A look inside the NCAA selection process

[REM08]

I know most of you well enough to confidently say that this will be worth your time. I learned a lot and no longer suffer some of the misconceptions I had before reading this.

This is a blog post from a Kentucky blog. Matt Jones was one of the 20 media members invited to attend the media-mock selection process this year and he wrote about his 10 biggest takeaways from the event. I realize this is a Florida board, but 95% of the info is non-UK related and actually much more pertanent to Florida considering UK probably wont' even make the field. I think you'll find it worth reading.

Here's an excert:

3. There is NO Set Criteria for Tournament Entry. Anything Goes:

Chalk this up as surprising fact #1 about the process. I assumed that there were some guidelines that Committee members were supposed to use in determining which teams should get in. I was wrong. As the Chair Mike Bobinksi said to me at one point, “anyone can use whatever they think is most important.” I found that somewhat astonishing, but in watching the process in practice, it is clearly the case. Do you think games on the road are most important? Fine, then that can be your criteria. Do you prefer out of conference scheduling strength? That’s all you baby, go for it. In fact, even the so-called “eye” test is fair game if a specific committee member wants to use it (and at least one near me did). As Bobinski said to the group, “as long as the member is consistent in applying whatever standard is most important, that is totally fine.” What was disappointing to me is how little so-called “advance metrics” are used in the process. During the entire two days, efficiency ratings were mentioned one time, KenPom once and BPI not at all. RPI wasn’t brought up a great deal (another misnomer that it drives the Committee process), but more than these more modern statistics. In ten years this may change, but for now the statistics are still a bit behind for my liking.

http://kentuckysportsradio.com/?p=127722

[gatorbogey]

like a mosh pit of 'anything goes' chaos. somehow after all the tug 'n war, they have their seeding.

how is it all arbitrated? each member w/ a vote? then they have to come up w/ the 'cover story' for why a team got in vs why a team got left out and why the seedlines etc?

[GatorRade]

Interesting stuff. Thanks REM.

[gatorbogey]

pretty astounding to me that they didn't use some of the objective metrics out there - like sagarin. i wonder if they didn't openly refer to it, but hide it in their back pocket (that's what i'd do )

[GatorJeff]

Sounds like the Gators' incredible offensive and defensive efficiency won't get much notice.

[REM08]

Originally Posted by GatorJeff

Sounds like the Gators' incredible offensive and defensive efficiency won't get much notice.

Same for margin of victory apparently. I agree with the writer that the committee seems a little behind the times stats wise.

[regurgigator]

Interesting paragraph above, but if someone is using the eye test, how in the world would anyone know if the committee member is applying that criterion consistently

The eye test is completely subjective.


Not that I don't think the eye test is relevant. I do. It's the main reason I'm so high on the Gators believe it or not. The various incredible stats we're putting up just kind of reinforce what I believe I'm seeing on the floor.

[REM08]

Originally Posted by regurgigator

Interesting paragraph above, but if someone is using the eye test, how in the world would anyone know if the committee member is applying that criterion consistently

The eye test is completely subjective.


Not that I don't think the eye test is relevant. I do. It's the main reason I'm so high on the Gators believe it or not. The various incredible stats we're putting up just kind of reinforce what I believe I'm seeing on the floor.

Thats the thing about it, no one would. That doesnt' seem to bother them either. I'm actually not sure it bothers me. You have to hope these 10 people are as without bias as they can possibly be and that they take pride in what they're doing. Any attempts to influence how they quantify teams might help, but it also might hurt.

[REM08]

I think I anticipated this topic being a little more interesting than it, apparently, is.

Here are the things I found surprising:

1. They have a very systematic voting procedure. I guess this isn't shocking, but I like how its designed.

1. All ten members select teams that are (a) In the Tournament or (b) Under Consideration. If a team gets 8 of 10 votes on the “In the Tournament” selection, this first go around, it is officially in. If it gets 3 votes for being either In the Tournament or Under Consideration, it moves to round 2. If it doesn’t get 3 votes for Under Consideration, it is out, with no further discussion absent a majority vote. It’s that simple.

2. After those “locks” are in (and in our case, 19 got in via this process), you begin rounds of voting. It goes like this…each member picks 8 teams that they believe should be in. The 8 teams that get the most votes are then moved to a second round, where each member ranks them 1-8. The four teams with the best total score are in the Tournament. This is then repeated over and over until the field is completed.

3. Once all teams are in, they are ranked seeded 1-68 (adding the conference tournament winners). This process is similar…each member selects his or her pick for the 8 best teams in the tournament. The 8 teams that are picked the most are then ranked 1-8 by each member. The top 4 vote getters become the #1 seeds.

Its a simple process and with the exception of the “scrubbing” feature (discussed later), it is how the bracket is created. Teams are debated multiple times and the layers of votes tend to rid the process of outlier votes. After watching it in action, I think it is a great way to create and seed the field.

2. While no specific objective team stats or ratings are encouraged over others, there is a cheat-sheet of sorts that comes with each team. This cheat sheet contains these stats:

Overall Record
Record vs Tourney Teams
Record vs RPI Top 50
Road Record
Strength of Schedule
Out of Conference Strength of Schedule
Losses to Teams below RPI 150
Non-Conference Record

Whoever picked these eight stats to put on the nitty-gritty sheet (and we don’t know who it is), by their very placement, made them the most important. He “Nudged” them to the most prominent spot. Teams should schedule accordingly.

3. Good wins matter much more than bad losses and margin of victory doesn't really matter.

QUALITY WINS MATTER MORE THAN ANYTHING ELSE

I put this point in all caps, because in terms of what actually makes the teams get in the tournament or not, I think it is the most important. Yes, we talked about lots of team qualities, but nothing got brought up more times than “who did you beat.” In every debate about a team, the question of their “best win” was highlighted. And more often than not (I would guess 80% of the time), the team with the best win won out on a close choice.

4. Teams are evaluated as they stand (roster wise) on selection Sunday.

5. NCAA Relies on Teams for Information: Another little juicy tidbit. How does the NCAA find out whether players will actually play in the Tournament while seeding? Short answer is that they rely on the teams. They ask the teams privately to tell them if a player will be ready, and that information does not have to be released to the media. Does that give a team incentive to lie? Why yes it does. Case in point, Ryan Kelly and Duke this year. The NCAA is relying on Duke to tell it whether Kelly will play, and the Blue Devils are supposed to give that information. What happens if Duke, or any other school, lies? Chair Mike Bobinski answered with a smile, “we learn not to trust them in the future.” But there is no actual punishment.

5. No S-curve. This just doesn't even exist. Its not like they just try to even things out but have to succumb to regional issues - they dont' even bother caring about this. If the best 1 seed, best 2 seed, best 3 and best 4 end up in the East - thats fine by them. All 1 seeds are treated equally.

6. According to this media member, he's now "anti-conspiracy theory" when it comes to how brackets are made. What stands out to me is that marquee matchups are not sought out. I was always sure this happened. Apparently its just a coincidence if it happens.

There are many things wrong with the NCAA, and college basketball in particular, needs a re-examination of some of its most fundamental principles. But after two days in Indianapolis, one thing that I do think is right with the NCAA is how they pick the Tournament. The people involved take their job very seriously and they know their stuff. David Worlock and the rest of the staff have the process down pat and spend a great deal of times making sure everything is in place. More importantly, I honestly believe that what they say, the fact that the Committee doesn’t consider conference affiliation, tv ratings, etc in selecting its teams, is true. I know many of you are rolling your eyes and saying, “come on Matt, whatever. They wouldn’t tell you they considered that stuff even if they did.” And you are right. But after going through the process, I am not sure how they even could. There are so many rules, layers of voting, etc that conspiracy theories simply wouldn’t have a chance to take shape. During our entire two days, I am not sure it was ever brought up once what conference a team was in. During the bracketing, marquee matchups weren’t sought out, even though a few were created randomly (such as a potential Rick vs Tubby game in Rupp). The process weeds out all the bias for as close to a fair bracket as can be created. It isn’t perfect, I think much more modern analytics should be used and the decision to not consider margin of error seems absurd, but it is pretty damn good. Going through the process made me respect what they do and want to be part of a real one. The latter may not happen, but the former will be a part of how I analyze ever bracket going forward.

[ufgator4ever]

' They ask the teams privately to tell them if a player will be ready, and that information does not have to be released to the media. Does that give a team incentive to lie? Why yes it does.'
So I sure hope yeguette plays in the SECtourney as is consistently being advertised by the team/coach publicly. If he hasn't played by selection sunday, not sure committee would have any confidence in the team/coach continuing with that same line he'd be available.

Unless of course they take whatever the team/coach tells them as gospel truth. In which case, I definitely think Noel will be available for the tourney. Privately, per calapari

[gatorbogey]

that is some interesting revelations.

i thought they still put the teams up on the S-curve and that set the matchups, then they would move teams around to avoid conf matchups, look at the geographics the also attempt to get balance in the regions. i don't think it would be too hard to do that w/ the process outlined.

so, instead of the S-curve, sounds like they just go down the seed lines and pick the 1's, then 2's and so on.