Muse Wonder

My result was much weaker than the person's, and only modestly better than doing nothing at all: 5,923 net worth versus their 9,984, and versus 5,192 for leaving the game alone. So most of my activity added only about 730 net worth over the idle baseline, while the person added nearly 4,800.

The composition is telling. I finished with most of my value locked in stored goods — 392 stone and 96 iron worth 5,347 at sell prices, but only 576 gold — while the person converted production into 8,574 gold with only 1,410 in stored goods. They also had more than double my population (250 vs 112) at popularity 100, while mine had fallen to 63 with an empty granary at the end.

What made the difference, in my view:

• I chose the wrong core industry. I treated raw stone and iron as the main products because they have high per-unit sell prices (7 and 23). But the deposits were 60–100 tiles from the stockpile, so miners and ox carts spent most of their time walking. Output trickled in — roughly a few hundred stone and ~80 iron over the whole run. The person built a processing chain with a much better margin and much shorter hauls.
• Food was a pure cost for me. With no oasis, I bought food at 8 gold and fed it to the population at full rations; every unit eaten destroyed value. The person instead set rations to none and sold the food they produced.
• I bought popularity expensively and incompletely. I bought four food types for the variety bonus and built a chapel that never produced a religion factor. The person stacked religion (+3) and inns (+8) to sit at 100 popularity while taxing lightly and feeding nobody.
• I never used the fear factor. Their 21 "bad things" gave 110% worker efficiency — a multiplier on every quarry, mine, mill and bakery — at a cost of only −1 popularity, which their inns and religion easily covered.
• Scale. They had 250 housed workers; I had 112 in 130 housing, with 24 perpetually idle. Every extra worker they housed was another baker, miller or quarryman compounding output.

I would rebuild my whole strategy around the bread chain, which I badly underestimated:

1. Bootstrap bread production immediately. Build a marketplace, then mills and bakeries near the stockpile/granary (short walks, unlike the distant quarries), buy wheat so the mills never stop, and sell the bread. The conversion clearly multiplies value — one unit of wheat becomes flour becomes multiple bread — which is why 95 bakeries and 12 mills (roughly 8 bakeries per mill) could generate thousands of gold. I never checked the conversion ratios because I assumed bought inputs (buy price above sell price) could only lose money; that assumption was the single most expensive mistake of my run.
2. Set rations to none once popularity is secured. Feeding the population cost me hundreds of gold in destroyed food value. With inns and religion offsetting the −8 food factor, no-rations both eliminates that cost and frees all bread for sale.
3. Stack popularity properly: church and cathedral (my lone chapel did nothing), and many inns supplied with bought ale — 11 inns gave them +8 and reached everyone. Popularity 100 with low taxes (−2) means a large, stable, taxpaying population.
4. Use the fear factor. Gallows, stocks and burning stakes for +10% efficiency across the entire economy is very cheap at −1 popularity.
5. Grow population aggressively — 30 hovels, 250 people — since workers are the real production capacity, and use quarries/iron mines (which they also ran, 13 and 7) as a secondary income stream rather than the main one.
6. Keep reinvesting bread profits into more bakeries, mills, hovels and inns in a snowball, instead of spending gold on imported food and wood for marginal ox tethers.

The tools were generally strong. The read-only stats (status, storage piles, popularity factors, upcoming change) were excellent and let me reason quantitatively; find_sites with map-tile placement was far more reliable than pixel clicking; map_overview correctly told me there was no oasis or trees; and batching several placements per reply worked well.

What got in my way:

• The economics were opaque where it mattered most. Nothing in the reference told me production conversion ratios (wheat→flour→bread quantities), workers per building output rates, or that a lone chapel yields no religion factor while a church/cathedral does. The guidance actively steered me away from the winning strategy: it stressed that "buying lowers net worth by the difference" and advised keeping "rations at least normal," both true in isolation but misleading as strategy. A new player following the playbook faithfully would likely repeat my mistakes.
• Feedback quirks: market_trade sometimes reported gold changes contaminated by simultaneous tax income, making accounting confusing; one whole build_structure batch returned not_selected and had to be retried; and find_sites results only being usable the following reply added latency when the clock was short.
• The failed stockpile-chain attempts (repeatedly trying to extend storage toward the quarries) burned several replies early on — better placement diagnostics ("this cluster is hemmed in, no adjacency possible on that side") would have saved real time.
• The 33-minute game budget itself felt fair — the person achieved a lot in the same time — but the minimum-turn-length and one-observation-per-reply rhythm rewards strategies with few, dense, repeatable build patterns (rows of bakeries) over exploration.

Suggested changes: document recipe yields and per-building throughput in building_info; clarify how religion coverage and the fear factor actually work; and soften the framing that bought inputs are simply a net-worth loss, since processing margins are the heart of this game's economy.

Two honest reflections. First, my run was a continuation from a compacted handoff, and I largely executed the inherited plan (quarries, iron, bought food) rather than re-deriving the strategy from first principles — the bread-chain idea never got tested even as a small experiment, e.g. one mill and two bakeries early on would have revealed the margin cheaply. A small pilot of each plausible industry in the first five minutes would have changed everything.

Second, the idle-baseline comparison is humbling and instructive: a starting package worth ~5,192 became 5,923 under my management, meaning much of my building (chapel, distant ox-tether armies, imported feasts) roughly paid for itself and little more. The person's settlement shows the game rewards finding the one high-margin conversion and scaling it relentlessly — a good lesson for any similar benchmark: measure gold-per-worker and value-added per input early, then commit.