We Need More Vanilla. The Old One's Expired.
Last time I wrote about vanilla creatures, it was prompted by question #48 's mentioning of "virtual vanillas" on the Great Designer Search 2 multiple choice test. Intrigued by the definition, I was attempting to figure out the magical ratio of virtual vanillas and French vanillas to every other kind of creature when designing a Magic: The Gathering set. That blog post is now outdated by this very post, but in case you still want to read it, here's a link to the original post.
What I didn't cover before were how many vanillas there should be in a set, virtual French vanillas, and the distribution of these different types of vanillas among commons and uncommons. Funnily enough, I'm doing all this analyzing and Mark Rosewater tweets me around the time after I posted the first vanillas blog post saying it isn't an exact science. Even if it's false irony, it was amusing to me nonetheless.
So, dear reader, if you're a fan of designing Magic sets and haven't yet solved the conundrum (sphinx) of how many vanilla/virtual vanilla/French vanilla/virtual French vanilla creatures to include in your set, then you're in the right place. Whether you're designing a large set, a small set, or a Core Set, read below for some numbers to help you out. Though, I should disclaim: I'm not a member of Magic R&D, and a lot of the data gathering was done manually by going through and viewing a lot of commons and uncommons. So, that means 1) I could have missed something vital that R&D keeps in mind, throwing the numbers off and 2) I could have made human error in my data gathering and some numbers are off in some places. At least you'll (like I said last time) be in the general ballpark. (Though, hitting a home run so well that it exits the ballpark runs contradictory to the phrase "in the ballpark" as being a good thing; but I digress.)
Before we begin, I want to make clear why it's so important to care about all these vanillas. The answer is board complexity. The more complex state the game is in, the more difficult it is to make decisions as a player. Since creatures are a large part of Magic, they make a larger impact on the complexity of the game when there's more rules text. There's a limit to how much an average Magic player can account for in his or her mind. It's important to have the right amount of complexity in a set's creatures, so the player doesn't get overwhelmed, heavily mentally taxed, and/or take too long making decisions. There may be a certain level of complexity for instants/sorceries/enchantments/artifacts/lands, but that's another topic.
Just recently, Aaron Forsythe mentioned board complexity in his introduction as a judge in the second episode of The Great Designer Search 2. Amidst his words, he mentioned word counts per creature (vanilla creatures help reduce that) and that Lorwyn (the set) was "incredibly board-complicated." It's no coincidence that Lorwyn's percentage of vanilla-ish creatures are among the lowest in all the sets currently in Extended (This is all the sets I data-gathered on my spreadsheet). And Morningtide (The small set in the Lorwyn block)? It has the lowest percentage of vanillas in Extended. It doesn't even have any vanilla creatures of any kind at uncommon.
So here we go.
What Does Vanilla Taste Like?
First, let's determine what makes a vanilla, virtual vanilla, French vanilla, and virtual French vanilla what they are. There are a couple of key factors that I went by for determining the differences among the vanilla types.
- The difference between a vanilla and a French vanilla is a creature-based keyword. A creature-based keyword is one that appears mostly on creatures. It also does not have to be evergreen. What this means is that while the staples of flying, deathtouch, and trample count, so does the block-specific keywords such as infect and exalted. Notice I say appears "mostly" on creatures, which makes exalted O.K. despite having appeared on non-creatures like Angelic Benediction. This is confirmed by Mark Rosewater. Cycling does not count as a creature-based keyword since it appears on all sorts of non-creature cards. And ability words like landfall and chroma are not keywords.
- The creature-based keyword must not cease to function after the first turn on the battlefield. The two keywords that come to mind are haste and flash. Haste is a creature-based keyword, which means you'd think a creature with just haste should be a French vanilla, but it's actually a virtual vanilla in disguise. After that first turn, it's just a plain vanilla. And flash only matters for when you cast it. Once you cast the creature, flash doesn't do anything anymore.
|That's no moon... It's a virtual vanilla!|
And then there are rare exceptions like changeling which don't actually do anything in a vacuum. Though, you could argue that it does count since the environment that creatures with changeling existed in meant that changeling made a difference. But I didn't count changeling as something that mattered when I did the spreadsheet. Do I regret it? I'm not sure. Anyway, here's the differences among the different types of vanillas:
- Vanilla: A vanilla creature has no rules text at all (except in the case of changeling, at least, by the definition the spreadsheet went by). If a creature has nothing but cycling in its text box, it doesn't count as a vanilla. That's because, when the creature is in your hand, the cycling contributes to board complexity by giving the player another option to consider in the overall picture of the current game state. A vanilla creature is also a virtual vanilla creature.
- Virtual Vanilla: This is a creature that functions like a vanilla creature after the first turn on the battlefield. These creatures could have abilities like "enter the battlefield" (ETB) effects or haste that aren't in effect after that first turn. Thus, all vanillas are also virtual vanilla.
- French Vanilla: French vanilla creatures have in their text box only creature-based keywords (and always at least one of them, otherwise, it'd be a vanilla). If it contains keywords like evoke or reinforce, then it doesn't get to be a French vanilla. A French vanilla creature is also a virtual French vanilla creature.
- Virtual French Vanilla: Virtual French vanillas (which includes all French vanillas) have at least one creature-based keyword and any additional rules text that don't take effect after its first turn on the battlefield. After the first turn on the battlefield, it must function like a French vanilla. A creature with flash and first strike is a virtual French vanilla while a creature with just flash is a virtual vanilla, not a virtual French vanilla.
- ETB effects
- Effects that trigger when you cast it, such as cascade
- Alternate or additional costs, such as suspend or the text on Flamekin Bladewhirl
- Other abilities I may have missed
Now that we've got what defines our vanilla-type creatures, there's the creature cards to look at to see how many of these vanilla creatures we need to include in a set. Since design standards change over time, I want to get as close to modern design patterns in terms of vanilla creatures as I can. I looked at creatures in all of Extended, which means Tenth Edition and onward, just to set a good cut-off for "modern." I didn't look at every Magic set in existence since what was done in Alpha may not apply nowadays when designing a Magic set. That'd be more of a historical study on how vanilla creature ratios change in sets over the course of the game's lifetime, which is still an interesting topic (I'm nodding in your direction for that suggestion, MTG Color Pie).
Also, when counting the average number of creatures in each type of vanilla, I kept in mind the change in set size that Wizards decided on starting with Shards of Alara. So, while I counted sets further back than Shards of Alara to determine percentages, I didn't let those sets skew the "number of vanilla creatures" counts that an average set would need.
So, after taking the data into account, here's the "boring" part of this post. Though, the difference between a large set and a small set are seemingly non-existent. The data shown below is just variance from actual results. The average large set and average small set's total vanilla creatures percentage of total creatures differed by just about one percent (39.45 vs. 40.49. Core Sets had an average of 61.4). Anyway, here's the numbers:
- 9 - 16 common virtual vanillas with the average being 11.75; 4 - 6 of those should be pure vanilla with the average being 4.75
- 2 - 4 uncommon virtual vanillas with the average being 3; 0 - 1 of those should be pure vanilla with the average being 0.5
- 10 - 15 common virtual French vanillas with the average being 12.5; 5 - 14 of those should be pure French vanilla with the average being 9.25
- 4 - 8 uncommon virtual French vanillas with the average being 5.5; 0 - 7 of those should be pure French vanilla with the average being 3
- 19 - 29 total common vanilla-type creatures with the average being 24.25; this is out of sets with 53 - 57 total common creatures with the average being 55
- 6 - 11 total uncommon vanilla-type creatures with the average being 8.5; this is out of sets with 25 - 40 total uncommon creatures with the average being 31.5
- 4 - 10 common virtual vanillas with the average being 6; 1 - 2 of those should be pure vanilla with the average being 1.66
- 2 - 7 uncommon virtual vanillas with the average being 4; 0 - 1 of those should be pure vanilla with the average being 0.66
- 8 - 12 common virtual French vanillas with the average being 10; 4 - 12 of those should be pure French vanilla with the average being 8
- 1 - 4 uncommon virtual French vanillas with the average being 2.66; 1 - 3 of those should be pure French vanilla with the average being 1.66
- 12 - 20 total common vanilla-type creatures with the average being 16; this is out of sets with 27 - 39 total common creatures with the average being 33.66
- 3 - 11 total uncommon vanilla-type creatures with the average being 6.66; this is out of sets with 18 - 22 total uncommon creatures with the average being 20.33
- 17 - 17 common virtual vanillas with the average being 17; 10 - 11 of those should be pure vanilla with the average being 10.5
- 2 - 4 uncommon virtual vanillas with the average being 3; 2 -2 of those should be pure vanilla with the average being 2
- 15 - 15 common virtual French vanillas with the average being 15; 13 - 16 of those should be pure French vanilla with the average being 14.5
- 7 - 8 uncommon virtual French vanillas with the average being 7.5; 5 - 6 of those should be pure French vanilla with the average being 5.5
- 32 - 32 total common vanilla-type creatures with the average being 32; this is out of sets with 51 - 53 total common creatures with the average being 52
- 10 - 11 total uncommon vanilla-type creatures with the average being 10.5; this is out of sets with 27 - 29 total uncommon creatures with the average being 28
And that's that. Is this really all there is to say about this topic? There's a bunch of numbers in that spreadsheet, there's definitions for how to identify vanillas, virtual vanillas, French vanillas, and virtual French vanillas, and there's the resulting analysis. Well, "analysis." I really do hope that this was helping in some way to you.
Oh! I thought of something. There's no study of vanillas in each of the five colors. But, if that were my goal, I might've jumped off a bridge by now. Just divvy up your vanillas as evenly as you can, and what's remaining, you could keep in mind for when you're squeezed for a spell-like effect or something, and you could make use of an ETB trigger virtual vanilla slot.
P.S. In case you don't know, I've started up a Great Designer Search 2 page that accounts for my design work related to GDS2. The first finalist challenge that was assigned to each of the eight finalists was also completed by me in a mock design challenge submission. I'll be "playing from home" along with the rest of the finalists for several reasons and posting them to the Wiki as well as most likely discussing related things on this blog.