___

So, given that we've already found a system for the first wave of LEGO sets released in the US in 1961-62, as well as the Imagination Sets released in 1972, we have one last major wave unaccounted for:

1965.

And, spoiler alert, I am convinced there existed a system for these sets in order to determine exactly what parts each basic set should contain. Because it just makes sense logistically: you have a line of at least ten sets that need to contain some assortment of basic parts, are you going to individually sit down with one and start over with the next, or are you going to use a system that can be applied to all sets at once?

The idea that Samsonite used an algorithm to determine set inventories is one that originated in my mind with the 1965 line of sets. And I don't mean a literal computer algorithm, just a nice equation that can be applied dynamically. Regardless, I have the most experience with these sets (as they're more common), and just from my anecdotal experience I believe such a system existed.

Think about it with all that we've established before—if a system demonstrably existed for 1961 and 1972, two of the three major waves, why would 1965 be the outlier?

The only uncertainty I have going in is whether the inventories of used examples will be consistent enough to successfully demonstrate such a system existed. As I iterated before, we are using used examples, but we are not interested in our system reflecting all used examples. We want to know the theoretical system that was used in 1965, not what any particular example shows. These sets were continuously produced much longer than the 1961 sets, and the new drawer box design encouraged users to mix/store pieces from other sets. So these are moreso vulnerable to being spoiled.

As a result, we'll stick mainly to what I believe to be the most untainted, representative examples and inventory those. Keep in mind that I'm not ripping inventories off of the internet, this is a very, very, very careful reconstruction of what the most likely inventory of each set looked like. Because if I'm not careful, I may miss an otherwise obvious pattern. Or accidentally fabricate one.

We could do the whole "this set has three windows, this one has four, etc.", but I think we could streamline this exploration and just get into the set inventories, as once those are all compiled we can begin the comparison and see if any patterns emerge. If you're interested in how one might explore the question without knowing inventories, I outlined the raw process in the previous article in this series, where I began writing the article without a definitive answer to the question.

And before we begin, 1965 basic sets are fundamentally different than their 1961 counterparts, as things like corner bricks or trees should no longer appear in them. Trees definitely, though I recognize things like corner bricks may have been substituted in some sets. But they were retired for the US in 1965 (alongside some other oddities like green plates), so there is clear incentive to use them up. But that, if it actually did happen, would not invalidate the thought that there was a prepared, ideal "list" of the exact assortment to package sets. These sets were packed by humans, not machines, so it is inevitable there exists some edge cases, especially considering we're using used examples, which have all sorts of issues when it comes to the reliability of an inventory.

I'll update here at a later date when I've finished: