Posted on

2021 Running Review

I am quite happy with how my running went in 2021 . As a quick background, I took up running as a pandemic hobby in 2020, albeit from a somewhat athletic starting point. I ran my first marathon (first running race longer than a 5K) in 3:32. I ran over 1,300 mi in total in 2020.

2021: My overall running was down a bit to 960 miles, due to two great life events. First, Kelly & I bought a house which meant a lot of my free time went to projects around the house & yard. Second, Kelly gave birth to our daughter Riley!

Despite this, I still completed my first half marathon race, crushing my goal time with a finish of 1:32:06. I was able to sneak this race in less than a week before Riley was born. Along the way I also PR’d the 10K in my first ever race win.

Overall, I’m most proud of continuing to get back on track to train & race through the two big events that came my way this year. I feel like this was not a quick fad for me and something I will keep up for many years to come. I’m actually starting to enjoy running more and more, whereas in the beginning it was just something I did to stay healthy during the pandemic.

2022: I plan to run 2 marathons and hope to break 3:10 so I can qualify for the 2023 Chicago marathon on time (I hope to break 3 hours in Chicago…). My second marathon of the year will be the Richmond marathon. I am currently targeting the Coastal Delaware Running Festival marathon for the spring but there is still time for my plans to change. I also hope to bump my annual mileage up over 1,500 miles. For my first marathon I will follow the same training plan I did for the 2020 Richmond marathon but for my second I want to follow a longer training plan.

I wanted to break 20 minutes in the 5K, which I did in the RRRC First Day 5K race finishing in 19:55 (previous PR was 20:50), so I am off to a great start!

Posted on

My First Marathon

I ran a marathon! At the end of March, I started running to stay active while my gym was closed. It went better than I expected so I kept running and running. I set a goal of a 2 hour half-marathon. As I neared that goal, I got notions of training for a marathon despite the uncertainty of a marathon even happening. This race was a bit unusual. They have the course open for 2 weeks, I relied on Kelly to be my water stations, and we had to cross some active intersections. But I did it! I set a conservative goal of 8:30/mile pace and an aggressive goal of 8:00 pace. I came dang close to that at 8:05/mile and felt surprisingly good throughout the race. A big thank you to Kelly who not only was my personal aid station today, but spent many mornings solo parenting while I was out running (sometimes for 3 hours…)

After my marathon, I have now gone on 175 runs this year totaling 1285 miles and 200 hours. To train, I followed the 12 week 55-70 mile plan from Advanced Marathoning by Pete Pfitzinger and Scott Douglas. Before that I was following base training plans from Faster Road Racing by Pftizinger and Philip Latter. The marathon training plan was quite time-consuming. I felt pretty sick of running the final few weeks leading up to the race. However I think it put me in a great position to exceed my expectations. While I set a goal pace of 8:00/mile, that only entered my mind on my final tune-up run on the Tuesday of race week. My initial goal at the start of training was four hours or 9:09/mile.

I spent some time yesterday analyzing my race data. I ran miles 6-20 all below my average pace. Before that, I was mainly warming up and getting down to race pace. After that, I was struggling to stay as close to race pace as possible. I was able to push myself at the very end and ran mile 26 at 7:53, twelve seconds faster than my overall pace.

A handful of people have already asked me if I want to do another marathon. There was a stretch during the race (I’d say miles 18-24) where I thought “ok I can check this off my bucket list, no way am I ever doing this again”. But I’m actually leaning towards doing more marathons right now. My training plan includes a five week recovery plan which will basically take me through the end of 2020. At that point I’ll actually have to decide if I want to start training for another marathon. Right now I’m considering attempting to qualify for the Chicago marathon, which would require a time of 3 hours and 10 minutes. A 22 minute improvement sounds difficult but plausible to me.

Posted on

Election Thoughts 2020

These are my initial thoughts the day after the 2020 election. You’ll notice most of them are about my understanding of what is going on, rather than what I wish happened and how it compares to those wishes. My wishes for an election aren’t remotely within the realm of possibility and I don’t treat them as such. I generally treat elections as data points to calibrate my understanding of the world around me.

After the 2016 election, I believed that Trump beat Clinton because Clinton was an especially unpopular candidate. I no longer believe that. In the popular vote, Trump kept his margin about the same and improved in percentage. Based on the 2020 results, I believe Trump’s appeal is more popular than I have realized.

Most striking is how far off the results appear from the polls. I think there are two explanations (both could be true): A significant segment of Trump voters lie to pollsters; pollsters have a hard time reaching all segments of Trump voters. A recent Gallup poll found 56% of registered voters said they are better off now than they were four years ago. This was hard to square with poor poll numbers for Trump. There are a few percentage points of people who support the job Trump is doing but won’t admit to voting for him.

The polling gap troubles me. I do not believe the claim that polls are presented as propaganda to make leftist positions and politicians seem artificially popular. There are certainly pollsters who do this. On the whole, pollsters were so embarrassed by 2016 they sought to remedy their errors. Clearly, this is still a problem. If we can’t get roughly accurate responses on how the country weighs in on politicians and policies, I think we are in for trouble. In a country of over 300 million citizens, no individual can get an accurate read on public opinion through personal experience. If we can’t count of polls to give us better guidance, we’re fumbling in the dark.

Looking at crosstabs, you have to notice Trump’s improvement with minorities. According to the New York Times, Trump did 11.5 points better in majority Hispanic districts in FL, GA, and NC compared to 2016. This wasn’t a total shock to me, I expected Trump to improve with Hispanics about 10 points, but this goes against the common narrative. I think we need to be cautious with saying this definitively right now. Exit polls are extra messy this year due to heavy early voting. There will be months and years of debating why this shift occurred. I don’t have a particular narrative myself, but I’m not sure how to square the anti-BLM protests narrative with a Biden win in Minnesota.

If you are curious, I did not vote for Trump, Biden, or Jo Jorgensen. In fact I did not vote for a single person on my ballot. I did vote, yes on proposal 1 and no on proposal 2. I voted for Gary Johnson in 2016 and I felt I could honestly tell people I voted for the candidate who I thought would make the best President. I was not sure I could say that about Jorgensen. Aside from my policy disagreements with him, I fear Trump’s abuses of power and disregard for the constitution. Aside from my policy disagreements with him, I fear Biden will not be held accountable for any of his abuses of power as shown by the media’s refusal to cover Hunter Biden’s business dealings in China. Ultimately, the one outcome I feared was one party holding the Presidency, House and Senate. I don’t think we will see that.

I placed two bets on PredictIt and one prediction on Twitter. In March I bet Trump would win the Presidency and back in September 2019 I bet the House would be Democratic but the Senate would be Republican. A few days ago I posted these predictions about how the Presidential race would play out. As I write this early Wednesday evening, this looks largely correct with a few mistakes, some of which may be contested in the days to come.

Trump has announced he will request a recount in Wisconsin. I predict he will do the same in Michigan and Pennsylvania. I would not be surprised to see lawsuits in these states as well. There has been a lot of talk about voter fraud today, particularly from Republicans. This is not surprising. Early voting was heavily Democratic, day-of voting was heavily Republican. This meant that while some states reported early voting first and some reported day-of first, most critical states had some kind of significant shift. My take on voter fraud is if you are an election observer and you tell me there is fraud, I will listen to you. If you’re a random person on Twitter reporting anything other than evidence from an election observer, I am likely to ignore you. There are many people with huge incentives to identify and fight any voter fraud against their interests. I will sit back and see what major cases are brought before the courts. And for the record, unless you are an election observer, I don’t think you can saw you definitively know there was no fraud in that locality. Blind dismissals of the notion of fraud is just as lazy as blind allegations that a heavily-Democratic big city must be hiding fraudulent votes.

I think Democrats may regret not only fighting for expanded mail voting but promoting it. Republicans have been criticized for not taking action in legislatures they control to allow quick counting, as is the place in Florida.  I don’t know enough about this area to confidently agree but this seems plausible to me. When Trump tweets out statements like this, I can’t disagree with the notion that Republicans seek to sow doubt on as many mail votes as they can. Even if there is no voter fraud, because of the strong partisan split in early vs day-of voting, you were going to see strong shifts of this manner.

If Trump loses, I think he will be the Republican nominee in 2024. He outperformed everyone’s expectations even if he loses the Presidency. As a party Republicans did better than expected even in the House. I think there will be a strong belief that Trump has electoral magic no other Republican can match, and any concerns about his 2020 performance will be chalked up to covid-19 being a wild card out of his control.

Posted on

Things Break

Kelly and I have lived in our condo for a little over three years now. Things around here have started to break at a consistent pace, likely accelerated by the fact that we are home all the time now during lockdown.

Our laundry has been coming out a little funky. This is not due to lockdown but the fact that Henry (nearing one year old) is getting his clothes covered in food. This forced us to discover some elements of the washer and dryer that are supposed to cleaned out regularly (not sure we’ve done that since we moved in).

Our toilet paper holders have become loose / fallen off. They have an element that needs to be tightened with an odd looking screw head. I purchased a computer repair kit (for another step in this process) which included the required screw head. With that tool, this was an easy fix.

We have two toilets that use Koehler flushless technology. It’s an incredibly stupid device that we hate. They have all kinds of issues: sometimes they run endlessly, sometimes they don’t run at all, and they use up batteries endlessly (I eventually bought rechargeable batteries). I looked into switching to standard flush tech, but that would run us around $150. Not awful, but I’m not sure how long we will live here. I spoke with Koehler and explained that our units have just stopped flushing even after replacing the batteries. The flush units we have have been discontinued but they were kind enough to send us 2 of the new units that have replaced ours. So far they are working alright.

My laptop (a 3 year old Dell) has been running a little slow and Kelly’s (a 6 year old MacBook Pro) has been running very slow. I installed new RAM (Kelly went from 4 GB to 16, I went from 8 GB to 16) and new solid-state hard drives. For Kelly I simply cloned her existing hard drive. For me, I started from scratch installing both Windows 10 and Ubuntu 20.04 in a dual-boot setup. Then I copied over the files I needed from my old hard drive. I would rate these upgrades a strong success. The computers are booting up and running very fast and I love Ubuntu for programming. I was very pleased with how easy it was to copy over my iTunes library including metadata like play counts. When I was pulling my old hard drive out of my computer, the SATA cable came out from my computer. I was petrified. It turned out there was a latch I have pulled out and once I tucked the cable under it (and aligning it up with some tiny pins) and put the latch down, everything was fine.

Our dishwasher is leaking. I did some research and it looks like there are a number of possible causes. It isn’t clear to me how to diagnose the exact cause and most of the fixes look above my pay grade. I think we will be calling a plumber for that one.

In our bedroom, we have noticed some serious paint blemishes. About 18 months ago our condo was the victim of serious damage due to a nearby construction project. Our unit was fixed up & painted including this wall. Based on my research of paint blemishes, it seems the two most likely causes are underlying moisture or improper painting (I.e. painting on a wet surface). I’m working with our property manager to see if we can diagnose the exact issue. If this is the result of poor work by the repair crew, I’d like to get them involved in repairing the poor repair.

Posted on

My Lockdown Experience So Far

Kelly and I have been working from home for over nine weeks now. The IHS office has been closed for eight, which is roughly how long most statewide shelter-in-place / stay-at-home orders have been in effect.

Here’s what a typical day looks like for me:

  • Wake up and spend time with Henry
  • Drop Henry off at daycare
  • Go for a run
  • Work
  • Do yoga & eat lunch or eat lunch & read
  • Work
  • Pick up Henry
  • Spend time with Henry
  • Henry’s bed time
  • Dinner and a TV show with Kelly
  • Read / play a videogame / write some code
  • Bed

In the beginning I was largely goofing off. There was a lot more videogaming and a lot less reading or running. After about two weeks I tried to mentally shift from a short-term vacation from reality to a longer-term attempt at normality.

For me, this is more or less working fine. I don’t mean to brag. Kelly & I are lucky to not only be gainfully employed but have a childcare provider still able to operate. I understand there are people whose lives are being absolutely destroyed under lockdown. I want us to return to normal as soon as possible. But given the situation we are in, I’m glad to have built a routine that works for me.

Even if Virginia lifted all restrictions tomorrow, I anticipate following this schedule for months voluntarily. I would not be surprised if I’m not in the IHS office again until 2021.

I’ve been meaning to write more during this period, both to document what it’s like for posterity and because I (in theory) have the time. I plan to write about:

  • My thoughts on the lockdown policies
  • Running
  • What I’ve been reading
Posted on

On COVID-19 and the economic costs

A friend sent me “The debate over ending social distancing to save the economy, explained” by Ezra Klein and asked for my thoughts. Here they are:

I absolutely think people need to be willing to have this discussion weighing the economic costs of essentially shutting off the economy, as uncomfortable as that may make them. In my opinion we should *all* want the economy to restart ASAP. Even if it really needs to be shutdown for a while, it should be as short as possible.

The hardest aspect of this is that the data on the virus is fuzzy. Very very fuzzy. How fatal is it? How many people who get it require a ventilator? I think you want enough social distancing to keep ventilator demand below supply.

There have been a number of arguments recently around the theme of: we have been overestimating the fatality rate. Some are more strongly argued than others.

I think the best way to know this is if we can do antibody testing of random samples to answer the question: Have a lot of people in the US already had it and developed immunity? If so, we can conclude this is actually not that fatal and we can end social distancing pretty soon.

Initial estimates from the WHO were that the fatality rate is around 3-4%. That is CRAZY high. The flu is like 0.1%.
For a while my best guess has been 1% but I’m just a guy on the internet. At the start of March my best estimate was 500k Americans dead. Really not sure how far down to adjust that based on the fact that we’ve gone hardcore social distancing.

I’m not sure when but at some point in the past 10 days I feel like I’ve gone from feeling like society is vastly underrating Covid to feeling like society is slightly overrating it.

I was definitely in favor of hardcore voluntary social distancing. Statewide shelter-in-place and shutdown of “non-essential businesses”…I don’t know. I know that’s a frustrating answer but it’s really hard to say and I felt even less certain when they really started picking up Monday. And I say this as a libertarian with a very strong skepticism of such harsh government measures.

As of today I’m leaning towards lifting them soon and we just need to worry about the major metro areas, but I don’t feel confident either way. I want that random sample antibody testing to make a determination about how many Americans we expect to require ventilators over the next few weeks.

I’m glad Ezra included the Hammer & Dance strategy, I think it does a good job of taking into account how miserable for society social distancing is.

One big point: Getting the cost/benefit between the possible deaths & the economic damage is basically impossible. Because of exponential growth, there isn’t really a point at which you can say “ok the deaths so far has been low, but we want to stop it at this point”. By that point, it’s already too widespread. Like a wildfire you can’t put out and just keeps spreading. So if you’re going to attempt mitigation, it is going to look premature.

I think the after-debate on this issue is going to be ugly. People are going to believe what they want to believe about how bad the virus really was and how much social distancing mattered and how much economic damage was done. It is going to be really hard to prove these theories one way or the other. We might attempt to compare say California (hard early lockdown) vs Florida (still not doing a shelter in place). But it will be hard and people will hand wave stuff away.

Posted on

Response to “The White House didn’t like my agency’s research. So it sent us to Missouri.”

I was sent “The White House didn’t like my agency’s research. So it sent us to Missouri.” an op-ed in the Washington Post, by someone I respect. Here were my thoughts in reply:

I think this is good (the actions of the Trump administration).

Above all else, this piece reeked of arrogance. If only we could all be so intelligent, so valuable, so noble as the kindly author.

I started working on machine learning tool that could ingest large amounts of data and use it to predict when and what farmers plant. Eventually, it could be the beginnings of a larger framework for integrating farmer behavior and economics into climate change research.

The last thing I want my tax dollars going towards is some data scientist working towards central planning of farming for his climate change agenda. As the CEO of Better Data, my official stance is that anyone who trys to wow you with the term “machine learning” is bullshitting you when what he really means is “overfitting my data set”.

I mean, is this comedy?

Because the publishing staff all left, dozens of reports on subjects from veterans’ diets to organic foods are delayed.

Oh, no! Won’t someone think of the…veterans’ diets studies?! And how can we live without the government studying organic food for us. Especially given their great track record. Can someone remind me if eggs are currently on the government’s “good” or “bad” list this year? How about red meat? It’s so hard to keep up.

I really resent this “oh I’m just a lowly public servant” shtick.

Contrary to the common talking points about cushy government jobs, we all knew that we could have gotten higher salaries in the private sector, faster advancement elsewhere in the government or more perks in academia.

Notice how he writes this. Private sector has higher salaries, but less job security; faster advancement elsewhere in government but less fulfilling work; more perks in academia, but have to deal with committees and grading. These are all trade-offs. The author and his colleagues have chosen to work at ERS instead of other opportunities, presumably because it was the best option available for each of them. You’re not acting out of charity, guy, this is your job. And one of the downsides of government work of course is your big boss can change every 4 years. And sometimes they’re brought into the job on a mandate to drain the swamp.

I laughed at the contrast of these two

Perdue insinuated that our research was politically motivated, telling reporters that the department needed to avoid past mistakes, in which it made decisions “based on political science rather than on sound science.”

My colleagues and I organized letter-writing campaigns, contacted our congressional representatives and voted to unionize.

How dare Perdue accuse us of being politically motivated! The author is going going to prove him wrong by…politically organizing. Oh.

Of course Perdue is politically motivated. But so is the author. All of us are. Personally, I’d be pleased if the ERS was shut down. If it’s going to exist, I think it and many other agencies should be relocated outside of DC. The DMV has the highest median income in the country. So long as that is the case, you’re going to have populists running on “drain the swamp / hang the elites” rhetoric. Whether or not those populist candidates like Trump are full of shit, the message is going to sell.

This article doesn’t change my mind on it at all. The author spends the whole article complaining about how he’ll have to relocate, and when he quits he gets a new job…in Philly, not DC.

Posted on

What I’ve Done With Salesforce

I work with Salesforce often. Salesforce is a popular CRM (customer-relationship management) software. In my day job at IHS, we use it to track our fundraising efforts. We also use it to track our programs and opportunities in higher education.

For a period, I focused my role at IHS on standard Salesforce administrative tasks. This includes user access, database structure, and workflow automation.

This work is important. A database is doomed to failure without a wise admin. I strive to go beyond the basics to deliver the best possible outcome. Here are a few examples of how I have do that:

List matching. At IHS, we send staff members to academic conferences to meet potential academic partners. When we have conference registration lists in advance, I run a query of our Salesforce database to see who we already know. This allows our team to reach out in advance, leading to maximizing the value of their time spent attending the conference.

Automated metrics. When I joined IHS, we were paying a vendor to track our social media performance metrics. This vendor logged into our accounts and entered this information into a form which stored the data in Salesforce. I built a tool that does this automatically by connecting to the relevant APIs (Google Analytics, Youtube, Facebook, Twitter, etc.) This saved the organization money on the vendor and ensured the data was in Salesforce in real-time.

Connected Salesforce to WordPress and Workplace. This meant that when a client clicked a button that fit their existing Salesforce workflow, accounts were automatically created on their WordPress website and Workplace by Facebook platform. This automation reduces manhours spent on creating and deleting these accounts.

Pushed Pardot email metrics to Salesforce. Pardot is a marketing automation tool sold by Salesforce inc for integration with the Salesforce database. I’m not sure why, but there is no out-of-the-box way to see Pardot email performance metrics within Salesforce. I built an integration automation that accomplishes this by creating Salesforce records with performance metrics for each Pardot email.

Built Pardot lists using custom Salesforce fields. Pardot’s main selling point is Salesforce integration, but we found at IHS that we could not create lists based on a custom field for the Campaign Member object. I figured out how to do this with a Python script, which I wrote about here. This saved our staff time they were spending manually exporting lists from Salesforce and uploading into Pardot.

Built a Youtube integration. Last but not least, I have built a direction integration between Salesforce and Youtube for performance metrics. This allows you to connect your Salesforce database to a Youtube channel, create records for each of your videos, and sync performance metrics. I’m working towards selling this as a Salesforce AppExchange app.

Posted on

Future of Sports Commentary

Last night I tuned in to a great NBA matchup, the Milwaukee Bucks against the Boston Celtics. I watched the broadcast on TNT but my TV was muted. Instead, I had my tablet open as two guys on their couch did commentary for the game.

I think this is the future of sports coverage.

Danny LeRoux and Nate Duncan, broadcasting from their couch

Well maybe not exactly those two guys on their couch. But I think we will see more options for consumers. Last night I tuned in to the “#NBACast” show by Nate Duncan and Danny LeRoux. They also run my favorite NBA podcast Dunc’d On.

There are two categories of NBA games, national broadcasts and local broadcasts. National broadcasts are those on networks like ESPN, TNT, or ABC. There is one neutral commentating crew. Local broadcasts have two separate broadcasts of the game by each team’s crew. League Pass subscribers like me can still tune in to these. 

The announcers for national broadcasts are unbiased and more ‘prestigious’ in their field. As a pretty extreme NBA fan I find their commentary bland. Some announcers like Jeff Van Gundy often go on rants unrelated to the game they are announcing. This actually makes some sense. The average NBA fan follows their team and probably tunes in to one national broadcast a week. That game is their weekly lens into the NBA at large so they actually find some value in Van Gundy complaining about LeBron’s defense during a Toronto / Philadelphia game.

But it drives me crazy when I’m trying to watch a game and the announcers are on a completely unrelated rant. I know I’m not alone. Cue the “#NBACast”. It fits my style because they approach the game from an analytics mindset. They also take viewer questions using the chat feature of their online broadcast. 

Anyone can tune in to this show for free but you can subscribe to their Patreon to financially support them. I have been contributing $7 per month for a few years now. Between their show and podcast, I get well over 21 hours of content a month. I get more than $0.33 per hour of new NBA information. I particularly enjoy their summaries of games I missed (especially West Coast games that happen on weeknights) and their “15 in 60” segments throughout the season where they quickly run through the status of each team.

Duncan and LeRoux certainly aren’t perfect. Lots of people can’t stand their style and even as a big fan I have some complaints with them. But I enjoy having the opportunity to tune in to their commentary of an NBA game instead of listening to Reggie Miller or Mark Jackson. I hope we all get more options for how we consume our content.