Week 1: way bigger than I expected

Nov 11, 2026

Welcome to my first Daily Banana dev blog. Every week I'll write about what changed, what I tested, what feedback you gave, and what's coming next.

Before we get started, I honestly thought Daily Banana would be a small, silly game my friends would play and maybe share with a few groups. But over the past week, it's grown beyond anything I could have imagined, and I've been making updates as fast as I can. I work a full-time job, so I moonlight as a game designer.

This first entry covers everything from launch week. If you want the origin story, I talked about it in an interview with Smidgen Games, the company I run.

Making exact guesses feel special

You used to get a standard banana for any 100% round. But if you nail the exact number, that should feel way more rewarding than a regular “good guess.” So exact guesses now get a 💯 and a golden banana. The golden banana went through a few design iterations as well.

I recently got feedback from a few players that they wanted to visually see where they sat within the player base for the day (similar to how other daily games do this). So I added a chart to your results that shows where your score sits among everyone who played that day. It makes sense. I just never thought the game would need one, because I didn't expect it to have more than like 10 players. With this many people playing, seeing where you land is half the fun.

The double-rounding bug

Someone I know messaged me directly about a round that didn't add up. Going by the numbers the game showed them, their guess should have been exactly on the mark, but the game said they were off by 10. Frustrating user experience when the number on screen didn't match the actual calculation.

When I dug into the code, I found I was rounding in two places: once before the calculation and again after it. Every time you round, you drift a little further from the real answer, and on some questions it added up to 20 to 50 items different. So I made sure that I'm only rounding at the end, after the game calculates the answer. The fix goes live today, so what you see during a round now matches how it's scored.

Moving to a real database

Players were seeing slow load times on some of the numbers, and my admin dashboard with analytics was taking 9–13 seconds to refresh for a player and 30+ seconds for my admin panel. Again, I didn't expect it to need a backend database for the 10 players I envisioned. But with this many players, and everything we record for every question, the old storage wasn't keeping up.

I used to be a backend developer and I'm comfortable with SQL, so I went with a relational database, Postgres. Load times dropped dramatically (1–3 seconds, down from 10–25 seconds), reads and writes are instant, and it should scale for many months to come. I wanted to do it now, while the player base is still small. If something was going to go wrong, I'd rather it happened now than after thousands more people showed up.

While I was in there, I fixed another bug. Upcoming dailies were being shipped inside the game's code. Now each daily stays hidden until its day, the way Wordle and other daily games work.

Where the share button lives

Something I noticed rather quickly: not many people were sharing their results, and I had a hunch it was the button's location. It sat below the list of question results, so you had to scroll all the way to the bottom to find it. Anyone new to the app, who visits on mobile, would just assume you can't share your results.

So I ran an A/B test with several hundred players. Half saw the share button near the top, above their results, and the other half saw no change. The players who saw it up top shared twice as often.

This is the change I'm proudest of this week. I never expected enough players to run a solid A/B test. But I noticed a trend, leaned on my UX background to set up the test, and got a definitive answer. And the more people share, the more people hear about Daily Banana.

Results or home page?

I was curious about another thing: when you come back after finishing your daily, is it better to land on your results instead of the home page? No surprise, it was. Players who went straight to their results on their second and third visits spent more time in the game and shared more than players who were sent to the home page every time. So now, all players who return after playing the daily get sent to their results screen.

Are hints worth it?

I was curious whether hints would help players out, so during the first weekend challenge I ran another A/B test. Some players got a size hint with each question, and the rest played without one. The difference was negligible: the average score was 290 out of 600 with hints and 282 without. Size hints didn't help at all, so we removed them. All they did was make each question take longer, and add more reading and confusion.

Brown or yellow?

A designer friend asked me why a darker brown meant a higher score than the brighter yellow, when yellow is the one closer to a banana. My first instinct was a scale that goes from light to dark as scores go up. After exploring both orders side by side, I decided to switch now, before the game gets too popular. Brown is now the lower score and yellow the higher one. It looks nicer and makes more sense. Thanks for the feedback, Adam!

New share emoji order

❌⬜🟫🟨🍌💯

That was part of a bigger round of design work this week:

• New banana art and app icons.
• An updated font, Sora, which fits the game much better and offers variable font weights.
• A design system in Figma, so I can mock up new pages quickly. I'm also exploring some changes to the menu.

I was lazy when I started Daily Banana and had way too many colors within the original design system, so I simplified things down to 7 essential colors and 4 assistive colors. The original design had almost 14!

Making it accessible

My friend Jason told me the background bananas made the text hard to read, and that it wasn't accessible. I already had settings to turn the bananas off and to switch on high contrast, but I wanted to fix the problem itself too. So I gave every container with text in it a solid background, and now the bananas never show behind the text. The bananas are just here for fun, so they can be covered when they need to be.

I also wanted to make sure almost everything meets at least WCAG AA contrast and works with a keyboard, whether you're on a computer or using an assistive input device. The game also works with a screen reader for anyone who needs one.

Keeping it fair

A few smaller decisions this week were about keeping the game fair and fun.

In user testing, players didn't enjoy trying to guess exact numbers most of the time. For big quantities, hitting the precise number is really hard and not much fun. We adjusted the rounding early on, and it seems to have evened out the playing field. I also updated the score stats to be fairer, and I'm keeping the details vague on purpose.

I've also gone back and forth on showing exact player counts publicly. Big or small, exact numbers could put players off, so the game shows the percentage of players instead. That works nicely for small groups and large ones alike.

Play it where you hang out

You can now add our Discord bot to your server and play Daily Banana with your friends right inside Discord. You can also play inside Reddit, though Reddit scores stay on Reddit for now.

How I plan to maintain Daily Banana

I design professionally, and this is my fun side project. It's a place to bring almost 15 years of my professional design and development experience together. For me, everything starts with an idea, which I usually brainstorm in Figma. I know HTML, CSS, SQL, and Python, so a lot of my designs start out as a Python script or a Flask server. I'm not well-versed in JavaScript, but I can work my way around the code when I need to.

Modern development tools now handle the tedious stuff: server maintenance, setting up variables, test servers, automated deployments, bug fixes, etc. I used to write scripts for all of that or do it by hand, and now it takes seconds. That leaves me more time for the parts I love. As a product designer who's in Figma every single day, I enjoy creating screens, mapping user flows, analyzing data, taking feedback, and tuning the experience for new and returning players. I really love building the brand and seeing what Daily Banana has become. I'm excited to see where it goes.

What's next

I'm watching the player base as it grows to decide whether translations are worth doing. Browsers can translate the words, but some words don't translate well, and grammar differs from language to language. If we do translate, I want to pick the languages that matter most to our players and get each one right. I think that will be one of our biggest hurdles.

There are a few other modifications I'd like to make around the look and feel of the menu screen. I also need to figure out how to recoup costs for server maintenance and database fees. It's either display/banner ads, Smidgen game ads to direct sales to the Smidgen site, or lock advanced features and reporting behind a paywall.

Thanks for playing, and see you next week!