MTG Salvation

How do you determine the skeleton of your deck? Like, how many creatures, how many removal, how many support? These ratios will differ from deck type to deck type but do you guys have any method on establishing these given a certain deck?

My skeleton is very basic. Commander slot, 36 lands and the rest takes care of itself.

As you said, different deck types demand different things. As such, I don't set an exact number on any one type of card. When I make a deck and fill out my 62 card slots, I do research. I look for cards that would/could go with the deck and a start to form a selection on which I would deem 'must have' and those that I like. Once the deck is assembled, I play it and test it out over a period of time, usually a few weeks. If I've deemed the deck worth keeping, I begin to modify it based on my experiences. I take out the cards I found myself never using and replace them with cards I believe will be more useful. As I do that, I also begin to collect more cards for the deck so that I may continue to modify it over time.

As such, the number of cards I have for a specific type or function change, sometimes drastically. As such, I don't believe in really pinning down the number of cards I think I'll need. I just let my deck evolve organically through play and whatnot.

This is a really, really rough guide I start with. Each of these is highly variable, but it makes sure that I don't get so focused on the Main Plain that I forget to consider just as carefully my ramp, card draw, removal, and manabase.

• Lands: 36-38. Always carefully plan out colors based on your cards. Consider not only your utility lands and color fixers but also your basics count just as seriously as you considered card choices for the rest of the deck.
• Ramp: 10. This varies wildly on count and type (manarock vs land ramp) based on the colors and the deck.
• Card Draw: 10. Don't get stuck top-decking. Running out of gas is a problem for half the friends I play with (or at least half their decks) and is probably the biggest factor in falling behind. Card advantage is not the same thing as draw, but it's also a big factor.
• These above two, I don't actually run 10 of each. I'll often run less ramp and draw than 10, but starting at 10 slots make sure you have to carefully weigh that slot for card draw against the card you want to run instead.
• Removal: 10. I also vary a lot here. I'm not sure what my average is, but it's probably lower. I like to have a couple wipes that work with the deck's Main Plan and then some varied spot removal/answers.
• Main Plan: 20-30. Whatever the main thing your deck wants to do to win should have a large number of cards devoted to it. This will also sometimes skew the other slots. Notable examples: Edric decks often run 30 or more 1- and 2-drop weenies. Daxos decks often run 25-30 enchantments. Decks that want to win by turning nontoken creatures sideways will often have around 30 creature cards. Jolrael or Kamahl would skew Ramp slots higher.

There are always exceptions to these numbers, of course. I've run a monored combo deck that had few slots for ramp and no removal because it was trying to go off turn 5. It actually had fewer card draw options, too, but they were more explosive. I've run a politics/control deck with 45 pieces of spot removal — but if Removal is the Main Plan, then the numbers are still pretty close to the rough guide.

I also try to reserve a few spots if possible for non-removal answers or speed-bumps, but these are often dependent on the decks I face the most and how much the answers/bumps interfere with the Main Plan. Things like City of Solitude, Song of the Dryads, Rule of Law, Nevermore, Rest in Peace, Null Rod, Omen Machine, Arcane Lighthouse, etc. At least 1 piece of targeted land destruction and 1 piece of graveyard removal are staples for me and I don't usually count them among my normal Removal category.

Every deck is different, there's no generic template. Usually after I have an idea for the deck, I first dig through my collection to find any cards that seem useful. Then I decide which functions will be important and allocate an appropriate number to each category, and divvy the cards into those categories. Then I pick the best x cards in each category, moving some numbers around if one category seems to have too many or two few good cards.

Find a commander who could pique my curiosity, look for a potential theme to go with it, then find cards that fit into that theme. Numbers and ratios come after I get the flavor cards in first, because I prefer winning with style.

For example, my Bolas and Three Gods deck would obviously include Nicol Bolas, three fallen Amonkhet gods, and his PWs. Then, I find cards that would benefit the flavor cards, preferably with high synergy between them, like Windfall and Animate Dead, Everlasting Torment + Goblin Bombardment, etc. In the end I usually have excessive amount of cards, that's when I start shaving for the best options and swapping for removal cards WITHOUT deleting the flavor.

The most beautiful deck you can make is where every single card helps directly with your mission. I hit a Pax Romana back in 2014 of this effect due to the nature of which decks I was building.
Let's take Trostani, Selesnya's Voice as a commander:
Forests, Plains, dual lands all help play the cards you need, so they are good.
I need removal! I use Trostani's Judgement and Sundering Growth! Not Naturalize.
I need protection! I use Rootborn Defenses and Druid's Deliverance! Not Fog.
With commander the more you can get your deck to harmonize the better it will perform!
BUT:
You need about nine cards in your deck dedicated to dealing with your meta. If you lost to Insurrection the last four games, you need to run Homeward Path. Too many indestructibles? You might want a Pacifism (or Jabari's Influence).
These cards won't harmonize with your deck's them, but think of them as calluses. Calluses are seen as an ugly waste in the world of beauty, but in the working and survival world they are highly practical augmenting your pain endurance, how hard you can hit for how long, and making the difference between a bleed, a bruise, and a I-barely-felt-that.

I don't play many skeletons. Reassembling Skeleton was okay but there are usually better options.

The Command Zone podcast, as well as many article written over the years going back to 2007 when there wasn't even a formal sub forum here and commander was in the Casual sub forum, the ratios of commander have been discussed at length. As you and others have said, specific decks will change based on the meta and need, but there are some core numbers most of us agree with.

There was a great article on the WotC forums, which have since been taken down, that broke down the math of commander decks. The author had three categories:

*Signets and other 2 casting cost ramp like Rampant Growth
*Concentrate effects: Cards which allow you to draw multiple cards at once
*Action spells: Every other spell, be it a creature, removal, whatever.

The author broke down the math of the format and compared it to a "control" deck. Not like agro/combo/control, but in the scientific sense, as in a "control group." The "control deck" had no ramp and no Concentrate effects. It was 44 lands and 65 action spells. The test deck had 33 lands, 11 2cc ramp spells, and 14 Concentrates, and 41 "action spells."

The author showed how the deck with ramp and card draw would, over the course of a game which lasts 8-12 turns, would see significantly more actions spells per game and have access to more mana for said action spells in addition to the mana spent on ramp and card draw than that of the control deck. Furthermore, the deck with ramp and card draw would have to mulligan less often, suffered far less mana screw or mana flood, and was just opperated smother all around.

The general conclusion was that far too many decks run far too little ramp and card draw than they should and suffer for it.

I apply this to my deck building and see the results first hand. Sure, there are very fast decks out there which try to end games on turn four. If that is your meta, then you shouldn't even be in this thread, as you are in a hyper competitive environment and don't need help tuning a "normal" deck. However, if you find yourself playing games which got to turns 8-12 or higher, then these are the numbers you are looking for.

==============================================================

The Command Zone, a podcast for Commander, also talks about the raw numbers of the format and says you want about 8-10 copies of any effect your deck is looking for. Keep in mind that some cards overlap, so try to be as efficient with your effects as possible.

Depends on deck, but usually 37 lands.