Hi, I've been working on the last few months (currently on pause) in a tool to rank commanders in an objective way, analyzing all the cards the deck has and so...
The thing is, I think I'll have to rework it a litle sjnce ai feel it's lacking accuracy. So far I'm considering only 3 Deck archetypes: Aggro, Control, and Combo. Do you think that's okay? or Should I include more types?
Also, another part of the tool is the ability to compare two cards and see which is better to fit your deck. This is to help to decide which cards remove and add when new sets come out... What do you usually consider for removing a card from your deck in order to add a new one?

EDHRec attempts to do this also, that's a good place to look for inspiration. They use deck construction statistics. Episode 239 of the Command Zone podcasts studied some games posted on youtube, so that might give you ideas for your technique.

Before you go too far down the rabbit hole you should understand that "objective ranking" in Magic the Gathering is functionally impossible. Commander adds a second layer of impossibility because, as a multiplayer format, perception of power or "threat assessment" of the individual players has a larger role in determining who wins than deck construction does.

Any choice between cards comes down to a subjective analysis. Some cards are better than others. Most are just different, better in some situations but worse in different ones. Assessing how often each situation comes up, how likely it will matter, how much better a card is in that situation, all of those things are highly subjective interpretations.

Playing with a program that helps optimizing decks can be fun. You'll get some decent suggestions here. But basically everyone on this forum will feel compelled to tell you "There is no way to objectively rank decks or optimize card choice in the community-scale Commander Metagame"

I can't overstate how much the threat assessment/perceived power effect is, especially in multiplayer. Key example: Pheldagriff is becoming a pretty popular head of your standard Bant control deck. While it is a rather efficient beater, you could argue that an incredibly similar deck with Derevi, Jenara, Rafiq, or even modern Arcades, are objectively more powerful. So why Pheldagriff? Because even though he is in practive a 4/4 flying trample pseudo hexproof for 4, he looks like a group hug card and therefore not really much of a threat. And as the Trojan river horse becomes more popular people might start to catch on, making it less and less useful.

There is no actual objective tier list. That's sortof impossible, because most Commanders aren't particularly well-tread territory. That said, you can find the closest thing to a consensus (the cEDH tier list) here. Highly recommend treading lightly with that information, though- as cEDH has very specific decorum, and mindset. Most of the information is purely speculative within their community, and there has been shifting in those ranks at a fairly constant pace. The generals are based on the power of established builds in a competitive environment. They get innovated pretty regularly, and their aggregate data is lacking, since it's based on experience instead of sourcing concept decks from the web.

BUT WAIT, THERE'S MORE.
If you want a functional model, which condenses available cEDH data, here goes...


There's no reason to have any more tiers than that. If someone can think of a clearer razor for a lower resolution summary that is closer to objective truth- good on you, homie. If not, I still won't call it objectively true, but hopefully clear enough to be a good model.

Compare in what way? Only two cards? A feature I would find more useful would be something that suggests similar cards/effects.

For example: I need something similar to Rites of flourishing in WUB. If I select a card and the color identity, it could return similar effects - in this case, cards that allow you to play an extra land (like Storm Cauldron) and cards that draw additional cards (like Righteous Authority)



If a card in a new set comes out that I am interested in testing, I will generally start by finding the other cards that fill a similar roll then comparing them side by side to decide which one(s) stay or go. For creatures, it may not be based on the role, I may compare by CMC (e.g. Kazarov replaced Chancellor of the Dross in my UBR Garza Zol vampire tribal) to keep the curve (if that deck cares about that).

That would be tough. But if you get good at using a card search engine and have a sense of how cards can be worded, you can usually find what you're looking for without too much else to sort through. I use Scryfall because it has a lot of features, is easy to use for simple things, and has working Boolean logic.

For instance:
https://scryfall.com/search?q=f%3Acommander+id%3Awub+is%3Apermanent+%28o%3A%22additional+land%22+or+%28o%3Aput+o%3A%22land+card%22+o%3Ahand+o%3A%22onto+the+battlefield%22%29%29&unique=cards&as=grid&order=name

Yeah, I already use MTGO for these kinds of comparisons. I was just remarking that if Chiky is trying to do a 1:1 card comparison, something like this would probably be more useful. Though it depends on what criteria is being used for the card comparisons.

Thank you for the feed back and answers.

For the engine searching I think I'll have to pass. I'm working on Microsoft Excel and I'm not good enough to add stuff like that...
The idea of comparing is putting both cards' stats, and the tool then compares both with your whole deck stats to see which one gives a better "score."

I think I'll rework and redo the whole thing to fix the scoring and sort the results by a numeric score (not defined yet) and a tier. The idea is to rank from jank (already made some commanders for this using a random legendary creature, random ammount of corresponding basic lands, and filling the rest with cards taking from Random Card in scryfall that matches the color identity) to competitive commanders.

I guess I'll keep you posted. For now, I want to ask some things, so I'll be asking here in this post. First, let's start with something relatively easy:

What do you consider to be the speed of a deck? Which things make a deck fast and which make it slow?