Week of 1/19 Bracketology

Illinois a 4 seed in the Bracket Matrix
bracketmatrix.com

Illinois #7 in the NET Rankings

4 of the top 7 teams in the NET are in the B1G.

Nothing about our profile says 4 seed… we’re a 3 or better imo.

SkyysNotFalling said:

Nothing about our profile says 4 seed… we’re a 3 or better imo.

Click to expand...

If you seed out the AP #13 rank, it matches ironically. All that data manipulation and a 4th grader could come up with the same answer!

SkyysNotFalling said:

Nothing about our profile says 4 seed… we’re a 3 or better imo.

Click to expand...

I wouldn't say nothing.

-KPI is 16th and Strength of Record is 14th.
-There are 10 teams with more Q1 wins and another 4 teams with the same number of Q1 wins but less Q1 losses.
-16 teams have more Q1 + Q2 wins than we do.

I agree that we're a 3 seed if the season ended today but we might be closer to a 4 than a 2 right now.

We really need to win 1-2 of our 4 big regular season games remaining if we want a 2 seed.

Noblesville Illini said:

If you seed out the AP #13 rank, it matches ironically. All that data manipulation and a 4th grader could come up with the same answer!

Click to expand...

Most 4th graders at least know how to use Google though...

The NCAA Tournament Selection Committee does not even consider AP Poll rankings.

Key factors the Committee uses:

- NET Rankings
- SOS
- Quality Wins/Losses
- Team Sheets (compiled metrics)

FreshlySqueezed said:

I wouldn't say nothing.

-KPI is 16th and Strength of Record is 14th.
-There are 10 teams with more Q1 wins and another 4 teams with the same number of Q1 wins but less Q1 losses.
-16 teams have more Q1 + Q2 wins than we do.

I agree that we're a 3 seed if the season ended today but we might be closer to a 4 than a 2 right now.

We really need to win 1-2 of our 4 big regular season games remaining if we want a 2 seed.

Click to expand...

Not worth debating with all the quality games left on the schedule, but our resume looks like a 4 to me. What's our best win, Texas Tech at home?

Calvin said:

Not worth debating with all the quality games left on the schedule, but our resume looks like a 4 to me. What's our best win, Texas Tech at home?

Click to expand...

Iowa at Iowa at least by current Kenpom. UT in Nashville was “neutral court” but likely as good as Tech at home

Calvin said:

Not worth debating with all the quality games left on the schedule, but our resume looks like a 4 to me. What's our best win, Texas Tech at home?

Click to expand...

Q1A games are top 15 at home, 25 neutral, and 40 on the road.

So we have two Q1A wins, @ #20 Iowa and @ #37 Ohio St. @ #20 Iowa is our best win. Home against #19 Texas Tech is fringe Q1A and so is neutral against #26 Tennessee.

Two Q1A wins is decently competitive. The top teams (1 and 2 seeds) tend to have more. But teams in the same projected seed range as us:

Houston: 1
Iowa St: 2
BYU: 1
Michigan St: 0
Vanderbilt: 1
Florida: 1
Louisville: 0
Kansas: 2

NarrowJ said:

Most 4th graders at least know how to use Google though...

The NCAA Tournament Selection Committee does not even consider AP Poll rankings.

Key factors the Committee uses:

- NET Rankings
- SOS
- Quality Wins/Losses
- Team Sheets (compiled metrics)

Click to expand...

Oh, I’m aware. Just thought the irony is notable. Most ‘experts’ would still put us on the 3 line.

Noblesville Illini said:

Oh, I’m aware. Just thought the irony is notable. Most ‘experts’ would still put us on the 3 line.

Click to expand...

Just thought your 4th grader comment was saucy so was giving it back a little bit

On a serious note I agree with others that we are closer to a 4 than a 2 right now

I know I’m going full Captain Obvious here, but the time period between now and February 7th presents a great opportunity to solidify seeding. Having road games at Purdue, Michigan State, and Nebraska will go a long way toward our eventual seeding.

OrangeBlue98 said:

I know I’m going full Captain Obvious here, but the time period between now and February 7th presents a great opportunity to solidify seeding. Having road games at Purdue, Michigan State, and Nebraska will go a long way toward our eventual seeding.

Click to expand...

I have some time this morning, so here's a metrics bracket update. I'll skip the preamble, just keep in mind the polls mean nothing, NET (highest in each conference) is used for auto bid placeholders, resume metrics (SOR, WAB) are used to fill the at-large pool, and efficiency metrics (NET, KP, BT) are used to seed the field.

Auto bids (highest NET as of this morning):
B1G - Michigan (1)
ACC - Duke (2)
B12 - Arizona (3)
WCC - Gonzaga (4)
BE - UConn (8)
SEC - Vanderbilt (13)
MW - Utah State (21)
A10 - Saint Louis (22)
American - Tulsa (51)
MAC - Miami-OH (52)
SLnd - McNeese (60)
MVC - Belmont (61)
Ivy - Yale (67)
WAC - Utah Valley (81)
BSth - High Point (87)
CUSA - Liberty (91)
BW - Hawaii (99)
CAA - Hofstra (100)
SB - Troy (104)
Horz - Oakland (111)
MAAC - Marist (119)
Summit - St. Thomas (123)
SoCon - Mercer (127)
BSky - Portland St. (136)
ASun - Lipscomb (152)
OVC - UT Martin (168)
NEC - LIU (189)
Pat - Colgate (192)
AEast - Vermont (200)
MEAC - Howard (258)
SWAC - Grambling (267)

At-large field (top 37 resume average): Nebraska, Purdue, Houston, Michigan State, Iowa State, BYU, Illinois, Virginia, Alabama, Clemson, Texas Tech, Florida, Kansas, UCF, Arkansas, Villanova, Miami-FL, USC, Georgia, St. John's, Louisville, North Carolina, George Mason, Iowa, Wisconsin, St. Mary's, SMU, Kentucky, Seton Hall, Tennessee, Oklahoma State, San Diego State, Auburn, Missouri, Stanford, New Mexico, Ohio State

Just missed: Texas, Virginia Tech, Cal, Murray State, Baylor, Texas A&M, Santa Clara, Dayton, UCLA, LSU, Indiana, NC State

Seeding (based on efficiency average, auto bids in bold, change from last week in parenthesis):

1 (1). Michigan (+0)
2 (1). Arizona (+0)
3 (1). Purdue (+2)
4 (1). Duke (+6)
5 (2). Houston (+4)
6 (2). Gonzaga (+2)
7 (2). Illinois (+0)
8 (2). UConn (-2)
9 (3). Vanderbilt (-6)
10 (3). Iowa State (-6)
11 (3). Florida (+0)
12 (3). Michigan State (+1)
13 (4). Nebraska (+5)
14 (4). Virginia (+1)
15 (4). Louisville (+1)
16 (4). BYU (-2)
17 (5). Kansas (+3)
18 (5). Alabama (-6)
19 (5). Iowa (+2)
20 (5). Texas Tech (+3)
21 (6). St. John's (-2)
22 (6). Tennessee (-5)
23 (6). Arkansas (+3)
24 (6). Utah State (-2)
25 (7). Clemson (-1)
26 (7). Saint Louis (+5)
27 (7). Georgia (+1)
28 (7). Kentucky (NEW)
29 (8). Villanova (-4)
30 (8). North Carolina (-3)
31 (8). St. Mary's (+1)
32 (8). SMU (-3)
33 (9). Miami-FL (+0)
34 (9). Auburn (-4)
35 (9). Ohio State (-1)
36 (9). Wisconsin (NEW)
37 (10). UCF (+0)
38 (10). New Mexico (+0)
39 (10). San Diego State (NEW)
40 (10). Seton Hall (-4)
41 (11). USC (-2)
42 (11). Tulsa (+4)
43 (11*). Missouri (-1)
44 (11). Belmont (NEW)
45 (11*). Oklahoma State (-4)
46 (12*). Stanford (-1)
47 (12). McNeese (-3)
48 (12). Miami-OH (+1)
49 (12). Yale (+2)
50 (12*). George Mason (-2)
51 (13). Utah Valley (+2)
52 (13). High Point (-2)
53 (13). Liberty (+1)
54 (13). Hawaii (+1)
55 (14). Hofstra (-3)
56 (14). Troy (+4)
57 (14). Oakland (NEW)
58 (14). Marist (+0)
59 (15). St. Thomas (-3)
60 (15). Mercer (-1)
61 (15). Portland St. (+0)
62 (15). Lipscomb (NEW)
63 (16). UT Martin (+0)
64 (16). LIU (+1)
65 (16*). Colgate (NEW)
66 (16*). Vermont (+0)
66 (16*). Howard (+1)
67 (16*). Grambling (-1)

Like last week, I don't think George Mason will hold on to an at large bid, either they'll sweep the A-10 and get the auto bid, or they'll lose and their resume will get passed by enough major conference teams that they'll fall off, but for now, they're 17-1 and have the resume. Miami-OH just passed Akron for best NET in the MAC which solves that minor issue, but I still think a 31-1 Miami-OH that loses to Akron in the MAC championship is getting an at-large bid, no matter their efficiency.

I keep thinking Indiana will eventually build their resume, and then they go ahead and lose badly every chance they get to improve it, so who knows at this point.

FreshlySqueezed said:

I wouldn't say nothing.

-KPI is 16th and Strength of Record is 14th.
-There are 10 teams with more Q1 wins and another 4 teams with the same number of Q1 wins but less Q1 losses.
-16 teams have more Q1 + Q2 wins than we do.

I agree that we're a 3 seed if the season ended today but we might be closer to a 4 than a 2 right now.

We really need to win 1-2 of our 4 big regular season games remaining if we want a 2 seed.

Click to expand...

Resume is a 4 (14th SOR, 13th WAB), Efficiency is a 2 (7 NET, 7 KP, 6 BT).

Win these upcoming Q1A games (Purdue, Nebraska, MSU, Michigan), and the resume will jump up fast.

FreshlySqueezed said:

I wouldn't say nothing.

-KPI is 16th and Strength of Record is 14th.
-There are 10 teams with more Q1 wins and another 4 teams with the same number of Q1 wins but less Q1 losses.
-16 teams have more Q1 + Q2 wins than we do.

I agree that we're a 3 seed if the season ended today but we might be closer to a 4 than a 2 right now.

We really need to win 1-2 of our 4 big regular season games remaining if we want a 2 seed.

Click to expand...

I agree, we are probably the 3rd best 3 right now.

It's REALLY misleading to say that the AP rankings just have NOTHING to do with the seed a team gets. Just because they aren't an official metric used by the Committee doesn't mean they aren't usually very, very closely related to what ends up happening. In other words, it's not like being ranked #6 in the AP Poll is a positive factor that helps to LEAD toward getting a #2 seed ... it's more like if you are ranked #6 in the AP Poll, that is probably an excellent indication that you have done OTHER things that have you in the #2 seed conversation.

So, I feel like the AP Poll is more like a parallel measuring stick, and to say it "doesn't matter at all" or anything like that simply because it's not an official part of the formula is not really true ... JMO. If you assumed the seeds followed the AP rankings on the Monday after Selection Sunday, you'd obviously have #1-4 as the #1 seeds, #5-8 as the #2 seeds, etc. This was last year's seeds by AP rank. Green teams are in the "correct" quad, while red teams are outliers here.

#1 Seeds
#1 Duke
#2 Houston
#3 Florida
#4 Auburn

#2 Seeds
#5 St. John's (NY)
#6 Tennessee
#7 Alabama
#8 Michigan State

#3 Seeds
#9 Texas Tech
#13 Wisconsin
#15 Iowa State
#18 Kentucky

#4 Seeds
#11 Maryland
#19 Texas A&M
#21 Arizona
#22 Purdue

#5 Seeds
#12 Clemson
#14 Michigan
#16 Memphis
#25 Oregon

Obviously just a sample size of one, but as you go down the seed lines, it generally correlates less with the AP Poll. This makes sense, given pollsters might overreact to a conference tournament loss of a team ranked #10-25 a lot more than the Committee would (who might not consider the loss at all), and I would imagine the metrics generally get muddier as you approach the #4+ seeds than with the top? Either way, if you are ranked in the top 8 of the AP Poll, I think you can reasonably like your chances of getting a top 2 seed.

TL;DR

The AP Poll is separate from the metrics used by the Selection Committee, but it is NOT unrelated. It's a parallel ranking system that uses a lot less "objective" tools to come up with a ranking, but that more subjective ranking often still ends up properly ranking the top 10 teams in a very similar way as the Committee will. So, while not a contributing factor to your seed by itself, it's probably clearly a good sign if you have a good AP ranking.

Counter-argument: Correlation does not equal causation. The AP Poll is not a rating, it's a poll. Nobody involved in the poll is in the selection committee room. The AP Poll is not part of the team sheets. It's completely unrelated to selection, as several teams that were ranked in the AP poll were left out of the tournament entirely (famously Syracuse about 10 years back, ranked #24 in the poll but on the outside looking in because their actual resume and metrics were not tournament worthy).

The AP Poll is related to tournament selection the same way media attention is related to tournament selection, it gets a team more media attention, as they show up in the top 25 scores and top 25 lists, and they get the number in front of their name on the score bug, but it is irrelevant to the actual selection done in the committee.

OrangeBlue98 said:

I know I’m going full Captain Obvious here, but the time period between now and February 7th presents a great opportunity to solidify seeding. Having road games at Purdue, Michigan State, and Nebraska will go a long way toward our eventual seeding.

Click to expand...

I would go so far as to say the way the conference schedule was built this year, our second half schedule is far more important in racking up valuable wins. Our first half conference schedule for the most part was about avoiding bad losses.

Piotyr said:

Counter-argument: Correlation does not equal causation. The AP Poll is not a rating, it's a poll. Nobody involved in the poll is in the selection committee room. The AP Poll is not part of the team sheets. It's completely unrelated to selection, as several teams that were ranked in the AP poll were left out of the tournament entirely (famously Syracuse about 10 years back, ranked #24 in the poll but on the outside looking in because their actual resume and metrics were not tournament worthy).

The AP Poll is related to tournament selection the same way media attention is related to tournament selection, it gets a team more media attention, as they show up in the top 25 scores and top 25 lists, and they get the number in front of their name on the score bug, but it is irrelevant to the actual selection done in the committee.

Click to expand...

I mean, you are kind of saying what I was trying to say anyway, though you downplayed how I would have said it. My point is that while entirely separate on its own, if you are getting recognition by the AP Poll saying you are a top 10 team, this separate recognition is probably a good sign that you are on track for a top 2 (or at least top 3) seed. I never said there was any causation, my point is that there IS a correlation, so folks are sort of going overboard when they act like the AP Poll is as good as toilet paper simply because it isn't an actual, baked-in ingredient in the selection process.

Fighter of the Nightman said:

I mean, you are kind of saying what I was trying to say anyway, though you downplayed how I would have said it. My point is that while entirely separate on its own, if you are getting recognition by the AP Poll saying you are a top 10 team, this separate recognition is probably a good sign that you are on track for a top 2 (or at least top 3) seed. I never said there was any causation, my point is that there IS a correlation, so folks are sort of going overboard when they act like the AP Poll is as good as toilet paper simply because it isn't an actual, baked-in ingredient in the selection process.

Click to expand...

But I think the point that was made was that it *is* as good as toilet paper to the tournament committee

AP Poll is not something that goes on any of the material they use to seed teams

It's useful to folks outside of that committee like all of us who are just talking ball, of course

Fighter of the Nightman said:

I mean, you are kind of saying what I was trying to say anyway, though you downplayed how I would have said it. My point is that while entirely separate on its own, if you are getting recognition by the AP Poll saying you are a top 10 team, this separate recognition is probably a good sign that you are on track for a top 2 (or at least top 3) seed. I never said there was any causation, my point is that there IS a correlation, so folks are sort of going overboard when they act like the AP Poll is as good as toilet paper simply because it isn't an actual, baked-in ingredient in the selection process.

Click to expand...

Also, the same thing could be said of any ranking service. A couple years ago St. John's was ranked #21 in KenPom. Didn't make the tournament. Because KenPom (which IS on the team sheet) doesn't cause teams to make the tournament or get seeded where they do. A team's performance does.

And teams who perform well are likely to have good metrics, get good results, which lead to good AP rankings, and a good resume.

FWIW, I've adopted a new view of the AP Poll. I think the end "product" of the poll is the final regular season poll before Selection Sunday. The weekly polls we see throughout the season are more or less just the process to get to that final poll. So, at any given point in time throughout the season, the poll might not look right but the pollsters have faith in their process and it's ability to identify the top teams in March.

Take a look at the top 10 of last year's poll before Selection Sunday and how it matched up to the end of year post-tournament KenPom rankings.

Duke (-)
Houston (-)
Florida (-)
Auburn (-)
St. John's (+9)
Tennessee (-1)
Alabama (-1)
MSU (-1)
TTU (-)
Louisville (+18)

So, basically, if you remove schools with a Rick Pitino affiliation, the AP Poll aligned pretty closely to the final KenPom rankings which came out 3 weeks later.

Comes down to your faith that the committee strictly follows their proscribed standards.

The AP Poll is created by humans, and the Committee is made up of humans. Those may be different humans, but I believe there is room for some overlap in their thinking, you see a team highly ranked in the poll all year and some part of your subconscious thinks they should be higher seeded. Or you have similar biases to the pollsters in the first place.

Analogous to me to the lawyer sneaking in a jab that the judge tells the jury to disregard. Can't put the genie back in the bottle.

Now we're getting kind of weird imo

Despite all of the readily available information says they do not use it.... well, they actually do, but it's just subliminal lol

Figured while I have the data, here are the teams that have the biggest gaps between their resume metrics (WAB,SOR) and their efficiency metrics (NET,KP,BT):

- Higher efficiency than resume (conceptually teams that are better than their resume): NC State, Cincinnati, Indiana, TCU, Providence, South Florida, Washington, VCU
- Higher resume than efficiency (conceptually teams that have good wins but haven't overall been a dominant team): Miami (OH), George Mason, Liberty, Murray State, Oklahoma State, Stanford, USC, UCF

As to what the committee prefers? It changes year to year, mostly, because the committee members are different year to year. I will say there are usually one or two high-efficiency teams that get inexplicably left out every year for some odd reason or another, but typically it'll be a team with a good resume and a bad adjusted efficiency.

Last year, the highest KP team that was left out was #44 Northwestern, who had a WAB of #70, below the cut line.
Two years ago, the highest KP team that was left out was #21 St. John's, who had a WAB of #53, still just above the cut line but just barely.
Three years ago, the highest KP team that was left out was #31 North Texas, who had a WAB of #40 and was screwed anyway. Maybe because they were in CUSA and the committee made a mistake (they went on to win the NIT that year also)