I'm a big fan of the blog Priceonomics and I posted four years ago about one of their data visualizations where they look at how diverse major U.S. cities are using a metric that was new to me at the time, the Herfindahl–Hirschman Index or HHI. HHI looks at how evenly groups are represented among a market, and it's a really simple formula. You just sum the square of each group's market share as a percent, so an evenly distributed market would have a value of 1/(number of groups) and a complete monopoly by one group would have a value of 1. 

In the case a market has four groups, a perfectly diverse market would be: 

(0.25)^2+(0.25)^2+(0.25)^2+(0.25)^2 = 0.25

and a complete monopoly by any one group would be:

(1.00)^2+(0)^2+(0)^2+(0)^2 = 1

Since I've been playing around with College Scorecard and IPEDS (Integrated Postsecondary Education Data System) college data recently, I thought it'd be cool to apply the same HHI method to each state's flagship university. The thinking is each flagship university is likely one of the most diverse colleges in that state, and is probably a good reflection of that state's diversity as well. So if I could find a HHI score for each state and each flagship university, I should not only be able to compare flagships to one another to see which are the most/least diverse, but also should be able to compare which states' flagship universities are more/less diverse than that of the state's population. Lastly I should be able to run a simple linear regression to see how good a prediction of a flagship's diversity just its state's diversity is.

**Disclaimer: I normalized the market share of each racial group across these five groups: White, Black, Hispanic, Asian, and Native American/Pacific Islander. This means I am not considering any student that identified as: two or more races, non-resident alien, or unknown. I realize this is affecting the accuracy of the data -- especially not including two or more race responses -- but it was the cleanest way to compare the most prominent groups. Another note is that this research has already been done by others, including this fantastic interactive piece by Ben Myers in 2016 and this Hechinger report from earlier in 2018**

Let's start with the rankings of each state's flagship university, from most diverse to least diverse (perfectly diverse among 5 groups = 0.2, complete monopoly by any one group = 1):

Generally, the greater the population of White students the less diverse that university is; this isn't a surprise. What I did find surprising is just how White some schools (and their respective states) are. Seeing it as a percent is one thing, visualizing that on a bar graph really puts feeling behind that number. This is why I love good charts or data visualizations: they pack a feeling of scale or proportion that numbers can just never achieve. [Again keep in mind that I normalized the actual percentages across the five groups, so each group's actual percentage is smaller than it appears -- however their proportions to each other remained the same]

Another cool thing with the above graph is it visually identifies "sister schools", or schools where their populations are roughly the same proportions (via the scientific "squint and see which colors are about equal" method). Theoretically, you could Truman Show-someone: switch every other student at their school with the other school's population (save for a few close friends or familiar faces) and that student would never know the difference. However, that student would notice the difference if the student population of Rutgers (NJ) was replaced with that of New Hampshire. Completely useless, but a fun thought experiment. 

Looking at this data leads to a whole other discussion about how we stereotype certain schools based on their populations; for instance, I'm not at all surprised that UNC and UVA have similar populations since they're bordering states, but I am surprised that UCONN has a very similar population mix.

Another case: it makes sense that ND and SD's flagships have nearly identical populations or that VT/NH/ME all are very similar, but ND/SD also are extremely similar to VT/NH/ME.

Last one of these, but even though they are on opposite coasts of the U.S., the University of Florida and the University of Arizona have very similar populations.

Most of the time these schools have very close HHI scores (meaning they are similarly diverse, sorry for pointing out the obvious) but interestingly there are cases when another school will slip in between with very different proportions. This is interesting since technically that school that in between (B) is closer in diversity to the top school (A) than the bottom school (C) yet the top and bottom (A & C) schools have much more similar student populations. For instance, the universities of Delaware and Kansas have very similar populations, but technically Minnesota and Mississippi are more similar in diversity to Delaware than Kansas even though Kansas' population is more similar.

I think there could be an economic principle buried in this closer-but-not-as-close oddity and the Truman Show-style thought experiment above but that's for another time.

Anyways... now let's look at how each flagship university's diversity compares with their corresponding state's diversity. The schools at the top of the diagram below are MORE diverse than their respective states, the ones on bottom are LESS diverse than their state. [Note: State population figures have been normalized across the same five racial groups, again disregarding responses of two or more races or unknown, for comparison purposes]

As you can see, most flagships are not as diverse as their respective state's population. This is somewhat surprising to me, as I would've guess it'd be more of a 50-50 spit for some reason. In general, schools in the South (SEC football conference) need to do a MUCH better job at making their student population more reflective of their state's diversity. And WTF Delaware??

Finally, let's test my theory that a state's diversity is a pretty good predictor of that state's flagship university's diversity. To do so, I'm going to see how correlated the state and flagship's HHI scores are. Plotting them on a scatter plot (with college.hhi depending on state.hhi) and finding the linear trendline and corresponding R-squared should do the trick.

As expected, they're highly correlated, with an R-squared of nearly 0.77. You can essentially read this number as "A state's diversity score explains nearly 77% of that state's flagship university's diversity score". Even if it explained 100%, it doesn't mean that each group's proportion to one another is the same at the flagship as the state, but the overall diversity of the population market is equal. Again, not entirely useful, but interesting nonetheless. 

That's it, a quick look at how diverse each state's flagship university is. I'm attaching a publicly available version of the data I collected on this Google Sheet. I got the school's demographic data from College Scorecard (download link fyi), the state's data from U.S. Census estimates in 2017, and a list of flagship university per state by just Googling it.

Howard Schnellenberger coined the phrase "The State of Miami" to refer to the wealth of talent in South Florida, from Orlando to the Keys basically. The thinking then (and now) was that all that the University of Miami had to concentrate on was recruiting football players from this area and they'd be good enough to field a championship team. And that's right for the most part.

Players from The State of Miami have been some of the greatest to ever play the game; NFL and college Hall of Famers and Pro Bowlers and first round picks. And they're continuing to do it, just not always at Miami [See: Jerry Jeudy]. Players from South FL are just different: they're hungrier, scrappier, and overall angrier not to mention faster and more agile. They play with a special kind of confidence, a swagger if you will, that is undeniable and feeds off of itself. This makes the same players that are some of the fastest and toughest in the world start to play faster and tougher -- they start to bring that extra oomph with every single hit and block -- and they sure let you know about it. Which only feeds into it more and the feedback loop becomes self-reinforcing. [See: Miami's beatdown of ND one year ago or the impact of the Turnover Chain]

But the problem with this style of play is when it tries to be something it's not -- when it tries to be slow, methodical, and precise instead of fast and loose and intuitive -- the self-reinforcing loop switches from positive to negative and the level of play implodes. That's how you can have world-class football players, players that are a lot more talented than opponents they're lining up across from (no offense UVA, BC, and Duke - you guys kicked our butts but with objectively less talented players) just look lifeless and directionless.

This is not to say that Miami is lifeless or directionless at all, but they look like they are. And if you were again objectively looking at the current state of Miami, you have to start with the man at the top.

Coach Mark Richt (CMR) has been around the game of college football longer than I have been alive and he's forgotten more than most fans could ever hope to learn. But fans of the Canes and football alike can still analyze a team with less football and inside information and be more right than those on the inside. And it's not personal; everybody sometimes lose sight of the forest through the trees, it's impossible not to. Here are the good and bad things CMR brings to the Canes as the man in charge.

Good:

Loyalty - CMR is one of the most loyal people in all of college football. He showed it by being at UGA for 15 years, for continuing to support Bulldog causes even as Miami coach (which is great), and for being loyal to assistants and seeing them grow under him. He's also very loyal to offensive schemes which is a good thing as you have a record of a coherent philosophy that you can point to and always refer back to. I mean the man donated a million dollars to his employer and alma mater to build the new indoor practice facility. I call it "The House that Richt Built".

Stability - If you're loyal to teams and people and philosophies, one of the benefits you'll see is stability. People love consistency even if they won't outright admit it. One of my old bosses put it best when he said people don't get mad at poor performance they get mad at inconsistency because if you always perform poorly or are always late, people know what to expect and can react accordingly. I can fire you if I know you'll always be a bad worker or if we're friends and you're always late, I can tell you an earlier time and plan on you showing up late. But if it's when we're inconsistent that people get really mad at. And CMR is an extremely consistent person (except for his flip-flopping of QBs this year although I heard he did that late in his career at UGA). You know what to expect from him, you know he'll always remain calm and promote a stable environment to the team during games -- that's why his outbursts and grabbing an official last year or him telling his whole Bulldog team to celebrate on the field against Florida receive so much attention, because they're out of character for him. So the stability that Richt brings to the Canes is great, especially if he can be at Miami for 15 years as well.

Precision - If you've ever heard CMR go deep into the things that he concentrates on with teaching QBs, it's precision he covets. Always making the right read or decision is paramount to him because he loves consistency. Making a great throw isn't good enough, the ball needs to be perfectly numbers-high and 12 inches from the chest or something. When the offense is always making the right reads and running the right routes and blocking the right people, CMR's offense is unstoppable because of its optionality and answer for everything a defense tries to do to it. There's no answer for it.

Organization - CMR is a very organized person which makes sense since he loves stable systems that are precise. He'd be a fantastic manufacturing floor manager or would have gotten his Six Sigma certification in another life. I still think his greatest contribution to the Miami program (besides the point mentioned below about the defense) was bringing in a SEC-caliber organized recruiting operation. He has been recruiting at a high level for decades and he knows how to run an efficient football recruiting system. Along with Matt Doherty who runs the recruiting department and his tireless assistants, CMR's organization definitely bumped up Miami's recruiting [See: Paradise Camp]

Intelligence - Like I said before, CMR has been around football for a very long time. He's one of the smartest people in the business and has seen it all. The other greatest contribution that CMR has brought to the Canes was the intelligence to adapt Miami's defense to a style of play that it had shown to be successful in the past and that perfectly suited the type of defender that the State of Miami produced -- a fast and aggressive defense (I wrote about this in 2014 arguing for them to move to a 4-2-5 defense to take advantage of South FL players which Miami has done with the Striker position). Manny Diaz was the perfect hire to run Miami's defense and they have shown what removing barriers of forcing them into schemes that didn't fit their talents [See: Al Golden's 3-4 defense] and letting them play fast and intuitive can do for a team. This was absolutely the right call by CMR when he came to Miami and was very smart.

Values - CMR is a man of values. He's talked at length about the importance that strong Christian values have had on his life and has consistently demonstrated this with his and his team's actions in the community. I'm not saying the Canes didn't have values before he arrived, but everyone in the college football world knows and raves about CMR's values.

Family - CMR is a family man and fits the tight family atmosphere of the Miami Hurricanes football program perfectly -- because he is a Miami Hurricane! It's fantastic to have a Cane coaching the Canes, and promoting the family atmosphere of the Canes with former players as well has been a great decision. That's one thing that sets the Canes football program apart from others, it's a family.

(photo via Getty Images)

Now it's time to talk about some bad things that CMR brings to the Canes as the man in charge:

Loyalty - It's no secret that CMR is loyal to a fault when it comes to moving on from his assistants and upgrading. This has hurt him many times in the past at UGA and is hurting him now at Miami

Stability - Sometimes CMR is too stable as opponents know exactly what to expect as he'll stick with the same plays over and over again, something he has been proud to admit in the past. Or he's too stable in that he doesn't use his emotions for good in reacting during the game, as a calm presence isn't always advisable in every situation. Teams need consistent environments to thrive but when an opponent is taking advantage of your stability, it pays to be unstable.

Precision - CMR's offense runs on precision but when the precision isn't there, it self implodes [See: 14, 13, and 12 points against UVA, BC, and Duke]. When CMR has a QB that can make the right reads AND the precise throws, the offense piles on the points [See: Matt Stafford, Kaaya and Rosier's multi-game stretches]. BUT focusing on precision isn't always the best thing to do. I can make the argument that Miami plays it's optimal offense when it is playing fast and loose and intuitive -- think backyard football or the 7-on-7 teams from South FL that dominate national competitions -- similar to how the Miami defense is currently thriving! We have too much speed and agility on the outside to continue to run a slow and methodical and precise offense. This is what fans mean when they say they want CMR to bring in a new playcaller - someone that will bring a pass-happy spread the field offense that takes advantage of speed. There will be more mistakes in this type of offense than in the current precise system, but the advantages will vastly outweigh the mistakes. When you're throwing the ball to speedy receivers  more than 20-30 times a game (which we have seen recently) you're going to have more turnovers but you're also going to have much more explosive plays and touchdowns. And since touchdowns have more expected points than turnovers (because the other team might not score), there's a rational argument for throwing the ball much more with Miami's offensive talent. This includes continuing to take advantage of your dynamic RBs with designed throws to them besides flat routes [See: Washington St., Oklahoma, and the LA Rams].

Family - This goes back to loyalty, but I think any reasonable person would conclude that CMR has hurt this team by having his son as QB coach. We all want the best for our family but if CMR really believes his son would be a QB coach at a similar program of Miami's caliber then he's blinded by nepotism. I understand CMR is the de facto QB coach but something's not working. It would be a slap in the face to every Miami assistant if CMR fired other assistants this off-season but kept his son as QB coach.

(photo via Matthew Emmons)

As you can see, Mark Richt brings a lot of positive attributes to the position of head football coach. But within all of his strengths lie weaknesses, just like in every other human. The current problem for CMR and the Miami football program is the weaknesses are outweighing the strengths right now. Forcing Miami's offense to be something that it is not is taking away some of its greatest advantages. Even Golden was able to recognize the type of offense that thrived with Miami's players when he brought in Jedd Fisch and James Coley and they ran a fast and throw-happy offense. CMR knew this with Kaaya as well but somewhere along the way the focus on precision took over and it has ground the offense to a halt. Being too loyal to bad assistants and family is hurting player recruiting and development.

The last three losses are unfortunately mostly a reflection of poor management decisions of the recent past, in both offensive scheme and player personnel. There's absolutely no excuse to have seen our (lack of) punting performance last year and go into this year without legit competition, that's completely a management decision. And it's literally losing us games due to poor field position, especially when combined with an inconsistent offense. You're giving the defense a horrible starting field position because of management decisions as play caller and special teams recruiting. To continue to defend the punter's practice performance is laughable when we have many games worth of film. Cut losses and move on but continuing to actively hurt your team and not making a plan to fix it in any other industry would be a fireable offence. That's what our punting has become, especially when combined with an offense that is inconsistent and punts frequently!

Overall I still think Mark Richt is the man for the job and will lead the Miami program back to greatness. But this feels like a critical adapt or die moment in his tenure. He can continue to be loyal to his scheme and assistants or he can draw on decades of football intelligence and realize that something needs to change. It's a hard pill to swallow for someone that professed how important running things himself was when he came to Miami. But it feels like the best path to greatness is to highlight his strengths as a program CEO - to be the stable organized force on top and lead these young men with your attention to detail and values. But look at the offense the same way you did with the defense and evaluate the players and talent pool's strengths and play fast and aggressive. Step back and let someone bring new life and energy into the offense. Step back and take an objective look at the current state of Miami.

As I have dived deep into the world of college access research over the last couple of months, I've noticed many researchers appear again and again across various higher ed topics (such as financial aid, college choice, and completion) with co-authors from different universities and across many different academic departments.

I thought it would be cool to see how these researchers are all connected; since even though 2nd or 3rd degree connections didn't directly work together, their ideas and co-author experiences were shared somewhat. And I hoped to identify some great researchers/universities doing some kick-ass work.

Above is that initial network graph of paper co-authors. Ideally I'd do this with citations but co-authors was an easier start. This shows which researchers published papers with one another. The bigger, bolder researchers are more connected across disciplines and topics.

Here's the basic process I used: 

1. copy-paste from Google Scholar into Excel
2. clean the data through formatting formulas (about 3400 papers)
3. narrow down papers by title, keeping only those related to higher ed (about 1000 papers)
4. format co-authors (nodes) to establish separate connections (edges) for each shared paper
5. create a simple Gephi graph (this was my first one)
6. cluster/organize the layout by Force Atlas
7. filter out authors with less than 3 connections (it cleaned it up nicely)
8. size the nodes by betweeness centrality
9. partition and color the nodes by modularity class

From June 22 - September 22:

Smino - Anita

Khalid & Normani - Love Lies (Medasin Remix)

Apre - All Yours

Drake - In My Feelings

Jerry Folk - Sweetness On My Tongue

La Felix - Comes & Goes

Koresma - Turquoise

The Carters - Apeshit

Maribou State - Nervous Tics (Feat. Holly Walker)

Emotional Oranges - Personal

SACRE - Lovesick (Mura Masa Cover)

Channel Tres - Topdown

Bronze Whale - One

Breathe – Are you All Good?

Slenderbodies - Toxic (Britney Spears Cover)

Here are some classroom posters for a high school math class that I created a while ago with the goal of bringing real world examples into the classroom. Feel free to use them if you see fit.