### A or B? Two Scoring Options for Belle of the Ball

After playtesting Prototype L a few times, I've seen analysis paralysis set and delay the game. This is partly from the peculiar scoring method. For some context, consider the symbols on the top left of each card to be what that guest is talking about. So Queen Jewel Jaite is talking about conversations, coffee and shields. Make sense? Okay, here are the two scoring options. First, let's see how it works now.

A: Cross-Preference

The left guest grants 1 point per guest with a speech bubble. (The middle guest and right guest qualify, noted by the arrow.) The middle and right guests grant 1 point per guest with a heart, but there are no such guests in this group. This group earns 2 points.

Here's a more complicated example. The left guest grants 1 point per guest with a heart. (The middle guest and right guest qualify, noted by pink arrows.) The middle guest grants 1 point per guest with a speech bubble. (The left guest qualifies, noted by a blue arrow.) The right guest grants 3 points per guest with a tree. (The left guest qualifies, noted by a yellow arrow.) This group earns 6 points.

Keeping track of these cross-relationships is a lot of mental overhead when you're also thinking about how to combine charms, tracking your opponent's party, and hoping to satisfy your Belles. So I am considering this simpler scoring method.

B: Matching
In this scoring system, I'd remove the whole group scoring diagram from the cards entirely. All that remains are the symbols on the top left corner. The new scoring system rewards you for having matching symbols in a group, one point per matching symbol. In the example above, the group has two shields and two speech bubbles. This group earns 4 points.

The group above has two hearts, but otherwise has no other matching symbols. This group earns 2 points.

Naturally, my inclination is to favor the option that takes away visual clutter from the card. It certainly makes more thematic sense that a group "discussing" the same subject would enjoy each other's company more. Think of it kind of like the Sims visual conversation. "You like trains? I love trains!"

I'm eager to see how the other playtesters feel about option A and if they'll respond better to option B.

1. I have an initial strongly positive reaction towards B.

The main problem I foresee in B is: This really really rewards stacking very similar guests, which could dramatically change the value of cards if there are many more (say) heart and fish guests than heart and tree guests.

2. Does this help?

There are 24 Hearts, Snubs, and Chats each.
There are 18 Cakes, Books, Cups, and Musics each.
There are 12 Fish, Gem, Moon, Shield, Sun and Trees each.

These three possible options are randomly, but somewhat evenly distributed. For example, within Snub, there are 6 cakes, cups, books and musics.

The final game may include an atlas of the full card breakdowns, to make all these probabilities clear for super-analytical types.

3. The "somewhat" in "somewhat evenly" matters a _lot_ more under B than A, because the value of a card goes up much quicker with the number of cards that share its partial or complete interest profile.

4. Definitely. Care to take a look at the spreadsheet?

Check the columns titled @ATT-A, @ATT-B, and @ATT-C.

5. Okay, I'm sold!

As a minor point, something funny is going on with Dan Dacrey, Tiny Toffman, and Ingrid Inverse: the Match column doesn't accurately report the match.

6. Plan B seems to make the game a lot more easier to read, lightens the card, and leaves more "brain time" for the other aspects of the game. Even if it means to rebalance a bit the spreadsheet, why hesitate ? :)

7. The bright side is that it actually doesn't even require editing the spreadsheet! I'm going to send playtesters instructions for an A/B test to see which they prefer.

8. Fixed! Thanks for catching that.

9. I have an immediate positive reaction to your simpler solution and I have two big reasons for this. First, your in-fiction reason is wonderful and very on point. Second, I'm thinking of my not-very-rules-patient wife who enjoys games but hates learning them. I think explaining it with an in-fiction justifcation that leads to the simpler scoring is win-win.

10. Thanks, Josh! When you get your prototype, mind doing some A/B testing with both scoring methods?

11. My vote is Plan B. Easier to see and calculate in head, and unclutters the card.

12. Maybe go for B+ -- you score 1 point if a symbol shows up on two of your up-to-three cards, but if that symbol shows up on all three cards, score 3 points. (Basically, a triple gets you a bonus. Or to put it another way, a triple is composed of three pairs -- guests A+B, A+C, and B+C -- so score all your pairs. That lets you figure out your score on a guest by guest basis, looking at what pairs you can form from each guest card's perspective, if that makes sense.)

I do think there's some nice texture to be had in A, and I don't think I'll have a problem understanding it and scoring it, but I wonder if there's more middle ground between the two, like my B+ idea plus a small number of lighter-touch bonuses that reward creating *dissimilar* sets.

Then again, *those* scoring incentives maybe live better on the Belles rather than the Guests. Belles would give you a small number of "tricksy" rules that change how you play your side of the game, and how you interpret the data on your cards; but that "tricksy" thing is held consistent for the whole game, so it's not like each new guest card is pushing new rules at you. You're merely seeing more or less value in various guest cards based on what bonuses you can earn.

Yeah... I'll happily test A, but B or B+ is already sounding good to me.

13. I'd agree with Fred Hicks' B+. Looking at it - I can do all this cross-stuff and it gives me a few more layers of complexity with scoring - or I can take B (B+) and I get simple, easy, clarity. And it makes sense to me.

14. What are your thoughts on Fred's B+ suggestion above?

15. Take a look at the next prototype card styles here for reference. A few others have suggested a B+ format wherein matching common suits (the top circle) is worth 1 point each, uncommon suits are worth 2 points each, and rare suits are worth 3 points each. Your thoughts on that idea?

16. Cool. I'm interested to see how that scoring turns out after some playtesting.

17. Definitely has some value to it. I'd still keep the "score each pair" thing so a triple is worth 3x a single pairing. Good incentives to set-build there:

Pair-Common: 1
Trio-Common: 3
Trio-Uncommon: 2
Triple-Common: 6
Trio-Rare: 3
Triple-Rare: 9

18. Assuming you meant:

Pair-Common: 1
Trio-Common: 3
Pair-Uncommon: 2
Triple-Uncommon: 6
Pair-Rare: 3
Triple-Rare: 9

That may require a little scoring guide included in the game, but I dig it regardless. I'm doing a post later this week explaining what those ribbons are all about, mainly they're an even more sub-divided category for the deck (9 ribbons, 8 each).

These may be used in an expansion:
* Belles that want specific ribbons
* Effects determined by number of ribbons in a party
* Targets against specific ribbons

* Pair of ribbons in the same group: Double-score
* Trio of ribbons in the same group: Triple-score

19. Looks good. It could even support a couple of types of scoring. B for your first few games while you are learning and B+ for the full experience.

(If you go with different scoring with rarities, you might get away with a tiny set of 1/3, 2/6, 3/9 labels under or beside each rarity circle to help remind of the scoring.)

20. I'm tempted to have the ribbons work a little differently. Like, maybe each Belle has a particular kind of set of ribbons as one of her score-affecters. (For every Star + Moon ribbon pair you build, get 10 points; Star + Moon + Sun, get 20 or 30. Something like that.

21. *phew* So assuming Belles reward grouping different ribbons, that makes 72 possible combinations? Clearly too many for the base game, but plenty of room for expansion.

For the base game, I would probably figure out a color wheel and narrow down to just analogous ribbons. (18 Belles)

But all this is assuming this doesn't create more analysis paralysis. One of the things playtesters have liked so far is that the guests and Belles are all focused on those same three suits, without divergence.

22. Right. So why are you introducing the ribbons, again?

23. Fair question. :P

I guess I'm pulled in two directions: On the one hand, I want to make the base game as simple and elegant as possible. On the other hand, I'd like to release small expansions that pull extra info out of the base deck.

To do the former, I worry I will limit the potential for expansions. Just not enough data on the cards.

To do the latter, I worry the base game visuals will be too cluttered. (See: Race to Adventure)

24. You could ditch the ribbons/tokens/whatever the hell they are (visually, I see the three core symbols as arranged upon a ribbon, so calling that dangly bit at the bottom the "ribbon" is weird), and still support that, though; an expansion would be about crafting runs (sequences of non-identical symbols -- that's info that's "in the cards"), playing off of wardrobe (you've talked about hats and scarves before), and introducing guests with a few new powers that you don't include in the base game.

Right now, after you wait a month or so for playtest feedback following the prototypes, I'd counsel you to do a design phase that involves cutting things back until you have the most basic version of the game possible, so you have a sense of what you *could* but don't *have to* have in the core game. You can add some of that stuff back in, making your complete core game, and use the remainders as fodder for an expansion.

Right now tho those dangly bits just seem dangly. Snip 'em off.

An elegant game that builds complexity using as few elements as possible is more likely to be approachable, fun, and appreciated than something carrying a bunch of at-least-in-the-short-term extraneous visual data.

25. Wise as always. You'd think by now I'd learn that "more from less" is always the better solution. :P

26. Most definitely! Also some testing of "explaining the rules before you even play them." :)

27. So I was just playing BANG! last night and admiring how they had several levels of information (roles, character powers, special effects, poker ranks, unique iconography for ranges) then on top of that we used the High Noon expansion, which sets a different game condition each round. None of it felt hard to analyze in the moment, though.

28. Yeah, that concern had occurred to me.

I'd be tempted -- though this is really head-bending work -- to engineer it such that any potential "killer app" combo of guest-powers was necessarily had by putting unlike guests together. But that requires a LOT of watchdoggery on how the card combos can be built, and careful rather than randomized distribution of abilities.

29. I'd lose the organic distribution, but this could easily be resolved:

72 cards
Primary Set: Divide the deck into thirds: 24 cards each A, B, C
Secondary Set: Divide primary set by four: 6 cards in each Secondary Set: 1,2,3,4
Tertiary Set: Divide each fourth by six: 1 card each Tertiary Set: i, ii, iii, iv, v, vi

So there is only one instance of each unique combination. There is only one A2ii. However, there are still enough instances of each individual suit across the whole deck to make valuable combinations. You'll just never find a perfect match.

30. Since the tertiary pairs are triples are highest value, that's where I'd impose some charm-based rules on the design.

Like, maybe it makes sense to say that "trees never duel" or whatever, thereby forcing you to go only pair your trees if you want dueling (or whatever the charm's called now) to be a part of your group.

Similarly, if there's a particular kind of charm that it's super valuable to have more than one of in a group, I'd make sure to distribute that charm so it only occurs once, MAYBE twice, per tertiary symbol.

But, yeah, from a design perspective, that's exactly why a 'symmetric' distribution of the pairing symbols might make sense. Maybe it's the charms where the organic feel comes in.

Then again, maybe you could start utterly symmetric, and then introduce a very small set of asymmetries into the mix if (and only if) playteresting indicated the symmetry was producing some kind of dullness.

31. That makes a lot of sense. So considering these charms in the abstract, see the card map attached for reference:

!,@,#,\$,%, æ and ^, are the rarest charms, each appearing only twice in the whole deck and are on cards that are totally different from each other.

&,*,(,-,+,ß and = are uncommon, each appearing four times. They sometimes share a primary attribute, but never a secondary or tertiary.

∑,†,¥,ø,π are very common, each appearing six times. They may share a primary or secondary attribute, but never a tertiary.

This creates a really brain-bending Sudoku puzzle. :P

32. Sure seems worth testing. :)

33. *phew* 19 charms, eh? I look forward to that bit of development :P

34. Oh! Sorry, I thought your symbols were tied to actual, already-devised charms (even if "in abstract"). What can you do with your actual quantity of established, understood, tested charms? :)

35. D'oh! Slight revision to that chart. Spotted some inconsistencies. Geez, Sudoku card game design is hard!

36. Revised the map!

37. Let's see, breaking it down by rarity:

Rare
-----------------
(These cards must be discarded from the party to activate their ability.)
Dismiss a guest from opponent's party
Lure a guest from an opponent's party
Lure group from an opponent's party
Score 1 point per (1) guest in your party
Score 1 point per (2) guest in your party
Score 1 point per (3) guest in your party
Score 1 point per (4) guest in your party

Uncommon
-----------------
(These are cards may stay in your party, but you must take a special action to activate their ability. Also, these allow your opponents to take actions of their own.)
Mingle move two of your guests anywhere within your party. Everyone else moves one.
Draw: Draw two cards from the deck. Everyone else draws one.
Backdoor: Draw two cards from the discard deck. Everyone else draws one.
Invite two guests to your party. Everyone else invites one.
Group three guests in your party. Everyone else groups two.
____ (Need to add a charm here)
____ (Need to add a charm here)
____ (Need to add a charm here)

Common
---------------
(These are cards whose abilities are passive, they do not require a discard but also do not need a special action. Also, these haven't been tested at all. Entirely hypothetical.)
+1 Draw: When you draw, you may draw an additional card.
+1 Invite: When you invite, you may invite an additional guest.
+1 Group: This group may hold one extra guest.
Lock: Guests in this group may not be split apart.
x2: This group is worth double points.

38. Or just remove one of these categories and have some guests just not have Charms at all. That would certainly be simpler.