What is the REAL cost of new molds in 2021?

[danth]

7 hours ago, howitzer said:

and as long as there's no consistent bias the inconsistencies tend to cancel each other out and you'll get a number that is surprisingly close to the correct answer

Exactly, that's what I was hoping for, but you explained it more succinctly than I would have.

[MAB]

It is still an order of magnitude calculation though. And is that a good enough answer for the problem?

We already know a constraint, that a complex mould (dice) is about $200k. So without even doing a calculation based on estimates, a guess of $20k (or $50k or something similar) is likely to be correct within an order of magnitude. That would mean we are confident the answer is between $2k and $200k (anywhere in that range is not more than an order of magnitude out, so not more than 10x out in either direction). If we had done a calculation and got $20k, we'd be happy we are correct within an order of magnitude based on the known  constraint.

If you estimate the  cost of a mould using various assumptions, and come up with any answer for the average mould cost between $15k and $150k, you'd probably be happy the answer is correct (within an order of magnitude). Does it really help solve the problem? If one person gets an estimate of $40k and another $120k, which is  more likely to be correct? We simply don't know without knowing the actual answer. Although we should be confident they will both be about the same level of correctness, since they are not that different (within an order of magnitude) even though they look very different on a linear scale. 

Here it tells you more about the validity of your input estimates than solving the problem. If anything, it is a waste of time trying to calculate it when we already know the order of magnitude of the cost of the average mould. It is $100k. Different people could give other answers of course. What that answer really means is the range $10k to 1m is about right. We already know higher than $200k is unlikely, so someone else might say go for $20k to mean $2k to $200k, cutting out the unlikely range above the  constraint and extending it slightly lower. 

This is not really a Fermi style problem, where we need to calculate the order of magnitude, as in could it be $1, $10, $100, $1k, $10k, $100k, $1m, $10m, etc as we already have the order of magnitude. If an calculation based on various assumptions gave us $100, we'd know our assumptions were incorrect. Similarly if we got $1m.  But if we got $40k or $120k, we'd assume we were close enough. Which leads to the real question - is this close enough good enough. What is thereal point of the question and is such an estimate good enough?

Edited July 18, 2021 by MAB

[LegoDW]

Feel the general consensus is correct, TLG does not re-release sets because they do not believe they will be as profitable as the ones offered today.  And I don't blame them.  

If you want an old set shell out the +$1000's for it or Bricklink something very close which in itself is usually a enjoyable experience.