 13.4
Horizon 14 15 14 12 14 13.8
MAC 13 10 12 18 17 14.0
CAA 10 18 15 24 15 16.4
Summit 12 20 17 19 16 16.8
Sun Belt 19 19 19 15 19 18.2
MAAC 23 21 16 14 18 18.4
Big West 11 13 22 21 26 18.6
Ivy 26 14 18 23 13 18.8
Patriot 26 17 20 17 22 20.4
WAC 28 30 21 13 12 20.8
Ohio Valley 20 24 24 16 21 21.0
Southern 16 23 30 27 23 23.8
Atlantic Sun 29 29 23 26 20 25.4
Northeast 32 25 26 20 24 25.4
Southland 22 27 27 25 28 25.8
Big Sky 27 26 25 28 25 26.2
America East 24 28 29 22 29 26.4
Big South 30 22 28 29 27 27.2
SWAC 21 32 32 31 31 29.4
MEAC 31 31 31 30 30 30.6

(data from realtimerpi.com)


Over that period the Ivy League's RPI average was very similar to that of the MAAC, Big West, and Sun Belt Conferences.  All three of those other conferences use a post-season tournament to assign their NCAA tournament bid, so they will allow us to look at this decision by the Ivy League clearly.

Evaluating basketball peer conferences to the Ivy League

Here are the per year conference champions and the conference's representatives for the NCAA tournament for the Ivy league and their basketball peers.

2014-15
Conference school rec reg. seas. tourney t. Record
Sun Belt Georgia State* 24-9 14 v. 3 1-1
MAAC Manhattan^ 19-13 16-13 16 v. 16 0-1
Iona* 26-8 24-7 - 0-1
Big West UC-Irvine^ 21-12 18-12 13 v. 4 0-1
UC-Davis* 25-6 24-5 0-1
Ivy Harvard* 22-7 21-7 13 v. 4 0-1
2013-14
Sun Belt UL^ 23-11 20-11 14 v.3 0-1
Georgia State* 24-7 25-8 0-1
MAAC Manhattan^ 25-7 22-7 13 v. 4 0-1
Iona* 22-10 20-9 0-1
Big West Cal Poly^ 13-19 10-19 16 v. 16 1-1
UC-Irvine* 23-11 22-10 0-1
Ivy Harvard* 26-4 12 v.5 1-1
2012-13
Sun Belt WKU^ 20-15 16-15 16 v. 1 0-1
MTSU* 28-5 27-4 11 v. 11 0-1
MAAC Iona^ 20-13 17-13 15 v. 2 0-1
Niagara* 19-13 18-12 0-1
Big West Pacific^ 22-12 19-12 15 v. 2 0-1
LB St.* 19-13 18-12 0-1
Ivy Harvard* 19-9 14 v. 3 1-1
2011-12
Sun Belt WKU^ 15-18 11-18 16-16 1-1
MTSU* 25-6 25-5 2-1
MAAC Loyola^ 24-8 21-8 15 v. 2 0-1
Iona* 25-7 24-6 14 v.14 0-1
Big West LB St.* 25-8 22-8 12 v.5 0-1
Ivy Harvard* 26-4 26-4 12 v. 5 0-1

So The Ivy League's champion has been seeded #12, #14, #12, and #12 in the last four years.  They have turned that into a 2-4 NCAA tourney record.

The other three conferences have had a team besides their regular season champion be given the conference bid 10 out of 12 times. 

10 out of 12 times their champion was displaced.  That should be jarring to you.

Based on that sample, the Ivy league can expect their best winners to lose their bid in the Ivy League post-season tournament roughly 82.5% of the time --- let's say 1 out of 5 years the champ will keep the bid.

And it gets worse.

Those replacements have been seeded #15, #16, #15, #15, #16, #16, #13, #14, #13, #16 and they combined for a record of 2-10 (the two wins came in the 16th seed vs. 16th seed play-in game...The NCAA tourney's equivalent of the overweight fan's halftime half-court shot attempt).

The two times their actual champions survived the post-season conference tourney without dead legs, colds, or fluke plays taking them down, they were seeded #12 and #14 and went 1-2. 

(It should be noted that two of their champions who were displaced had strong enough wins during the season to earn at-large bids, but the committee punished them by making them play a matching seed for a play-in game. Both schools lost in the play in round, so 0-2.)

Conceptually what you should get is it is very bad to be seeded 14th -16th.   We will cover this in a second.

Why NCAA tournament seedings matter

Here are the records of seeded teams in their round of 64 matchups over the last 19 years.

1996-2014
matchup record percentage
1 v.16  0-76 0.0%
2  v. 15 3-73 3.9%
3 v. 14 8-68 10.5%
4 v. 13 16-60 21.1%
5 v. 12 32-44 42.1%
6 v.11 27-49 35.5%
7 v.10 33-43 43.4%
8 v. 9 35-41 46.1%

It's a statistical bitch slap to be a 16th seed.  You get a participation ribbon.

A 15th seed is only slightly better. A 14th seed has about a 10% chance of survival. A 13th seed is double that.  Then a 9th to 12th seed are fairly good odds for an underdog.

Why is this?

Most media types who follow NCAA basketball are familiar with the basketball concept of  "The Seven Power Conferences" even if they call them something else.  Those are the 5 football power conferences plus the new Big East and the Atlantic 10.  Those conferences are considered elite because they regularly earn multiple berths into the NCAA tournament.

Usually the top 12 teams seeded as top 3 seeds in the 4 regions in the NCAA tournament are largely comprised of the top 2 teams in those 7 conferences---their champions and the team that they edged out for the title.  (Well there is a little more to it...You could probably subtract the SEC and Atlantic 10 runner ups and you maybe throw a Gonzaga in there. ...)

The point being that the top 12 teams are usually not only super talented, but also know how to win as a team.   The next 20 or so teams? (The 4th, 5t, 6th, and 7th seeds?)  Usually almost as talented, but with tragic flaws that make them very beatable.

"They have a dominant center...but they lost their starting point guard a month ago and their backup went down last week.  Will their 5'6" walkon freshman 3rd stringer be able to bring the ball up the court?"  That kind of thing...

Champions from the 25 conferences beneath the Power 7 tend to feature teams heavy with upperclassmen with one or two lesser stars who might get drafted by the NBA who make their teams dangerous.

Those kinds of champions are not usually going to beat the top dozen teams, but they are more than capable of knocking off a middling power conference squad in a 4th to 7th seed, because they are the sounder squad and the more experienced winners. (If you caught something there that sounds like I have a major problem with NCAA seeding philosophy...It's because I do. An article for another day.)

Here is a compressed view of the odds to make things clearer.


Seeding Records Odds of winning
14-16 11-228 4.8%
13 16-60 21.1%
8-12 127-177 41.8%

As I was saying earlier, the NCAA selection committee judges any school a conference gives an automatic bid to by that team's basketball resume for the year.

When you swap a proven conference champion from the lesser 25 for a crap team that only beat your best in the tournament because of an injury ....or due to the fatigue of playing 3 games in rapid fire succession ----something a champion usually isn't forced to do in the NCAA Tourney --- you turn your seeding from a likely 12-14 seeding to a 15th or 16th seed.  Your odds of seeing the next round drop from at high as 40% down to potentially 0%.

You move from the competitive seeding tier to the semi-competitive one or from the semi-competitive one to the participatory ribbon tier.

This not only costs the conference exposure, it directly costs the conference money.

The NCAA Tournament pays a team a tourney share for each game it participates in.  This means each time you win and earn the right to play another game, you get another share of the big pot of TV money.

Forbes estimates a share today is worth $1.65 million.  Shares add up, as this excellent article by the Washington Post explains.

Most of the lesser 25 conferences earn a share a year for their single team that makes the tournament and gets whacked. 

Using their current methodology, the Ivy League earned 6 shares in 4 years while their basketball peer-level conferences earned a 3-14 record, seventeen shares, which divided by 3 equals 5.4 shares.

OK, financially that doesn't look like a huge difference.   It's only $990,000 of direct tourney money the Ivy league would have potentially surrendered.

(Now it could have been $10.89 Million, if the Ivy league had been able to maximize their setup and get their second best team in and properly seeded as well by tweaking strength of schedule, but I get the fact that the Ivy League doesn't appear all too concerned about strength of schedule --- they likely weigh it appropriately vs. the hardship on their student athletes...)

The Ivy peer conferences also had the dumb luck to send the second worst team to the tourney two years out of our five year sample skewing the difference.  By propping up their stats with a totally underserved win in a play-in game vs. an even mangier dog, these conferences were gifted two shares (what divides down to another $660,000 per school) with their fancy participation ribbons.

 Let's be clear if I haven't been.  Those games are not really part of the tourney.  No viewer gives a crap about the "first four" in the tourney---especially the 16 vs. 16 matchup.  Most do not even watch them.

You have to think about the value added.

Which of the four conferences looked most competent over the last 4 years? 

Which conference's teams looked the most dangerous?

Which is most likely to have earned fans among the teens going to college today?

That is the value of the NCAA Tournament---promotion of your brand nationally.

Playing a conference tournament to decide who will represent you is like choosing to accept a "participation trophy" for the NCAA tournament each year. 

A lesser 25's dogs are not going to win anything.  They will never be the Cinderella of the tournament.

It wastes the open mindedness and goodwill of the viewers.  Sending your dogs just earns a conference the contempt of TV viewers.

Certainly you chose what exactly you want to value, but if the 2-3 best Ivy League schools chose to ramp up their OOC schedules with stronger teams from the bottom 5-10 conferences, there is a fairly good chance an Ivy without a post-season tournament might send two teams to the NCAA Tournament fairly regularly with both participating Ivy teams seeded in that #12/#13 sweet spot where they can do some damage.

Now I have been a little unfair to the Ivy League in that there are some valid reasons to consider a post-season tournament.  There really are.

My point is that it would be smart not to use it to determine who gets your automatic bid. 

Use it to showcase teams for the NIT and the other tournaments and to excite alumni, but don't fall into the trap of burning your chance to advance in the NCAA tournament. 

One tournament is out facing to the greater public and one faces in to people who already have a vest interest in the Ivy League.  Don't make the mistake everyone else makes and sacrifice the more important tournament to prop up your less important one.

All I am saying is...Just be smart about it.




2 comments:

  1. I always advocated having a conference tournament, but only invite the top 4 or 8 teams at the end of the regular season. That way, the regular season has greater significance, while the conferences still get the financial windfall that comes with a post-season tournament (more on that below). It sounds like the Ivy used to try something similar, but oh well. Straight-chalk tournaments are always more exciting IMO--not these bloated brackets with all 12-15 teams playing over the course of three nights before you even get to the semifinals. A conference tournament would be more exciting if it was something you had to EARN over the course of the regular season.

    I may be wrong about this but, based on previous understandings, the financial model for Men's Basketball has always been infinitely simpler (by that I mean as opposed to football obviously). In other words, men's basketball (unlike football) generates revenue for a conference primarily in one of three ways: 1) Television contracts; 2) Number of bids in the NCAA tournament; and 3) conference tournament ticket sales. Unfortunately, I think schools are leaving too much money at the table these days by passing on Option #3. It also explains why conference tournaments include the full membership, as opposed to just the best teams. The revenue sharing model for ticket sales is different than most other types of the "equal revenue sharing" implemented by most conferences (Here I will use the ACC as an example because I am most familiar with it. I only assume most other conference follow a similar model). In terms of ticket sales, each conference member receives an "equal" allocation of tickets. But it is up to each individual school to sale their share of tickets and therefore generate income from those sales. Thus, schools like Duke and UNC benefit handsomely from conference tournaments, because they always sale out of their allocation(s) well in advance.* Small private schools like Wake Forest or Boston College may do well (generally) because they have much smaller alumni bases and so they have more tickets to go around. Schools like Clemson and FSU, on the other hand, don't sale as well. Usually (Clemson anyway) has tickets remaining and attempts to sale them directly to students, fans and boosters after making the initial offering to season ticket holders. Sometimes (I think) they even end up selling some of those tickets back to the conference, who then sales them directly through the host venue (e.g., Greensboro Coliseum, ect.). Ultimately, though, this is why everyone jumps on the bandwagon--and why we can't make those bandwagons any smaller by limiting the number of seats available to ride in.


    *[This coincidentally was also the main reason why UNC and Duke always opposed the ACC's initial expansion back in 2003--because it meant a smaller allocation of tickets would be made available to each school].

    ReplyDelete
  2. I apologize for my delay in responding. It certainly was not due to quality... This was a top notch response which pretty thoroughly covers the valid reasons I mentioned to actually have a post-season tourney.

    I share your opinion that there is a sweet spot in there for a cap on the number of invited schools. If the conference is going to put their slot up for grabs, there is a reward that combined with slot scarcity could drive attendance improvements and ticket prices. Perhaps at some point for some conferences that could balance out against the annual damage giving your slot to an underserving school creates...not sure.

    Thank you for spelling out the main options for revenue generation. It was a gross oversight in my article and you did it well.

    The Ivy league was the last DI conference not playing a post season tourney, so no one else is "leaving money on the table". I don't think most conferences are actually gaining much financially with a tourney. The vast majority of conferences do not draw huge crowds to their tourney games. Even if you are working from a premise that each school can sell their lots at a premium, do the costs of say FGCU sending their team to x random location for the Atlantic Sun Post Season tourney and selling their allotment of say 500 tickets outweigh what they could have playing a home game in front of 4500 fans?

    I think there is a point where the ACC conference tourney model becomes a financial and media loser as you go down the conference ranks. And the cost to their real champions is unforgivable.

    ReplyDelete