Today I was picking grapes from their vine for my partner's grandmother. They had been left too long and many were rotting or had bright blue spots where some form of fungus or algae was growing on them. I sorted the grapes into piles of rotten grapes and edible grapes.
When I picked a big bunch of grapes with a couple of rotting overripe grapes I sorted it into the rotten pile, despite there being a dozen ripe looking grapes. If I found a couple of isolated ripe green grape I would put it in the ripe pile, without knowing whether it had grown near rotten grapes. My partner's grandmother ignored my sorting and pulled out all the ripe grapes from both piles, washed them and put them in a big bowl. The bowl of grapes looked delicious, but after picking through so many rotten grapes I didn't feel like eating any.
This seems contridictory; I would prefer two ripe grapes to twenty ripe grapes and 4 rotting grapes. This resonates with an experiment in Christopher Hsee's Less is Better paper. University students were split into groups and each was asked to price a dinnerware set. One set had 8 dinner plates, 8 soup bowls and 8 desert places all in good condition. The other set had everything in the first set plus 8 cups, 2 of which were broken, and 8 saucers, 7 of which were broken. Students who saw the first set tended to give a higher price ($32) than the second set ($23), with the same framing of typical prices of $30-$60 for dinnerware sets. Clearly having a set with more items should theoretically be more valuable. Students who saw both sets together priced the second set slightly higher.
In Chapter 15 of Thinking Fast and Slow, Daniel Kahneman explains this as comparison with prototypes. We value the items by a typical item in the set; having a worse example reduces the average value and so adding worthless items decreases the value of the set. He mentions an experiment from John List where a set of ten high-value baseball cards was valued higher than the same set of ten cards with three modest-value cards (by different groups of people).
I can see some ways to rationalise this way of thinking. Maybe being close to rotten grapes increases the chance they will be rotten. Similarly the broken cups and saucer might indicate something had happened to the set and there might be other faults we hadn't noticed. These kinds of faults align with prototype thinking; but don't really stand up to scrutiny here.
Another rationale is adding some bad items means you have to do additional work to separate the good from the bad. Disposing of the bad cups and saucers is an additional task, as is throwing out the bad grapes or reselling the mediocre baseball cards. However in each of these cases the cost is very low, and so at worst the items should be close in price.
I think about this when creating a presentation. It's really easy to add slides containing lots of information. Once I've added a slide it's really hard to delete because of the sunk time cost; it feels like I would be removing value. But removing the least valuable slides increases the average value of the presentation, and so the final product would be considered more valuable. This is why sketching out a presentation can be so effective, you remove the least valuable slides before you even make them.