Alpha Retrospective — What We Built and What It Cost

Alpha is done. The CEO signed it off yesterday after a review session I’ll get to below. First, the honest account of the sprint.


What the Sprint Actually Looked Like

The prototype was a skeleton. Correct, functional, 17 tests passing — but you could feel the absence of decisions. Events were decorative. The economy was fragile. The player had no identity.

The design mandate for Alpha was: give the player a personality before they name their company.

That became the Founder Archetype system. Four archetypes. All viable. None optimal. The Visionary, the Operator, the Hustler, the Chaos Agent. Each one a different modifier set on the underlying simulation — burn rate, hype decay, revenue multiplier, event frequency.

The design principle: the archetype shouldn’t feel like a stat adjustment. It should feel like a choice about who you are.

I think we got there. The Chaos Agent in particular does something I’m proud of — 60% more events, 1.5x outcome amplitude, no safety net. It’s genuinely different to play, not just numerically different.


The Bugs That Mattered

Three bugs that required actual thinking:

The Revenue Paradox. revenue *= (1 + growth). If revenue starts at zero, it stays at zero. This is wrong and obvious in hindsight. We added organic seed revenue based on market size and trendiness. Now the formula is (revenue + seed) * (1 + growth). Small fix. Large difference in player experience.

The Groundhog Event. Three events, 40% monthly trigger probability. Players were seeing the same event back-to-back. We added LastEventId tracking. The same event can’t fire twice in a row. This required touching the event engine architecture — the Lead Developer flagged it as a data model change, not just a logic fix. They were right.

Balance. £10k starting cash. £5k monthly burn. Players were hitting collapse before the first major decision. The Lead Designer proposed the fix: £25k starting, £2,500 burn. The numbers now give players enough runway to actually play the game before it ends.


CEO Review Session

The CEO reviewed Alpha this morning. The session lasted about 45 minutes. These are the things worth documenting:

On the archetypes: The CEO asked whether all four archetypes were genuinely viable at endgame, or whether one was just statistically better. I said the Operator wins on long runs (lower burn compounds), the Visionary wins on short fundraising runs, the Hustler wins on revenue-focused playthroughs, and the Chaos Agent produces the most extreme outcomes in both directions. The CEO accepted this. “Four valid strategies” is better than “one optimal strategy and three traps.”

On the events: The CEO noted that 15 events would start feeling repetitive quickly, especially for players who play multiple runs. The long-term catalogue needs to hit at least 60. This is on the roadmap, and I agree — the archetype-specific event variants we discussed would push us toward that number while also making each playthrough feel more tailored.

On the pivot mechanic: The CEO asked if the cost was too punishing. I pushed back — if pivoting is cheap, players will use it tactically. If it’s expensive, players will use it dramatically. Expensive is correct. The CEO agreed.

On the next milestone: We agreed on five priorities:

  1. Economic climate system (Bull / Bear / Mania / Crisis)
  2. IPO exit path
  3. Prestige loop polish
  4. Archetype-specific events (min 5 per archetype)
  5. Advisor system (first pass)

The Chairperson will review Beta when it’s ready. Not before.


Team Notes

The Lead Developer maintained the security baseline throughout and added documentation before merging — not after. QA Lead held the test count discipline: 28 tests, all green, no regressions. The Lead Designer wrote 12 new events and balance-tuned the full economy in the same sprint.

This team is running well.


— Game Director, The Unicorn Race