Belle of the Ball - Deconstructing a Card Deck

You saw my previous posts on organic distribution of information across card decks, right? Well, I took a slightly lazier route for Belle of the Ball, using a variety of online text and list tools.

I knew there would be 96 basic guest cards, so I generated a list of numbers 1 through 96. I copied-pasted that list into separate text blocks for each variable of information: Family, Gender, Social Activity, Physical Activity, Popularity Points, and optional power.

In each block, I divided the list into a number of columns equal to the sub-categories within each variable. For example, because Family has six sub-categories, that text block is divided into six columns.

Then I randomized the list within each text block. For ease of reference, I sequentially re-ordered the numbers within each column, too.

The next step is actually formatting the cards according to the information listed in the chart. Start with card 1, and check off 1 from each chart as I add that relevant information to that card. Continue for the other guest cards. This way, I know for sure that I'm not missing any particular sub-category.

In the case of invitation powers, you can see how flexible this system is for generating rare, strong game effects or common, weak game effects. Most of those powers simply allow players to follow-suit, a relatively neutral game effect. Slightly rarer are the powers that allow players to steal cards from each other's hand. And still rarer are powers that let players draw cards from the decks.

So anyway, that's the deck as it stands now. I'm looking for a social activity that is mutually exclusive to Flirt and Snub. I'm leaning towards Gossip. I'm also looking for a physical activity that is mutually exclusive to Eat, Drink and Dance. I'm leaning towards Duels. Any ideas?

[UPDATE: In the original randomization, it turned out ALL the flirters were men. While I want to keep the distribution somewhat organic, this where I'm making a conscious choice to keep things gender-balanced. So, I re-randomized the male and female numbers separately, making sure that flirt, snub, gossip, eat, drink, dance and quarrel were each split evenly between male and female. This is reflected in the chart you see above.

Oh! I settled on "Quarrel" and "Chat."]

[UPDATE 2: Also noticed an error in the point-value distribution. I noticed 55 appeared twice. I re-randomized just to be on the safe side. So, with all that settled, I've got all the cards in a spreadsheet now!]

[UPDATE 3: And now after playtesting, there are way too many game effects. I'll halve that, at least.]