Hacker News new | ask | show | jobs
by knlinux 2848 days ago
That's not true and what they did was probably optimal. The probability of losing the whole collection over time is the same. By spreading collections around into N locations, you would have, on average, N times more fires like this one, but fire would be engulfing a smaller collection at a time (1/N). Moreover, by having it centralized you can reduce costs, hence save money and use it to lower the risk of fire.

What they did by centralizing the collection was the best thing they could do to save it, but apparently they didn't even have money to protect a single museum from fire hazard. Spread the collection around the country and you don't even have money/people for proper maintenance and protection against theft in each individual location.

2 comments

> By spreading collections around into N locations, you would have, on average, N times more fires like this one, but fire would be engulfing a smaller collection at a time (1/N)

If the likelihood of a fire starting in a non-residential building is roughly proportional to the square footage of the building, then you'd have roughly the same total number of fires across all buildings but 1/N of the collection destroyed each time.

That would only be true if the risk of fire is evenly distributed, which it obviously isn't, and that a museum burning down now and then would not motivate better protections for the rest.

In any case, the strategy they chose obviously didn't work and they lost it 100%.

It probably isn’t but it’s safe to assume that this location was at least around the median in terms of fire risk (this location had probably better infrastructure than the average Brazilian city), so the fact that there might be less risky locations doesn’t change much since you’ll then have places with higher risk, where the collection will have burned down even sooner. Even if you knew the best location with the least hazard, optimally you would still want to move all collections there anyway.

The strategy didn’t work since any strategy was likely deemed for failure due lack of funding, not because it was suboptimal.