# Category Archives: Uncategorized

## Engaging the Media: Telling Our Operations Research StORies to the Public

My article “Engaging the Media: Telling Our Operations Research Stories to the Public” has been published in SN Operations Research Forum. I wrote this article after appearing in the news many times, especially after the SARS-CoV-2 pandemic began (see my media appearances here). The paper’s abstract is as follows:

Academic research occasionally captures the attention of the media. When this happens, there is a small window of opportunity to disseminate the real-world impact of our research and the value of our operations research and analytics expertise to the public. To do so, we must package our messages for public consumption. In this article, I summarize principles for interacting with the media, describe what various media interactions are like, and offer tips for capitalizing on one’s expertise. Finally, I reflect on what we have to offer to journalists and the value of telling our stories to the public.

I feel strongly that applied operations research should be outward facing to some level. Otherwise, the research has little chance to make an impact beyond the discipline. I wrote a brief statement about my outreach philosophy at the end of my research statement in my promotion dossier:

My research formulates new operations research models and algorithms for solving important and interesting real-world problems of national interest and concern. I take this responsibility to serve my profession and our nation quite seriously. I believe that it is essential for researchers who are working on problems in the public sector to disseminate their research findings to the public through outreach in addition to dissemination in academic journals. Translating research concepts into practical messages is critical for influencing public policy and transitioning research concepts into practice. This is a common theme that permeates all my research activities.

## Punk Rock OR zoom backgrounds

I made a series of Punk Rock Operations Research zoom backgrounds for upcoming fall meetings and classes held on video conferencing software. I’ve been using the black background, and I wanted some variety for the many upcoming meetings. The images are below. If you want to use one, you can right click on the image and download it.

Punk Rock OR T-shirts, laptop stickers, water bottles and mugs are available here.

## Opening the economy is not the problem — opening without a plan to control the risk is the problem

Three weeks ago, I wrote an op-ed on government plans to contain the COVID-19 pandemic. I was underwhelmed and felt that many were asking the wrong questions. I wrote an op-ed that was published on Fox News today: You can read it here:

Opening the economy is not the problem — opening without a plan to control the risk is the problem

I outline three risk-management strategies that may be the difference between recovery and a second wave:

1. Regulation of gatherings with super-spreader potential
2. Face masks required in businesses, in buildings, and on public transit
3. Robust contact tracing

It’s up to the rest of us to do our part.

Right before the op-ed was published, I tweeted a thread that reflects my more recent musings on the subject of containment and risk management:

## a soccer win probability model

Last year, I tweeted about a win probability model I created for soccer (or football, depending on where you are from) and the 2019 FIFA Women’s World Cup case study. I promised to blog about this case study I developed for my probability models course. This is a long overdue blog post on this topic.

Here is a portion of the soccer analytics case study.

### Soccer win probability model

Soccer (or football, which it is called outside of the the United States) is based on a 90 minute match. The data from FIFA women’s soccer indicate:

• Home teams score 2.34 goals/match (in regulation)
• Visiting away teams score 1.71 goals/match (in regulation)

We assume the goals are scored according to a Poisson process with exponentially distributed arrival times. Assume each team scores independently of one another and of the score. We can use the home team data for the team that is favored in the match.

Consider the situation when the home team is down by 1 goal with 4 minutes in regulation. Find the probability that the home team wins. This is a win probability. We do not consider pulling the goalie.

In one possible solution, we divide the match into small increments of, say, a half minute in length. We recursively solve for the home team’s win probability for any score differential and any time. This way, we want to be able to answer questions like this repeatedly for different score differentials and lengths of remaining time.

#### The derivation is next. You can skip the math if you want and jump to the figures below.

Let the random variable $W_i(d) =1$ if the home team wins with a score differential of $d$ with $i$ increments to go. Otherwise, the team loses.

We want to find $P(W_i(d)=1)$, the probability that the home team wins if there is a score differential of $d$ with $i$ increments to go. To keep the math simple, we simplify this to $P(W_i(d))$. In our problem, the home team is down by 1 goal with 4 minutes in regulation, yielding $P(W_8(-1))$ for a win probability with a score differential of -1 with eight 30 second increments to go.

We use a recursive expression to compute the probability of scoring in small intervals of time and put the solution in a spreadsheet. The spreadsheet approach computes our solution, and it allows us to assess a variety of situations, including different score differentials with different amounts of time to go. The boundary conditions with 0 time increments to go are $P(W_0(d))=1$ if $d > 0$ (the home team is winning when time expires), $P(W_0(d))=0$ if $d < 0$ (the home team is losing when time expires), or $P(W_0(d))=1/2$ if $d = 0$ (the match ends in a tie).

There are two ways to compute the probability that a team scores in a small amount of time $\delta$, which is the length of an increment: (1) we can use exponential interarrival times, or (2) we can use the Binomial approximation. I’ll illustrate the latter approach below. We have to make the time increments small enough such that having at most one goal scored during the time interval is a reasonable assumption.

We compute $P(W_i(d))$ by conditioning on 𝑌, where 𝑌 captures what happened in increment 𝑖 with a home goal (+1), an away goal (-1), or no goals (0):

$P(W_i(d)) = \sum_y P(W_i(d) | Y=y) P(Y=y)$

After taking advantage of independent increments, we can simplify this to $P(W_i(d)) = \sum_y P(W_{i-1}(d + y) ) P(Y=y)$. Here, we recursively solve for $P(W_i(d))$ by conditioning on what happened last and formulating a new expression based on the win probability with $i-1$ time increments to go.

Here, $P(Y=1) = \lambda_H \delta / 90$, where $\lambda_H =2.34$ home goals per match. Likewise, $P(Y=0) = \lambda_A \delta / 90$, where $\lambda_A =1.71$ visiting team goals per match.

#### Let’s look at the answer

We can put this into the spreadsheet and estimate the probability of 0.048 that the home team wins when down by 1 goal with 4 minutes to go. We can see answers to other scenarios. A home team wins with a probability of 0.517 if the match is tied with 5 minutes to go.

What is more interesting is that we can move across these spreadsheet to estimate real-time win probabilities. Every time there is a goal, we jump to another row in the spreadsheet. We jump up a row if the visiting team scores and down a row if the home team scores.

I made two win probability charts for the USA v ENG and USA v FRA games in the 2019 World Cup. I set the USA Women’s National Team as the “home” team since they were favored to win in each of the matches, even though the matches were played in France. You can also see that the home team has a probability of 0.61 of winning when the game begins.

Earlier we assumed that goals are scored according to a Poisson Process. Is that a good assumption? Not exactly (see this post using Premier League data) but it’s not a bad approximation except for the end of game situation when a team pulls their goalie. The model we built above can be easily changed to have time-specific scoring rates. Pulling a goalie is trickier but doable. When pulling a goalie, we have to consider new Poisson scoring rates that depend on time and the score differential.

On a side note, the Poisson process assumption holds up better with National Hockey League data.

## A Conversation about COVID-19 with Economists, Sociologists, Statisticians, and Operations Researchers

David Banks organized a conversation with a few researchers about the COVID-19 pandemic that turned into a publication in the Harvard Data Science Review [Link]. I was one of the experts included in this conversation. A description of the article and a link to the full article is below.

A Conversation about COVID-19 with Economists, Sociologists, Statisticians, and Operations Researchers by David Banks, Laura Albert, Jonathan Caulkins, Sylvia Frühwirth-Schnatter, Fiona Greig, Adrian Raftery, and Duncan Thomas

Summary. The participants in this discussion are leaders in areas of economics, sociology, statistics, and operations management that are relevant to the COVID-19 pandemic. They felt that the economic toll of the pandemic would be large, adversely affecting the U.S. and the world for years to come. Even if an effective vaccine is found quickly, some changes (e.g., telemedicine, increased working from home) are likely to be permanent. Some portions of commerce can be safely restarted now, with appropriate social distancing protocols, but other portions (concerts, theaters) cannot. Social distancing is an effective tool for reducing the reproduction number of the disease, and small changes in that figure have nonlinear impacts on the scale of the problem.  It is important to get better estimates for epidemiological modeling, and for this we need disease and antibody testing on representative samples of the population. We also need to re-engineer work and other activities in ways that allow the economy to begin to slide back towards normal operation without increasing the reproduction number.  As restrictions loosen, there will patchwork outbreaks throughout 2020.  Universities will face challenges to their traditional ways of doing things, but will also have opportunities to improve what they do.

Related posts:

## Planning a sabbatical from the beginning

Several professors asked me for sabbatical planning advice and how to get started. I’m not sure what the best way to plan for a sabbatical is, but I am more than happy to share my experience along with lessons learned from my sabbatical at RWTH Aachen in Germany. Your mileage may vary.

Two years ahead of time, I started thinking about where I wanted to go. In the back of my mind, I always wanted to have an extended stay in Germany. I speak German and even earned a minor in German in college, when I studied abroad in Darmstadt, Germany for a summer. Additionally, my children could take German as a foreign language in school before and after the trip, and that was the deciding factor. I also considered the Netherlands after giving a couple of talks there and meeting many excellent faculty members at Dutch universities. Germany borders the Netherlands, so collaborations across the border would be doable.

My colleague Jim Luedtke was the source of inspiration. He took an academic year sabbatical to Chile with his partner and three children, two of whom attended Chilean schools. He was positive and adventurous, and he gave me the confidence to take my family with me on a sabbatical. Others I talked to were overwhelmingly positive about their experiences overseas with a family.

I looked into applying for a Fulbright scholar award early on. The application requirements are country specific. I used that the Fulbright guidelines as a template for overall sabbatical planning. Jim Luedtke lent me his sabbatical application for Chile, which was helpful in understanding what to include and what level to write about my proposed Fulbright project. Fulbright scholar awards in general have a maximum length of four months, which could be wrapped into a longer stay overseas. This would allow me to do a semester long or academic year sabbatical.

Fulbright applications are due in August a year before the trip starts. The spring before the Fulbright was due, I contacted Marco Luebbecke at RWTH Aachen about the possibility of a sabbatical. He said yes and asked me if I would be willing to teach a course in the summer semester (April – July). At this point, it became clear that an academic year sabbatical would be too long, because I would stay for the academic year and summer. This helped me narrow my scope for the sabbatical timeline, and I committed to a Spring semester sabbatical.

The Fulbright application was due in August 2018, and my university sabbatical application for the 2019-20 academic year was due a month later in September 2018. I could work on both applications at once. I proposed a sabbatical to start in January 2020, almost a year and a half before leaving for Germany. A few months later after I submitted my Fulbright application, I was delighted to learn that I was selected for a Fulbright award and that my university sabbatical was approved.

Living in Germany for more than three months requires becoming a resident and embracing German bureaucracy. U.S. citizens who visit Germany can stay on a tourist visa for 90 days. A longer stay requires applying for a residence permit, registering with immigration, and other requirements. This meant that I had to have a lease (not have an extended stay with Air BnB), arrange for health insurance that is required by law, open a bank, and have all sorts of documents for my children.

The Department of State and the Fulbright Commission were fantastic. The Fulbright Commission helped me with health insurance and gave me checklists and verification letters for all the requirements. I cannot say enough about the support I was given. They helped me acquire necessary international health insurance (required for residence in Germany) and helped me understand the residency requirements in Germany. They also provided a letter that I needed to register as a resident, register with immigration, and open a bank account. And importantly, they were extremely helpful during my sudden return to the United States in March 2020 during the COVID-19 pandemic.

~

Read more blog posts about my 2020 sabbatical here and see my tweets using the “PunkRockORinGermany” hashtag on twitter. I have posts about my sabbatical plan and how I prepared for a semester overseas. I had to end my trip to Germany early due to the COVID-19 outbreak.

## articles about re-engineering society post COVID-19

Here is what I’ve been reading about how COVID-19 will forever alter the world in which we live.

To start, the Centers for Disease Control has mitigation resources for re-engineering communities, schools, workplaces, and events (“Guidance for Where You Live, Work, Learn, Pray, and Play”). It is worth visiting. They have guidance for many institutions, including:

Re-engineering our facilities

1. How ten cities throughout the world adapted public transit in the wake of COVID-19
2. Possible improvements to airplanes to reduce virus transmission
3. How one theater company is bucking the trend by not canceling plays and implementing a slew of precautions, including removing 70% of seats and having no intermissions. I hope the arts can adapt, because I would love to see a play and the symphony in the future.
4. Citylab: how the United States public transit can survive the coronavirus has
5. Why are meat processing facilities so vulnerable to COVID-19 outbreaks? “At least 9,400 positive cases have been tied to meatpacking plants across 28 states.” The work is labor intensive in crowded, poorly-ventilated environments where workers may encounter 1000 other workers per day. From the Post Crescent.

Re-engineering our processes

1. NPR: 9 ways K-12 schools will be different when the reopen, including staggered schedules, smaller classes, and new calendars.
2. In an op-ed, Drs. Sheldon Jacobson and Janet Jokela suggest ways big land-grant research universities could slowly and intentionally reopen while mitigating risks to faculty, staff and students/.
3. It will not be business as usual for universities in the coming academic year. The Chronicle has a list of college plans for the fall.
4. German barbers and hairdressers can reopen with new, strict requirements, including styling chairs at least 1.5 meters from each other; hair washing before a cut is required to help kill off possible viruses; no face to face services; and no magazines or free coffee in the waiting areas.
5. Runners World covers running during a pandemic.

Re-engineering our social groups

1. New York City requires wearing face masks in public. NYC is distributing face masks in public parks. And Deutsche Welle explains the face mask laws in Germany.
2. Slate’s parenting podcast “Mom and Dad are Fighting” discussed “quarantine pacts” where families can group together to manage care and work responsibilities as a method to manage risk in a time of quarantine living.
3. The Nation discusses many of the challenges for non-nuclear families introduced by social distancing and quarantines that also hints at what solutions should encompass.
4. Child care allows every other industry to work, but it’s not clear what child care during the pandemic will look like. Citylab has a roundup of coronavirus child care related issues
5. Workplaces need to adjust to new leave laws and family leave obligations and homeschooling realities. The New York Times reports on how mothers are managing homeschooling with work, and the Washington Post note that there are many lawsuits over family leave and discrimination

Related posts:

## articles about forecasting and the Covid-19 pandemic

1. Nature has a nice article on forecasting: The simulations driving the world’s response to COVID-19
2. The Kaiser Family Foundation has a nice introduction to forecasting models for epidemiology: COVID-19 Models: Can They Tell Us What We Want to Know?
3. Fotios Petropoulos and Spyros Makridakis published an article in PLOS ONE entitled Forecasting the novel coronavirus COVID-19
4. FiveThirtyEight on Why it’s so freaking hard to make a good COVID-19 model
5. Finally, science writer Ed Yong has written a few articles about COVID-19 for The Atlantic that focus on the future, including How the pandemic will end based on what we know from other related viruses,

Related posts:

## articles about supply chains and Covid-19

Here is what I’ve been reading about supply chains in the COVID19 era:

1. Vox on food and meat supply chains
2. Why there are toilet paper shortages
3. The Society for Risk Analysis recorded a webinar called “COVID Conversations on Risk” (Ep 3) about food security
4. The Washington Post on why meat supply chains are vulnerable to Covid-19 outbreaks: they are massive facilities with thousands of workers in close proximity
5. Why you can expect shortages of pharmeceuticals but not toilet paper (nice job, Anna Nagurney)
6. On personal protective equipment (PPE) supply chains
1. CDC guidance for optimizing supply chains
2. The Washington Post reports on a factory that produces a chemical used in PPE
7. The modern supply chain is snapping from The Atlantic. Note this came out a month ago.
8. We need a stress test for critical supply chains (nice job David and Edith Simchi-Levi)
9. And finally, Belgians are urged to eat French fries at least twice a week as changes in consumption during the pandemic lead to a massive potato surplus

Related posts:

## how to organize your Tupperware cabinet

Rumor has it that many people are taking advantage of the extra time at home during the COVID-19 pandemic stay at home orders to organize their food container/Tupperware cabinet. It’s such a popular pandemic activity that The Onion published a satirical article about it.

My food container cabinet is always organized and optimized. I took a picture below. The secret to organizing my cabinet is to formulate it as an integer programming model and then solving it to optimality. I have my kids resolve the integer programming model every time they unload the dishwasher and put the dishes away.

I’m kidding.

I organize my cabinet with a heuristic. I nest my food containers in a few stacks that fit in the cabinets. Once I organize the cabinet, there are a few food containers that do not nest with the nicely organized stacks. The most important part of this process is to remove the few food containers that do not nest with the other containers (see the bottom picture). A few problematic food containers can lead to chaos and disorder. I do not throw these out. Instead, I use them to store items throughout my house (e.g., in my toiletries cabinet and tool cabinet).

I have not purchased food containers in awhile. When I do, I purchase them based on their nestedness properties.

How do you organize your food container cabinet?

The key is to remove the pieces that don’t nest.

Related posts: