Wednesday, November 14, 2007

The Lowdown

Also wanted to make a quick note: Devin Low (the Magic Director of Development) has been putting up some articles that totally kick my ass.

In the past, I never found the Development articles to contain enough technical nitty-gritty to help me out. But Devin's articles have really been an eye-opener. Especially this one - I have been waiting a long time for that article.

These articles gave me the nudge to figure out the problem I described in the last couple posts (why weren't my mechanics showing up in draft??)

Thanks Devin!

Planning Mechanics

I love it when a good plan comes together; I just wish I'd had one a few months ago!

I'm going to reiterate a bit of previous stuff to show why seeding mechanics is so difficult if you don’t have a deliberate strategy.

Previously, I mentioned that I had a ridiculously low quantity of common cards that play into the “token” theme, yet thought I was right on the money. How could I have been so off-track??

If you look at the roadmap for mechanics I gave in the last article, you’ll see that my new target is 38 cards that are part of the “tokens” theme. I also said that I figure roughly 60% of the cards in a theme should be at common. (In this case, I’m actually going with 55% at common.)
Tokens: 38 cards
Target: 55% of these at common
Total target for common cards in theme: 21 cards

How many did I actually have at common?
18 cards.

18 cards instead of 21 – not bad for having worked without a plan, right? And yet something went wrong – we never had enough tokens in our games to use the cool effects that relate to tokens…And I had to dig a bit deeper to find out why:
Tokens theme: 18 cards (yay!)
Cards that generate tokens: 5 (doh!)
Cards that punish tokens: 8 (double-doh!)
Cards that leverage/reward tokens: 5

So my roadmap has to contain more detail than just “qty of cards with mechanic X”. I actually have to identify – for each mechanic – how many cards have the mechanic; how many punish/prevent it; and how many reward/leverage it. And figure out (through trial and error) what makes for a good balance in these categories (generate/punish/reward).

And that’s the real subtlety (or at least it was too subtle for me to realize intuitively):It isn't good enough to say "There are 18 cards in the token theme at common", because it matters what those cards are doing with the theme.

Instead of having 13 common cards that generate tokens, I had only 5! Not only that: instead of a mere 3 that punish (i.e. destroy) tokens, I had 8! Ouch!

I hope this is of some use to somebody out there. I’ve sunk hundreds of hours into my set already, and now I have a bunch more ahead of me to shape it into what I really want it to be. If I’d planned better up front, I would be much further along now.

(Having said all that, I had to go through a certain amount of brainstorming-type design simply to hit on mechanics/themes that I liked enough to build a set around. But I should have switched to “planning” mode way earlier on…)

Tuesday, November 13, 2007

Seeding Mechanics

In the previous post, I showed how poorly I planned the "seeding" of my mechanics throughout the set. To resolve this problem, I did some Gathering and came up with the following list of "card counts by mechanic" from a few sets:

Bushido (CHK 17/291)
Splice (CHK 15/291)
Arcane (CHK 53/291) (It's actually 68 but I subtracted the number of Splice cards)
Soulshift (CHK 12/291)
Flip cards (CHK 10/291)

Morph (ONS 47/335)
Cycling (ONS 31/335)
"creature type" (ONS 34/335)
Zombie (ONS 25/335)
Soldier (ONS 24/335)
Cleric (ONS 30/335)
Goblin (ONS 23/335)
Elf (ONS 20/335)
Beast (ONS 36/335)

Threshold (OD 44/335)
Flashback (OD 30/335)
Discard (OD 47/335)
Sacrifice (OD 61/335)

Summarizing and percent-izing:
Morph 14%
Cycling 9%
"creature type" 10%
Zombie or Elf or Goblin or etc ~7%
Threshold 13%
Flashback 9%

(I think the CHK counts are so low because a ton of space in that set went to Legend stuff, but it would be a little more time consuming to come up with useful numbers there. In any case, I'm ignoring them as a result.)

I've come up with the following guidelines for myself:
  1. Most of the cards that have or cause a particular mechanic will be in common; say 60%
  2. Most of the cards the punish or leverage a mechanic will be in uncommon; say 25%
  3. Any cards that push the envelope on power level and/or are relatively complex will be in rare; say 15%

And here is my order of importance for the mechanics in my set.
  1. Resonate
  2. Tokens
  3. "CMC matters" (CMC = Converted Mana Cost)
  4. Debris (lands as a resource beyond mana generation...)
  5. Mindfly (if this is revealed from the top of your library...)
  6. +1/+1 Counters

Putting the above together with the fact that I want a 289-card set, I've come up with the following roadmap for "seeding" mechanics:

...And off we go!!!

Theme, or lack thereof

So despite the debacle of the last playtest (organizationally speaking), the guys had fun and generated a ton of useful feedback.

One thing that was interesting to me is that (despite my claiming to have solved this problem) there still were nowhere near enough cards at common to allow players to really use the different themes at work in the set...

So I did some counting...

Cards that generate Tokens
Common: 5/115
Uncommon: 15/85
Rare: 11/65

Cards that punish Tokens
Common: 8/115
Uncommon: 6/85
Rare: 6/65


So, 4% of my commons generate tokens (the 2nd most important "mechanic" in the set), and yet I walked into the playtest thinking the idea was well-represented!!!

Suffice to say, I'm about to count up the other themes as well, and have some major work to do before print #4...


Worst Executive Decision Thus Far...

Darn, I never wanted to have a 6-week delay between updates, especially during set development! But the reality is that it's taken every ounce of my free time to keep pushing the set through design, never mind chronicling the whole process.

But here's a quick summary:
  1. I did lots of design work to get the planned 289 cards into the set.
  2. Printed all cards for use in playtest 3. (I was editing the set right up until the print-shop store closed; I snuck in with 2 mins to spare the night before the playtest.)
  3. After printing, but before playing, I found a bunch of issues in the set (too many uncommon/rares, not enough commons; etc.)
  4. Corrected as many errors as possible (staying up til 4AM to do so; this comes at a great cost to me as I have a toddler and a newborn!)
  5. Reprinted about 30% of the set (printing 3.5)
  6. Tried to have playtest 3...

Unfortunately, I made a huge strategic error: I decided we should try to reuse whatever cards we could from existing printruns 2 and 3. I figured we could simply sift through the existing cards from print 2, swap in any card that was changed in 3, then swap in any card that was changed in 3.5 And by "existing cards" I mean "little squares of printer paper cut up into 600 Magic card-shaped pieces and printed in Draft quality".

I totally underestimated how time-consuming and confusing this process would be, and absolutely failed at organizing it. This error cost us 2 of our 5 playtesting hours and most of my playtesters' sanity...And then I realized that this process hadn't accounted for any cards that had been deleted at any time after print 2 or 3 (i.e. more than 80 cards...)

When we finally did get to draft, it was with a horrible abomination of a set that included several versions of many cards. If Magic sets were movie scenes, then this was the transporter-malfunction from Star Trek 1 (or possibly a post-accident Goldblum in The Fly...)

The worst in-game moment was when Luke played a printrun2 Advanced Mindfly common at 3U for 2/2 Flying with an ability, and then on my next turn I played a printrun3.5 Advanced Mindfly rare at 3U for 2/3 Flying with the same ability :(

The one positive note is that despite all of the above, the team claims to have enjoyed themselves :) Thanks guys!!!!! (Mike N, Mike P, Luke, Trevor)