I left the gambling industry in 2014 with an idea for a blackjack game bouncing around my head. I finally launched it on March 1st, 2026.
So just twelve years, three stalled attempts, and roughly eight weeks of actual work at the end to get it over the line.
The origin story
I spent years as a product manager at online gambling companies.
We always wanted to make the games fun to play. This tended to be much easier with the flashing lights and sounds of a slots game compared to traditional table casino games.
The house already has their mathematical edge, so the last thing anyone wants is a dealer that makes the player feel bad. The vibe tends to be pretty low key and functional. Just calling out card totals and binary hand results.
But if you’ve ever played cards with your mates you know the best part of the game isn’t the result of the hand. It’s the table talk. It’s your friend calling you a *cough* “idiot” for hitting on 18, or chasing a straight draw in poker.
What if a blackjack game had a dealer that actually reacts to you? (And not politely)
The idea amused me at least. I just struggled to get anywhere with it.
The graveyard
Attempt zero: the easter egg (2013)
I was Product Manager on the Paddy Power Vegas app, and managed to smuggle a couple of personality moments in.
We had Paddy himself in the studio to record dealer commentary, and added a “don’t touch the chips!” line that triggered when players tapped the chip rack.
It seemed to go down well with players, and perhaps we could have got away with more given the irreverence of the brand. But it was always going to be awkward in the highly regulated real money arena.
Attempt one: the Alexa skill (2020)
Six years later, at the peak of the Alexa gold rush (was it just Jeff that got the gold?), I decided a voice-first blackjack game with a talking dealer made sense.
I had paid a real human voice actor to record a bunch of commentary for an earlier dice game, and this felt like the only way to get anything approaching the personality I wanted.
What killed it was the medium. Voice UX turns out to be a brutal format at the time for fast, reactive banter. The latency flattened the punchline, and any tweaks to the commentary would mean booking another studio session.
There is a half working version still sitting in my developer account, but it quietly died there.
Attempt two: the standalone app (early 2020s)
At some point I took another run at it as a standalone mobile app. I even hired a guy on Fiverr to knock out some designs. And they are stupendously horrific.
This one died the most obvious death of all: I got excited, did the fun part (the designs 🤮, the persona, the script), and then hit the wall of everything that was not so fun.
Game logic, authentication, leaderboards, hosting etc. The boring 80% of getting something to actually work. I’m kinda technical so it wasn’t just a skills gap…
I’m a procrastinator at my core. If something doesn’t excite me, it can stubbornly sit forever near the bottom of a to-do list. No guilt, because it’s still on the to-do list right?
What actually changed
I could have built some version of this in 2020, or 2022, or 2024. But it would have taken me six months of evenings, I would have hated every minute of the boring bits, and most likely would have just given up again.
Fortunately, the cost and speed of the unglamorous 80% plummeted.
Two concrete examples:
The voice acting: In 2020, giving the dealer a voice meant paying a human to come into a studio. If I forgot a line, or wanted to add a new scenario, I had to book them again. That friction alone was enough to kill any iteration loop.
This time round I used ElevenLabs. New lines cost pennies and take about thirty seconds each.
The boring nuts and bolts: Split edge cases, authentication, leaderboard backend, daily challenge reset logic, responsive layouts, the bit where you edit your profile…
AI coding assistants collapsed the time cost of the parts I didn’t want to do from months to hours. When the boring parts stop being a six-month commitment, it allows you to concentrate on the fun parts.
I started the renewed effort in January. Had a working beta by mid-February. Launched on March 1st.
Twelve years of “one day” became eight weeks of actually doing it.
What shipped
Blackjack Smack is a daily challenge blackjack game. You get 21 hands, a £1,000 starting stack, and a dealer whose entire personality is to make you regret every decision. Every day there’s a leaderboard showing whose bankroll survived.
There are no subscriptions, no coin packs, no “buy more chips” loops. Given my background I had strong opinions about what I did not want this to be, and avoiding those spammy mechanics was top of the list!
With a core of regular players joined by newbies every day, I’m still adding new features and looking forward to introducing a new dealer soon.
The obligatory takeaway
I’m sure everyone reading this has an app or game idea they’ve been carrying for years. What is really holding you back?
Maybe it was technical proficiency, or the expectation that it would take ages and cost a fortune.
Go and check whether that’s still true.
The boring 80% of most side projects got dramatically easier in the past couple of years, so a lot of ideas that weren’t worth starting suddenly are.
That doesn’t mean they’re gonna be any good, but it’s much quicker and cheaper now to find out!
If nothing else, go play Blackjack Smack. The dealer’s waiting, and they’re not in a good mood.
