A Formula for Determining your Deck's Power Level

[fire5167]

There's something interesting about the numbers in this formula. Kinnan scored 99. It's the winconcount-term that gives immense power to combo decks with a few lines and tons of tutors. In Kinnan's case it's (10(17+2)) / ((2-3)^2 +1) = 190 / 2 = 95. That needs some adjusting.

I think you’re right. I never incorporated the number of tutors into the denominator, so it’s just increasing the value of the term linearly. It might work to add the TutorCount to the +1 on the bottom, and then remove the  TutorCount from the 10 coefficient in the numerator.

 (17+10(2)) / ((2-3)^2 +17 + 1) = 37 / 19= 1.947

I’m not sure how that changes the total result but it should tether down the value a bit.

[Morganator 2.0]

Okay so I have a question that ties into this discussion. How would you rank the following things in terms of importance for rating deck power? Which are most important, which are least important?

Average CMC of deck
Combo potential
Draw power
Interaction
Land base
Mara ramp
Stax
Synergy
Tutors

[Schau]

I would say land base and mana ramp would be the top two followed by draw power.

[ApothecaryGeist]

Okay so I have a question that ties into this discussion. How would you rank the following things in terms of importance for rating deck power? Which are most important, which are least important?

Average CMC of deck
Combo potential
Draw power
Interaction
Land base
Mara ramp
Stax
Synergy
Tutors

ok.  So the mathematician in me is intrigued.  I've thought about this a bit this morning.  I attempted a mental exercise pitting (imaginary) decks together that each have 1 of categories but don't contain the rest.  I've ordered them and reordered them.  I think this is where I ultimately land:


1. Ramp with CMC <=2
2. Stax/Interaction **
3. Tutors


4. land base and color fixing ++
5. CMC
6. Synergy & Combo @@
7. Card Draw ##


8. Graveyard recursion $$
9. Ramp with CMC >=3 %%


**Stax generally prevents your opponents cards from functioning, while "interaction" generally just gets rid of them.  Subtle difference.  But not really a lot of difference between a Propaganda and a Damnation.  Also, is Counterspell stax or ineraction?  Some cards are hard to differentiate.  Both categories seem similar and of similar importance for power level.  So I combined them.


++ Land base refers to having an adequate land count (not too many or too few) as well as not having lands that ETB tapped.  It should also factor in color fixing.  This is color fixing generated by dual lands AND by the ramp.


@@ Maybe I'm overthinking, but as I thought more, the lines between synergies and combos began to blur for me.  They're both important.  I combined them.


## Card draw.  Only Quality card draw should count.  What is this?  I don't have a concise definition at this point.  It is mainly card draw that doesn't cost you other resources.  For instance, Commander's Sphere costs you not only that permanent, but the ramp and color fixing that it provided.  Commander's Sphere is not quality card draw.  It is consolation card draw.  Same for Sad Bot.


$$ Using the graveyard as a second hand is very important.  it's power level could fall somewhere between tutor and draw.  But while you can always draw cards, recursion relies on something being in the yard in the first place.  And something worth digging out.  So I ultimately rated it a bit lower.  Flashback spells, Animate Dead, and Muldrotha would all fall into this category.


%% The original equation did not count ramp with higher CMC - Cultivate, Kodama's Reach, Commander's Sphere, Black Market, Gilded Lotus.  Having these is certainly better (more powerful) than having no ramp at all.


^^ You'll also notice line breaks in my list.  These are intentional.  It breaks the list up into three groups.  Within each group, the power level of each component is almost the same.  I feel there is a much bigger break between the groups.  For instance, if you have 2 lands that ETB tapped, swapping out those for lands that enter untapped is slightly better than swapping out two other cards in your deck in favor of cards with a lower CMC.  But putting one more tutor in your deck is far more powerful than either of those options.

[WizardSpartan]

Land base
Tutors
Mana ramp
Combo potential
Draw power
Average CMC of deck
Interaction
Synergy
Stax

I value tutors just below mana base because having a lot of good tutors means you can find whatever you need (even ramp if necessary). I put average CMC so low just because if you have a good landbase, ramp, tutors, combo potential, and draw power, it's not a big deal if your average CMC is blisteringly low in my eyes.

Interaction, synergy, and stax are on the bottom because, ultimately, they aren't necessary if you can just win. Am I biased, as a green player whose favorite type of removal is player removal? Perhaps.

[Core]

I only did the calc for my Anowon, the ruined thief deck, and with no tutors I still got to a 10.7 due to high draw and interaction with low cmc, and that was the old list, I've added tutors and more interaction sense then.

[Xaarvaxus]

Okay so I have a question that ties into this discussion. How would you rank the following things in terms of importance for rating deck power? Which are most important, which are least important?

Average CMC of deck
Combo potential
Draw power
Interaction
Land base
Mara ramp
Stax
Synergy
Tutors

I feel strongly the first four should be ordered: Combo Potential/Tutors/Draw Power/Mana Ramp

My reasoning being I think its well established that combo decks are the strongest in the format and become stronger the more competitively you play to the point it seems its the only thing being played in cEDH, so its the most important point in determining power level.  The ability to draw into combo pieces therefore determines how effective the combo potential is with quality [tutors] ranking above quantity [Draw Power] with mana ramp being what helps determines how quickly one can execute finding the combo and firing it off.

Next: Stax/Synergy/CMC

Not as sure on the order here.  Stax is important to slow down everything in the above tier.  Synergy to determine how much additional power and effectiveness your pieces have when played in conjunction, especially synergies that reduce casting cost I think.  Avg CMC mostly for determining how quickly you can get to the point of firing off multiple spells per turn though I'm not sure if we're talking printed CMC which most sites go by or the effective CMC that deck builders assume [delve and such being a thing].  I could see a case for any of these three points going in any order of 5/6/7.

Last two: Interaction/Land Base

Again, I feel either could be 8 or 9 but think both come at the end.  Interaction seems a necessary evil in the highest octane decks in case it becomes urgent to stop something *right now* but if you're faster at executing your game plan than everyone else, you probably didn't fire off a single removal spell.  Land base is probably mostly about playing shit that doesn't come into play tapped and that enough of it produces colors you need.  All basics with the correct mana ramp package probably doesn't suffer very much.

[Aetherium Slinky]

How would you rank the following things in terms of importance for rating deck power?

1. Average CMC
2. Draw power

3. Mana ramp
4. Tutors

5. Land base
6. Synergy
7. Interaction
8. Strategies (incl. stax, recursion, combo etc)

In my opinion 80% a deck's power is determined by average CMC and draw. Those two together enable you to do everything you want to do and they're just essential to every deck. You cannot have a powerful deck without sufficient draw and low enough curve to cast a lot of things. Mana ramp serves an important role in pulling ahead and tutoring can add consistency to your plan. Tutoring takes up a lot of mana (usually) so ramping is necessary for tutoring to be effective. With a combo deck it's obviously a little bit more pronounced, maybe more important than ramp.

Finally you get the rest of the things and I'm not entirely sure about their order. Interaction I rate highly because I just happen to like a bit more control over the game but you could just as easily say that your main strategy is more important because you want to be the threat instead of stopping the threats. All the rest are really close to each other anyway. Would group them together in a formula. Land base is obviously a little important because it works with ramp but usually people know how to craft a decent land base and ramp fixes the rest of it so it's not as important as ramp. Haven't heard of colour screw stories in a long time. I had some problems with Mind over Matter in a three-colour deck but that's probably a very extreme example. Land base becomes way more important in multicolour goodstuff decks that *need* the colours - like Scion of the Ur-Dragon and that's the reason I rate it a little bit higher than overall strategy or interaction.

[Slyvester12]

MustaKotka makes a good point about colors and land bases. I would make land base directly proportional to number of colors. Mono color decks don't find it important, but four and five color decks consider it top three, probably.

[Morganator 2.0]

Alright, I think I have a first draft for my version of the power level calculator.

Goal

This formula is not supposed to give an accurate representation of power level, it's just an estimate. However it should get pretty close. A deck that gets a score of about 8 should be able to go toe-to-toe with an approximately 9 deck. Decks at 6 and 5 should be able to have fun with each other, and so on. The formula does this by using a combination of discrete variables and categorical variables. A discrete variable is something you can put a precise number on; average CMC of the deck, number of ramp cards, etc. A categorical variable is one that you need to give a self-rating for; for example, how well does the deck combo? How strong is your land base? Here's the formula:



Discrete variables
A: Average CMC of the deck. Technically this is a continuous variable, but whatever, it works all the same.
D: Number of draw spells in the deck.
I: Number of removal, boardwipes, counterspells, and protection spells (not stax).
R: Number of ramp cards in the deck.
S: Number of stax cards in the deck. Count mass land destruction as stax.
T: Number of tutors in the deck.

Categorical variables
I based these off of the powertable made by Judaspriester and dexflux, with slight modifications based on comments made in that thread. Each statement has a number linked to it. If that statement is true for your deck, you use that number in the formula.

C: Combo strength of deck
0 - no (infinite) combos.
1 - janky multiple card combo(s) without redundancy.
2 - janky multiple card combo(s) with redundancy, or a combo that is hard to setup.
3 - typical combo wins without redundancy and are harder to setup (Pili-Pala + Grand Architect).
4 - typical combo wins with redundancy and are easy to setup (e.g. Kiki-Jiki combos, Worldgorger Dragon).
5 - Strong combo wins that's easy to setup (e.g. Dramatic Scepter with commander as a outlet).

L: Strength of land base
1 - basics/taplands/utility lands.
2 - basics/taplands/few untapped nonbasics.
3 - basics/taplands/many untapped nonbasics.
4 - None, or next to no lands that enter tapped unconditionally.

Y: Non-combo synergy of cards
0 - essentially just a bunch of cards with no effects.
1 - the barest minimum of synergy - cards hardly interact with each other at all.
2 - useful synergies, but relatively harmless (e.g. blinking Mulldrifter).
3 - less harmless (e.g. repeated blinking of Mulldrifter, Deadeye Navigator synergies).
4 - dangerous (e.g. Brago blinking the board for five triggers or more per turn).
5 - extremely dangerous (e.g. Doubling Season into planeswalkers).

Justification for weights

Each variable has a weight attached to it. You're probably wondering how I got these numbers. Simple: I inserted the values for each of my decks in an excel sheet, and then played with the numbers until I got power level values that were close to what I would expect for those decks. That does mean that these values are largely arbitrary, and have a massive bias to them. To help fix this a little, I included some precon decks to give representation to the lower limit.



In case it matters, the order for most relevant to least relevant is:

Combo potential
Synergy
Average CMC
Tutors
Draw
Stax
Mana ramp
Interaction
Land base

Flaws with this formula

• You'll notice that individual card strength is not taken into account. A Swords to Plowshares contributes just as much as Assassin's Strike, even though we can all tell which of these cards is trash.
• Choice of commander has little effect on the discrete variables.
• Numbers are very much based on my decks, and a few precons. Has a heavy bias in this sense. Like, you'll notice that all my decks have a "4" rating for their land bases. I really hate lands that enter tapped.
• You'll have to use your judgement depending on your deck. Should every high-cost card in Jodah, Archmage Eternal just have a CMC of 5? Does everything that can sacrifice itself count as card draw for Korvold, Fae-Cursed King? It's up to you to decide.
• Graveyard synergies are not taken into account. Try to use your judgement on these things. If someone has a good way of incorporating graveyard strategies into this formula let me know about it.
• Color identity of the decks are not taken into account. Obviously mono-green is going to have a much easier time with color fixing than 5-color. Conversely, 5-color has a much wider selection of cards to use. I'm not sure how I can get the formula to take both of these into account.
• For some bloody reason Mishra still ranks really high. Seriously, it's a really crummy deck. The entire deck hinges on Possibility Storm and then luck. It's really not good. Might need to fix up the combo and synergy sections to get a more accurate rating.
• In line with this, the precons seem a little high, but that might just be because I'm used to people giving them ratings of 2 or 3. That and their card choices are rather weak, which this formula does not take into account.
• I only graded two stax decks, both of which are way outdated. Not exactly a good representation.

Next steps

This has been a fun weekend, but now it's over. Once again I've added a template attached to this post. It will calculate the power of your deck for you. If you feel like it, put your commanders into the equation, and see how their power comes out. I've included spots for 10 commanders to be inserted at once, but you can get more if you know how to copy and paste. Let me know if you have any feedback that you'd like to give about this.

[WizardSpartan]

What do we consider commanders for? Would Kumena, Tyrant of Orazca count as 2 instances of card draw? 3?

[Morganator 2.0]

I counted them as 1, but making them count as 2 cards is only a slight change in the result.

[WizardSpartan]

With that being said, I've attached my decks' results, including my Obuun deck which is basically the precon + a couple cards I had lying around.

The numbers seem reasonable. You already noted the reasons ratings might be off (the amounts are okay in Obuun, but card quality is poo poo, so that's why its only 1.5 points below Meren).

[Aetherium Slinky]

Is Rhystic Study stax or draw? Is Smothering Tithe stax or ramp? EDIT: Counted them by their utility value rather than as stax. I also counter Propaganda et al. as stax.


A couple of things jut out:

Kami Draw is definitely not that strong. It's the number of draw spells and mass bounce that just bloat the value by a lot. Most of the draw is either symmetrical or expensive so one could argue that the card quality is poor. The bounce spells are also relatively expensive to play (apart from Cyclonic Rift and Chain of Vapor) compared to what they achieve.

Sygg isn't a strong deck, it's barely at precon level. Card quality is extremely poor due to it being an ultra budget deck. It should score a 5.

Same goes for Dakkon, it's a voltron deck with a very fragile strategy. Card quality is incredibly good due to the deck being based on the Esper staples template but that doesn't make it a strong deck by any means.

Sooo my wincon-less Esper staples deck scored nearly 10. It's a template, not a deck.

Kinnan is at it again: dat value. It's a strong deck. Nice.

My Merieke would definitely lose against your Edric, Krenko and Ruric Thar, Morganator. It's a fairly fragile combo deck (except for Dramatic Scepter that's also in the deck) that dies to removal and it's also slow at that requiring a lot of things to go right. I gave it a little minus in terms of combo because Dramatic Scepter isn't usually the combo I win with but one that is definitely a tutorable combo but my Commander isn't a part of it. Ok, the deck has an obscene amount of interaction. I give it that. One could argue that it hits the ten if I added Tainted Pact or Demonic Consultation. As it stands Merieke should probably score an 8 - at least from what I've gathered from online plays it loses to the subjective "8s" and wins the subjective "7s"

Other than that I feel like the decks are in order. Stompy decks get the correct values. Atraxa Proliferate lands is also a very weak deck, I think 6 describes it pretty well. It just proliferates storage lands and finishes with a finite X-spell like Exsanguinate. Card quality is poor.

One should probably divide each category by the average CMC of the respective category. I know that requires a lot of manual counting but it would better reflect the card quality in a deck. My Sygg and Kami decks would get abysmal values as they should.

[Morganator 2.0]

Right, so it seems like I need to change a little bit with the formula, but the first draft did turn out pretty well. I mostly need to find a way to separate the stronger combo decks (Sisay, Kinnan) from the weaker ones (Mishra, Merieke). I also need to get the pre-cons down to a more expected level.

Not sure how I'm going to do this though. I know that first I need to find a way to differentiate between low quality cards and high quality cards. That was one of the stark contrasts between the pre-cons and player-made decks. If you have any ideas on how to tell the difference, let me know.

After that, I'm thinking about turning the interaction and stax sections into a categorical variable like "How well can you stop another deck?". 5 would be something like "I can consistently stop a combo on turn 3 or sooner", 4 would be "I can occasionally stop a turn 4 combo, but no issue with combos later on", and so forth.

Lastly, and this is something I should have done in the first draft, I'm going to make CMC a negative regression, as deck power tends to go down with a higher CMC.