The false starts of MVP

In 2023, I had added most of the features that I thought were needed in DraftKick, and I started looking toward my next thing. Although it has felt like I was fumbling around aimlessly during the past couple years, I'm feeling optimistic that it has all been leading me here, to MVP.

So let's chronicle my list of failures along the way!

Attempt #1: Fantastic

My first crack at this was in December 2023, with a project called Fantastic.

Using Python and Django, I built a Yahoo-esque fantasy interface. The neat thing was that this was built on a database I had made letting you create a league starting from any point in baseball history. Here are a couple of screenshots showing a league from 1963 and another from 2000.

Leading up to this, I had become borderline obsessed with fantasy UI. For example, I had noticed that every mobile app had coalesced around four primary navigation links (League/Home, Team/Roster, Players, and Match/Matchup), a choice that made it into my own minimalist design.

As you probably realize, nothing ever came of Fantastic. But the idea of creating my own fantasy platform (whether modern or historical sim) kept coming up over the next couple of years. The next iteration (some months later) changed the name from Fantastic to League but it too never got close to a releasable state.

Attempt #2: MVP Mark I

A little more than a year later, I again started digging into figuring out what made a good fantasy interface. One preliminary research project was studying the the history of Yahoo Fantasy designs on the Internet Archive.

This time, I wasn't thinking of building my own platform, but of refining the existing platforms.

My motivation, in hindsight, stemmed partially from my first real Fantrax experience. A longtime home league made the switch from CBS in 2024, and I found myself dreading to open Fantrax to check on it. (Note: Not a recipe for fantasy success.)

Wow! The Fantrax UI is terrible!

I toyed with replacing the Fantrax UI with something more bearable. I came up with a prototype that pulled real Fantrax data and then cleaned things up:

You can see the influence of Attempt #1: This version still has the four primary actions, and it draws from the Yahoo style that in my opinion is the best available.

As you can see from the screenshots, I had named this project MVP. What does that stand for? Nothing, really. ("Manage various platforms"?) I just liked the positive sports connotation and the association with being valuable.

This was in February 2025, and I soon found myself busy with the peak of fantasy baseball draft season. As with many of my projects, this never made it any further.

Attempt #3: MVP Mark II

As draft season wound down, I found myself getting sucked back in to the idea of my own fantasy platform (the idea that had once been called Fantastic and more recently League). We're now in April and May of 2025.

This time, I tried switching out Django for the slimmer Python web-framework Starlette. I also switched the name (again). My Fantrax-UI-replacement project had gone under the moniker MVP, and I still liked the coy ambiguity of that abbreviation. With nothing to show from the earlier project, I figured the MVP name was fair game for this one.

A new idea here was the introduction of position colors, a departure from the earlier Yahoo-based explorations. I think these first screenshots use the Flexoki palette. While I love the palette, it just felt wrong for this project—too dull.

I decided to try a more vibrant palette, with muted tones on the position badges:

At this point, my recurring issue with building league platform idea once again reared its head: The project is just so big. There are so many pieces to try to get in place, I predictably fell off the wagon at my usual spot.

What have we learned?

You probably won't be surprised that I gave the league platform another attempt in 2025, under yet another name. Also not surprising that I got overwhelmed and gave up.

But I did gain a very valuable insight in 2025, maybe something that can break the cycle I've been in for the past two years: I learned how to integrate with every existing fantasy platform.

This past year, I was building a draft sync into DraftKick, and so I was studying the internals of every league site to figure out how they ticked.

Which brings me back to that alternate UI idea that I just dug into once in 2024, the original MVP idea.

• I've spent the past year learning how to completely integrate with each fantasy platform.
• Building my own league platform has repeatedly shown itself to be too big of a project for me to take on. But the UI idea...that seems more manageable.
• People have been asking for the in-season equivalent of DraftKick. What if our custom UI also included all of the features that make DraftKick so valuable?

Hmm...