Tri-Color cards - A Practical Experiment

Background

    Recently I was inspired by this post by Channel Fireball's Jason Waddell to give tri-color cards another attempt in my cube.  Like many cube designers I had slowly eliminated all tri-color cards from my cube over the past several years as I tuned and refined my list, cutting the cards that were commonly picked last.  Still, I was inspired by Jason's ideas on how to increase the chances of cards being relevant, and with all the sweet new goodies from Khans block I figured I'd give it a shot.
    I'll be assuming you've read Jason's article in my comments below.

Tri-Color Options

    The first thing I needed to decide was which 5 Tri-Color factions I wanted to support.  I'm no mathematician, but I believe I figured out all of the available options.  Aside from the obvious options of playing just the five shards or the five wedges, there are ten other configurations available. 

    I started off with the desire to play both Temur and Grixis, because I wanted to include Nicol Bolas, Planeswalker, and Maelstrom Wanderer.  There are only two combinations in the list above that include both of those factions.  I settled on the version in the top left: Bant, Grixis, Abzan, Temur, and Mardu.

Cube Construction

    My original goal was to play two cards in each tri-color pair, and cut two cards from each relevant guild in order to make them fit.  I never was able to settle on a second Mardu card I thought would actually draw people into three colors, though, and so eventually I just decided to only play one Mardu card and only cut one card from Boros.  In doing so, I ended up with the following multicolor configuration.

    In an attempt to further increase the incentives to play the supported three color combinations, I also included the uncommon tri-lands for each faction (although I decided to use Murmuring Bosk for Abzan).

Drafting and Gameplay

    I did two ten person drafts with the cube in this configuration.  I didn't take specific notes for the first draft but our consensus in discussion afterwards was that the tri-color cards were fun, but there was very little noticeable incentive to choose a faction.  In a 465 card cube, 10 cards here or there simply wasn't enough to shift the draft strategies one way or another.  Roughly half of the three color drafters in that draft were in supported factions, which means the support provided had no real effect on the draft.  In the second draft, we saw the exact same results.  There were only four tri-color drafters.  Two of them were in the supported factions, two were not.
    In both drafts, however, tri-color cards saw play.  People splashed to include them, people wanted to play with them, and they made space to fit them in, even the ones that arguably weren't the best for their strategy, because they were fun and interesting.  The consensus among the people I spoke with was that they enjoyed having the three color cards, especially the splashy ones that did unique things they couldn't get elsewhere, but that they didn't think all the extra stuff I did made a difference.  Before the first draft I made sure everyone knew which factions were supported, and everyone agreed that once the draft started, they promptly forgot which ones I had mentioned and it never seemed to matter.

Conclusions

    I've drawn two separate conclusions from the drafts we did with the cube in this configuration.

    If you want to support tri-color drafting in the way Jason suggests, you have to go further than I did.  Eliminate off color guilds entirely, instead of just backing off the support of them.  Actually decrease the available manafixing for those guilds and increase the fixing for the guilds you want to support.  Make enough of a change that it resonates throughout your cube.  Changing 14 cards in a 465 card cube simply wasn't enough to make a noticeable difference when you were looking at a pack during a draft.  The downside to this approach is obvious (and Jason mentions it in his article): The tri-color cards in existence just aren't deep enough to make it worth it.  You have to really warp your cube for the support make a difference, and when you do that you've made a huge change to your cube just so that someone can draft Siege Rhino.  Until we have better tri-color cards available to us, I suspect this approach won't catch on.

    The more interesting conclusion, to me at least, is the natural followup to the first one.  If changing 14 cards in an intentional effort to warp the cube in the direction of specific tri-color factions didn't make a noticeable impact on color balance, why not just play the few tri-color cards that are actually worth cubing with in a separate multicolor section, and forget about trying to balance it all?  There are few enough of them that you'd want to draft anyway that it simply won't matter.  Based on this experience, I have trouble believing that including only Nicol Bolas, Planeswalker and Maelstrom Wanderer in my cube would have any real impact on Izzet as a guild during a draft.  Two cards one way or another won't skew the draft to the extent that colors or guilds are any easier or harder to draft.

    I know some cube designers have long advocated unbalanced multicolor sections, and I'm coming around to their way of thinking.  The guilds are all deep enough at this point that I won't be unbalancing them anytime soon, but the next time I make changes to my cube, I expect to include a few of my favorite tri-color cards, and I won't worry about the color balance when I do so.