Dungeon depth, bag space, loss aversion, and diminishing returns on wealth gain

Value / Weight
This is the ratio that dominates my play in Fallout. I can't say I enjoy it very much. It's repetitive and best performed by a computer, except if all my loot came with a pre-calculated and displayed value/weight ratio, I don't think I'd enjoy that much. I don't know why I wouldn't. Just one of those human irrationalities.

But there is another problem: vendor cost may be lower than the value I place on it and the value I place on it is a bit murkier. Does a crafting material for my rail spike gun have a value related to the sell price of the gun? No! Well, sort of! No! If I valued the gun the same as the vendor cost, I'd be selling it, not firing it. So not only is the material vendor price almost meaningless, but the proportion of the material to the gun to its sell price is also meaningless. I still need to figure out what the gun is worth to me.

Picture yourself deep in a vault, dungeon, mine, or whatever else you can go deep inside and have a lot of trouble getting out. If you are a young child who knows Lassie, then you are in a well. With a bag filled with rail spike gun parts and boar spleens. Protip: empty your bags before falling in a well, noob.

Okay, you're deep in there and you don't exactly want to pull out a calculator after every kill to see what everything is worth and compare it to what you are picking up. Alright, 507/1.2 vs. 692/2.3 (yea, this one is obvious), but there is also an item of indeterminate value, aka the value I put on it, and weight of 7. Is 507/1.2 greater than x/7? I don't know. Okay let's see, if I can make that gun with this spike and a few other things, but I have extras of these, so maybe I should vendor them, but maybe later on I will run out and then I'll want these, and I can't really value something with an unknown supply...

Are you picturing yourself having fun yet?

We could easily bypass this problem by removing limits on bag capacity or in this case, carrying capacity. Everything would then have a value greater than zero, which given that there is always an empty space, zero is the alternative value to consider. Is this fun? I don't think that is much fun either, just running through the halls filling up bags that never fill up. There are no decisions at all.

Somewhere in here we want to decide a few times, a little bit, without being overwhelmed. By we I mean me.

Dungeon depth / Bag space
Here's the ratio to look at. Ideally over the course of a dungeon we want slightly more than we can carry, bags overflowing with valuable loot, so we're trying to pick out which items are gems and which are colored glass.

But what about the other two things?
Loss Aversion
People don't like losing things. Except bad things, but those are a special case. People don't like losing things that they like to have. This is not at all unreasonable. It is the opposite: very reasonable. And, we try to avoid losing things that we like to have, as an obvious extension of avoidance of negative experiences, in this case, the experience of losing things that we like to have.

I like having parts to make rail spike guns. I don't like dropping or destroying parts to make rail spike guns. I cannot quite determine how much I dislike losing parts to make rail spike guns because I don't have a value for them, so the bad could range from slightly more than none to infinite. I could, theoretically, be psychologically destroyed by the act of destroying a part for a rail spike gun. I suspect this is unlikely.

But beside this there is another set of words that I am going to put in bold type:
Diminishing Returns on Wealth Gain
Five bits of time into the dungeon I pick up an item with a value/weight ratio of 1/1. A pretty crappy item. I can also see that my value/time ratio is 1/5. Five bits further in I find a 2/1 item, but my bags are full, so I drop the 1/1 item for the 2/1. Now I have found 3 value in 10 bits, but can only carry 2 value, meaning that despite finding a more valuable item, I am still at a steady 1/5 value/time ratio. Another ten bits in and I find a 4/1 item, so I drop the 2/1 and now I have 4 value but over 20 bits and I'm finding that despite going deeper in, each unit of time I spend is not yielding any greater return than the one before.

"So where are the diminishing returns," you ask. Every bit I run in, I must also run out. But obviously all that is doing is doubling the time cost of everything. Now what if I'm also spending more time shuffling around these items and calculating costs, which is trivial with 1/1 items, but the 507/1.2 and 692/2.3 items are a bit more difficult, and we've not even considered the crippling indecision of the x/7 item.

Now consider that you run in and every bit you grab a 10/1 item. Then you run out of those and every bit you're finding 15/1 items. And then 16/1 items. You went from 10/1 value/time to 5/1 value time (replacing 10 with 15 rather than gaining 15) to a mere 1/1 time, as bad as if you were in that crappy first dungeon with 1/1 items laying around.

With generous teleporting you might just leave and come back, emptying bags each time, but that is effectively just the unlimited bag space with loading screens thrown in.

The Point
It's not fun to have really long instances with tons of stuff dropping in them, particularly when the relative value of the items is hard to determine. If there is a lot of 1/1 stuff but a clearly distinct set of 50/1 items, then players can more easily reject the 1/1 items as pointless. Absolute value doesn't matter as much as relative value. Choosing between a 1/1 and 50/1 is close enough to choosing between 50/1 and 2500/1 that it is probably better to start with the former scenario and only use inflation if you're trying to drive an economy or create the illusion of progress.