Goosepoop: Year 1

Hi All!

It’s been exactly 41 days since we dropped our last newsletter, our inaugural State Finals Viewing Guide.

So, before we jump in, a bit of an apology from Mothergoose: December got away from me a little bit, and I promise to be more consistent in posting throughout the rest of the off season.

With that out of the way, I want to use this ‘first post back’ to celebrate the milestones of Goosepoop’s first year. We’ve come a long way since our first post on the SEARN forums in August, and I think it’s always a good practice at year end to reflect on where you started, how far you’ve come, and where you’re going next.

In this way, we aim to give Goosepoop some accountability: when we pull up at year end in 2022, you as part of the community will be able to grade us on whether we made progress vs. last year, and, most importantly, did we build what we said we were going to build.

One last thing before we start - make sure you subscribe to our newsletter. As we note below, we’ve got several new products in store for 2022; by subscribing to our newsletter, you’ll make sure you never miss one.

Where We Started

Because 2021 was Goosepoop’s inaugural year, we begin this year’s reflection with the ‘idea’ phase: when did the vision for Goosepoop really start to take shape? 

The first glimmers started to appear after I encountered an Mlive article detailing the switch in playoff point systems at the end of the 2019 season. That article naturally lead me to the MHSAA website, where I read this document detailing the switch in point systems.

Even after reading this several times, it was difficult for me to comprehend how exactly this would play out: would 4-5 teams make it? would 7-2 teams be left out?

You could do some pen and paper models on your own, but, given the need to model out all of your opponent’s wins (and then all of your opponent’s opponents wins), it quickly became clear to me that something with a little more computational power was required. Thus, sometime in the Spring of 2020, I started to build a web app to do this; this was the first version of what would ultimately become the ‘Playoff Wizard’.

Then COVID happened. For most of the Summer, I wasn’t sure whether we would even have a season; as such, progress on building the Playoff Wizard completely stopped. By the time that football did resume, it was already week 4, and it wasn’t clear whether there would be playoffs at all, let alone what playoff point system would be used. This lack of clarity lead me to halt any production until further notice.

Side note: while I didn’t code anything useable in 2020, I did end up making a spreadsheet for simulating week 9 outcomes. This was only possible given the COVID ‘bubble’ that most teams played in (i.e. OK White / Green / Red teams played only themselves in 2020, so modeling out the final playoff points of each team required only simulating 24 teams’ schedules)

What We’ve Done

The Build

Sometime around March of 2021, when it became clear that we would have a normal season AND the new playoff point system would be used, I once again set out to build the Playoff Wizard. I worked all through the Spring and Summer to do this; my first bit of code for the new app was pushed to the cloud on March 9th and my last bit was pushed on July 26th.

Launch Day

By early August, the app was built, and I was ready to release it to the world. The SEARN forums were a natural starting point: having read threads there for years, I had a feeling that the community there would have fun playing around with it.

This initial strategy got the app more traffic than I expected - on the first day, we surged to ~80 users, with roughly that number returning a few days later after another couple of posts on the forums.

Twitter

As you can see from the above graph, after the first few posts, user activity on the Playoff Wizard quickly stalled - in hindsight, this was most likely due to the time of year I was posting (early August vs. October), but, at the time, I wanted to be sure it wasn’t just due to ‘maxing out’ our channel via only marketing to the forums.

So, we launched the Goosepoop Twitter in mid/late August. The thought was that more Michigan HS football fans were on Twitter than coming to the forums, and by having a presence there, more people would learn about the app.

Our initial Twitter marketing strategy was to just tweet some informational descriptions of the new playoff point system. Once the season rolled around, we leveled up this approach by taking data off of the Playoff Wizard and tweeting out how users were picking the outcomes of each game.

This got us some traction - thanks to the Fox 17 guys (Jason Hutton and Zach Harig), we were able to piggyback the Playoff Wizard’s user picks to the release of the Fox 17 sportscaster’s picks each week (we called it ‘Hutton vs. the Crowd’). This competition resulted in a lot of new followers for our Twitter account, which, in turn, transpired into additional Playoff Wizard activity (As an aside, across 8 weeks of competition, Hutton ended up beating the Goosepoop users, so props there, @jhutt5).

Newsletter / Peak Week

Playoff Wizard activity picked up steadily from there - throughout September, we consistently averaged somewhere around ~75 daily users. This increased substantially as playoff selection time neared, peaking with our all time high daily users of ~580 at the end of October.

As this was happening, I was A) having a lot of fun compiling the data each week and B) starting to realize that this fun would stop at some point, given that there would be no need for the Playoff Wizard post playoff selection time and users would stop coming to the website.

This is what compelled me to start the newsletter - I knew that through the data I had gathered for the tool (points for, points against, common opponents, etc.), there were some interesting facts I could continue to share with people long after playoff selection time.

With that in mind, we launched Goose Droppings (what you are reading today) in early October. To attract subscribers, we posted a few of our newsletters on the forum, and we also added a pop up on the Playoff Wizard’s team selection page. Up until the start of the playoffs, this strategy was pretty effective - on the last Saturday of the regular season, we had grown our newsletter subscribers to 160! In the post season, we continued to grow, albeit at a much slower pace - our current subscriber count is just south of 180.

What’s Next

And this is where we sit today - thanks to you guys (and everyone who doesn’t subscribe to Goose Droppings but has visited our site), I can proudly call Year 1 of Goosepoop a success: we have taken something a lot of people have passion for (Michigan HS Football), and added a statistical ‘flavor’ to it. There’s no reason we can’t talk about our corner of the world with as much data & tools as other sports may have.

As we look forward into next year, we want to continue on this mission: how do we continue to add ‘data -centric’ content such that the conversation on Michigan High School Football is healthier than ever?

To accomplish, we’ve got a number of new products we’re considering, the first of which is the debut of our Goosepoop rankings model: come the start of the 2022 season, we will be using an algorithm to rank every single team in every single division.

This algorithm will then feed into our Playoff Wizard app - many of you reached out during the season last year to express your dissatisfaction with the way win probabilities were assigned to various games. To remedy this, we will be updating the Playoff Wizard to use our algorithm; we expect this change will make the app’s forecast much more accurate.

We’re also debating expansion into other mediums to broaden our message: currently, we are just on Twitter and Substack (this is the newsletter platform you’re reading from right now). Do you guys want a weekly ‘video’ version of our newsletter on YouTube? We will be beta testing just such a show sometime in August; stay tuned for that.

Finally, we’d also like to expand our level of interaction with you guys. Outside of using the Playoff Wizard, the current version of Goosepoop is really just me writing at you guys, with minimal feedback besides Twitter DMs or forum replies. How can we shift this? One thought is to expand the number of interactive apps on our website: we’ve got some ideas of what to do next, but we’re not yet ready to share. Stay tuned for later on this year for more updates here.

That’s a wrap on 2021 - thanks again to everyone who made this mission possible. I’m truly thankful to be able to code & write for you guys, and I’m really excited to see where 2022 takes us!

If you liked this post, would you share it on your social media? We greatly appreciate the support!