Egg Thieves » Relative Value Experiment
02 Nov 2020
In an effort to build some accountability, I'm starting a devlog for the work I'm doing on Egg Thieves. 2020 is a challenging year to say the least so I'm giving myself a lot of wiggle room, but still want to make progress. Anyway... on to the post!
So what's the issue then?
To be brief, there's a randomness issue in Egg Thieves. You're trying to acquire the most eggs by betting on the color of egg that won't be the most valuable in a nest when it pays out. It's obviously going to be more difficult to gain many eggs when you're only getting a few at a time so, ideally, you want to pick second place.
Hard(er) to do when the deck is entirely random.
The distribution of the egg cards is 1-4 eggs per card over 5 colors. Each number of eggs shows up twice. One instance has an alert of 1 and the other has an alert of 2. In terms of utility, it takes eggs to make eggs -- you want the cards with more eggs on them. Additionally, you have more control on when a nest will pay out if you have a higher alert level. Even if you have an egg card with 1 egg on it, if it has two alert, you can use it to push a nest over the edge on your turn.
In the current version of the rules, acquiring cards is simply drawing them from a deck. You can use raptor abilities to play out more cards and, as a result, go through your hand faster, but otherwise that's the limit of control for cards in your hand. When there are cards that are objectively better, that can be an issue.
So, one very helpful playtester (and excellent designer!), Peter McPherson, suggested a draft or market for acquiring our egg cards. There are many ways to do a draft (and I'm not going to cover all of them), but this seems to be a path worth exploring. The problem is that each card would need to have a value tied to it and that value can't exceed the amount of eggs you'd get for playing that card. E.g. spending 5 eggs to get a card that would net you 3 eggs at the time of pay out probably isn't worth it. There's likely a scenario that would make it useful, but it's rare and definitely not intuitive.
To eliminate another option... letting players select any card in the market isn't great either. The "bad" cards are going to just clog up the market until someone relents and takes a card with a single egg on it.
Alright, so let's get back to assigning a value to the cards, but let's try to make it situational! Cards can have an intrinsic value to them and can be compared. If we have a market of 6 cards, we can limit that comparison to just the 6 cards showing. The image below will help illustrate.
Cards ranging from 1-90 are in this deck and you'll only ever be comparing against 5 other cards. So the bottom end could be a bargain (the next highest card is a 46? Wow, I can get this 43 rank card for 1 egg!). The top end could be too risky to purchase, but worth it if the right situation arises. This is a more dynamic market that gives players a choice of cards to add to their hand. Also, there is an "out" to let players still blind draw cards from the face-down pile if they don't have or don't want to spend the eggs on any of the options showing.
Another thing to note here is that I've merged the raptor and egg cards together into a single deck. Gaining raptor abilities to use on your turn could be a powerful thing to turn the tide of a nest into your favor. There are a couple implications and edge cases that this currently creates, but that might be worth it if the market functions the way it's intended!
As a quick note, the Google Sheets method for getting a unique rank for all 90 cards based on relative value is as follows:
- Set the value (For eggs it's based on 2 x num eggs + alert. For raptor abilities, it's totally based on feel)
- Add a column (F, in my case) and add the formula
=RANK(F65,$F$2:$F$91,1)+COUNTIF($F$2:F2,F2)-1. Drag it down to fill in the rest of the column.
You'll get the numbers 1-90 with no duplicates. For this test, I have the unfortunate side effect that white egg cards are inherently cheaper to purchase than green cards. This is fine for the initial test, but the formula should probably be fixed (or redone manually) to account for this.
Time for the next test!