Analytics thread

[stable genius]

In the midseason predictions thread I was surprised at the number of people who are picking UNI to win 17 or 16 games. That just seemed crazy to me. UNI is currently rated 69 by kenpom, which is solid and certainly the best in the Valley this year by a decent margin, but not really close to as high as several valley frontrunners have been in recent years.

I did some kenpom research to analyze the question of whether or not it is reasonable to expect #69 UNI to win 16, 17, 18 games. There are a couple different ways to look at it.

First, a couple general statements that should be considered:
1) Right now UNI is #69 but they are trending up, we're only at the halfway point and the dust is still settling. Just because they're not in the top 15ish right now (what I'd expect from a team picked to win 17ish games), that doesn't mean they won't end up there or close to it. Loyola 2018 started the year at around 100, was still at 93 at the beginning of MVC play, ended the regular season at 45, and ended the NCAA tourney at 30).
2) The valley is not as strong as it has been in years past, so its better to be #69 than it would have been in 2007. I've tried to account for this below.
3) All this data was from the kenpom era, 2002-2019

Anyways, here's what I found:

Scenario 1, I looked at the teams that finished with 18, 17, 16, etc. wins and looked up what their average Kenpom was. Here is what I found.

---18 wins: 1 team, avg kenpom was 6 (WSU 2014)
---17 wins: 4 teams, avg kenpom was 28
---16 wins: 4 teams, avg kenpom was 24
---15 wins: 7 teams, avg kenpom was 30
---14 wins: 6 teams, avg kenpom was 52
---13 wins: 2 teams, avg kenpom was 43
---12 wins: 18 teams, avg kenpom was 79

Scenario/question 2: Look at the other teams with a Kenpom around #69 (specifically, the 59-79 range). How did they do in valley play?


18 teams, from 2002-2019, had a kenpom in the range of 59-69
---zero went 18-0, 17-1, or 16-2
---one went 15-3
---one went 14-4
---one went 13-5
---four went 12-6
---four went 11-7
---two went 10-8
---two went 9-9
---three went 8-10
---zero finished with fewer than 8 wins.

Based on this figure, it looks like UNI has an uphill climb to get to the 15 or so wins it will need for at large consideration. My opinion. However, The valley is not as strong as it has been in the past. Being #69 in 2019 is a lot different than being #69 in 2007.

Scenario/Question 3 - This is basically the same as the senario above, but this tine I removed the teams from the list that played in the conference when the conference was obviously [somewhat/significantly] better. So here, we're just looking at years where the valley was about as good as it is in 2019.

Under this criteria there were 9 teams. Here is how they did:

---zero went 18-0, 17-1, or 16-2
---one went 15-3
---one went 14-4
---zero went 13-5
---three went 12-6
---two went 11-7
---zero won less than 11

So, the picture here is a little better for UNI. But 12-6 is still looking a lot more likely than 17-1.

Scenario/question 4 - trying to look a this in a more optimistic way as a UNI fan, I would say that according to kenpom, UNI is better than all of the other teams by a decent margin. UNI is currently 37 spots higher than the next closest team (Loyola, #106).

Were there other years when one valley team was 30+ spots higher than everyone else, and if so, how did they do?

Answer: yes. Here they are:
2018 Loyola (#31). 15-3
2017 WSU (#8). 17-1
2016 WSU (#13). 16-2
2014 WSU (#6). 18-0
2011 WSU (#27). 14-4
2010 UNI (#27). 15-3
2008 Drake (#17). 15-3

So if you're looking for the rosiest outlook to make the case for UNI winning 15+ games they'll need for a shot at an at-large, here is where you'd start. I would still argue that all of these teams are in the top 30, and UNI is currently #69. There is still a large gap between a top 30 team and a top 75 team IMO. But this year may be a good opportunity to find out what happens when a good, but maybe not great (again, UNIs kenpom is still trending up for the time being) team is head and shoulders above the rest.

Lots of people will say that numbers aren't everything and kenpom doesn't know everything. That opinion is fine and they are not wrong. But I'm basically convinced at this point that kenpom is an incredible metric and tells you pretty much everything you need to know about how good a college basketball team is.

[stable genius]

Here is the underlying MVC data that I used:

Year/kenpom rank of conference/number of top 100 kenpom teams

2020 (current)...11.....1 (Go Loyola, Bradley, and the Trees!)
2019.....15.....0
2018.....9......1
2017.....10.....3
2016.....11.....3
2015.....12.....3
2014.....11.....2
2013.....9......6
2012.....9......5
2011.....11.....3
2010.....9......4
2009.....10.....3
2008.....7......6
2007.....7......8
2006.....7......6
2005.....7......5
2004.....11....5
2003.....15.....3
2002.....13.....2

[stable genius]

The goal here was not to create a thread about UNI, so here is the kenpom data for MVC teams at this point>

UNI #70 (fallen since I posted earlier)
Loyola #106
Bradley #115
Indy St #122
MSU #150
Drake #152
Valpo #165
SIU #191
UE #200
ILSU #222

[MissouriValleyUnite]

NET
9 - WVU
22 - UNI
23 - Colorado

How accurate is NET?
Only 1 year sample size but in the 2019 NCAA Tournament:
NET went 43-9 thru 2 rounds with a perfect 16-0 in Round 2.

I’m leaning towards NET’s assessment of UNI, who looks the part of a 2nd round NCAA Tournament team. They’ve proven the ability to defeat a top 25 team in their house. They led for 90% of the game against a Sweet 16-ish squad. They’ve taken care of business for the balance of their non-con, demonstrating consistency.

UNI checks off so many boxes. A threat to take over any game (Green). A 55% 3PT shooter (Berhow). A lockdown defender (Brown). Efficient scoring down low (Phyfe 65%/Dahl 74%). 3 guys off the bench who add value (Dahl 74% shooting / Pickford rebounding / Kimmons speed). The team plays cohesive, the players understand and excel at their roles.

You’re in for a special year in Frostbite Falls.

[VUGrad1314]

As a fan of not only a Valley team but of the MVC as a whole I hope you're right. Go UNI (This year anyway)!

[just ducky]

I'll put the updated MVC team 3 point shooting numbers here for convenience. Our numbers are impressive.

ISUb. #5. 40.7%
UNI. #35. 37.7
Bradley #71
Drake. #78
Loyola #82
ISUr. #92
VU #112
MSU. #120
Evansville #131
SIU #189 33%

This really shows how dependent the Valley is on good outside shooting.

Valpo finished the 2018-19 season almost dead last in Div 1 (349) at just above 27%. What a difference a year can make and they are still without their best outside shooter (Fazekas).

[PantherSigEp]

Be interesting to know what UNI's #'s would be if AJ had been consistent from game 1. Seemed like he was shooting under 20% until the South Carolina game. I think he's been hitting at a ridiculous 47%+ clip since UNI lost to West Virginia.

In fairness to him, he's had people keying on him and in his face all season long and almost never gets a shot at an open look. Makes his recent shooting performance all the more impessive.

[VUGrad1314]

We've got a shot at having 3 or possibly 4 KenPom top 100 teams this year.

UNI sits at 80

Loyola 105

Bradley 113

ISUb 117

I'm confident that those first two will make it in. ISUb needs some wins.

[VUGrad1314]

UNI was so unfairly taken down after that loss in Normal. Net is 61

ISUb 93

Bradley 109

Loyola 131

Valpo 143

Missouri St 144

Drake 145

I'd love to see ISUb and Bradley get into Q1 status and have Loyola Valpo Missouri State and Drake become Q2. That would be very good for the conference and its eventual tournament representative. If it's possible for Loyola to be a Q1 game as well so much the better.

[squirrel]

This isn't an analytics post so much as it is a trends post, but Bradley is 13-4 over their last 17 games against league competition, including Arch Madness last year. Hosting SIU at home this weekend, they can push that to 14-4.