Automated Commander Power Level Evaluation

[pokken]

Added power and toughness filter support for such beauties as this:

{"card":["Recruiter of the Guard"]}	{"type":["creature"],"toughness":["<3"]}	1	0	
{"card":["Imperial Recruiter"]}	{"type":["creature"],"power":["<3"]}	1	0	

[pokken]

Just a note that I've updated the magic card database for this, although none of the cards are added to the rules data yet.

If anyone has any interest in doing some work on the rules set for this please let me know. I'm happy to update it as needed, it's pretty easy to do.

[3drinks]

I see this thread is still active months latre, how has the data been further assessed, for lack of a better word. Please, spare me the technical jargon I won't get.

No idea how to use this, but feel free to run my Kari Zev through this. I suspect it should rank high power, though I have the cards to add in at least Dualcaster/Twinflame if my meta called for such a strategy.

edit: made some updates to the curve multiplier that changed the end total slightly.

+--------------------------------------------------+
|           3Drinks_Kairi-Sane -- 272.42           |
+-----------------------------+-----------+--------+
| Card Name                   | Commander | Points |
+-----------------------------+-----------+--------+
| KARI ZEV, SKYSHIP RAIDER    | Yes       |      0 |
| MANA CRYPT                  | No        |    100 |
| SOL RING                    | No        |     85 |
| ANCIENT TOMB                | No        |     20 |
| PURPHOROS, GOD OF THE FORGE | No        |     15 |
| SKULLCLAMP                  | No        |     12 |
| SCALDING TARN               | No        |     10 |
| ARID MESA                   | No        |      5 |
| BLOODSTAINED MIRE           | No        |      5 |
| WOODED FOOTHILLS            | No        |      5 |
+-----------------------------+-----------+--------+

Scores in the 250s usually mean you're playing some good mana acceleration. Looking at your deck it's likely that the payoffs are not being assessed quite correctly since it looks fairly strong to me.

Cards like dreadhorde arcanist and sword of feast and famine need to have points eventually, and I have not gotten to categorizing the mid power cards yet. Hopefully get some voluneers soon

[pokken]

I see this thread is still active months latre, how has the data been further assessed, for lack of a better word. Please, spare me the technical jargon I won't get.

I'm out of gas on making up point rules after exhausting most of my knowledge - I might try going through a few of my old decks and creating some rules at some point if I get bored.

So it's out there but it needs some crowdsourcing of the data to be useful

[Inkeyes22]

I see this thread is still active months latre, how has the data been further assessed, for lack of a better word. Please, spare me the technical jargon I won't get.

I'm out of gas on making up point rules after exhausting most of my knowledge - I might try going through a few of my old decks and creating some rules at some point if I get bored.

So it's out there but it needs some crowdsourcing of the data to be useful

This is something that is very interesting to me, I did play with word press and Dreamweaver a little bit but would be practically useless on the technical stuff. But I would be happy to share opinions on points/power level debates.

[pokken]

This is something that is very interesting to me, I did play with word press and Dreamweaver a little bit but would be practically useless on the technical stuff. But I would be happy to share opinions on points/power level debates.

You can download the TSV file here:



I believe @Rumpy5897 wrote a little blurb somewhere in here on how to write synergies. The format is important.

The key concepts of synergies are:

1) All card-only synergies go in the leftmost column (e.g. rings of brighthearth + basalt monolith)
2) Generally aanything can synergize with anything, there're demos of examples e.g. Maelstrom Wanderer with CMC > 3 and <8 and tag:ramp with tag:advantage and tag:selection with itself
3) All card synergies should be tagged as to what they are, there are examples of everything. e.g. Protean Hulk is a tutor. Obviously it's banned but I didn't see any sense in removing it yet

[Inkeyes22]

Maybe it is silly, but I find it odd that Elves of Deep Shadow is the same point value as Birds of Paradise. Don't get be wrong I think the power level is close, but if BoP is a 15, Elves are probably a 12 -14. Only producing B and losing 1 life isn't a major draw back, but being able to attack for 1 or kill an X/1 doesn't really make up for not being able to make W, U, R, G in my opinion. Avacyn's Pilgrim/Llanowar Elves being a 15 kinda makes sense-ish, but not flying makes me want to knock off at least a point there too.

Also the Talismans not being the same kinda makes sense, but I think the non-G ones would probably be better than the G ones.

[pokken]

Maybe it is silly, but I find it odd that Elves of Deep Shadow is the same point value as Birds of Paradise. Don't get be wrong I think the power level is close, but if BoP is a 15, Elves are probably a 12 -14. Only producing B and losing 1 life isn't a major draw back, but being able to attack for 1 or kill an X/1 doesn't really make up for not being able to make W, U, R, G in my opinion. Avacyn's Pilgrim/Llanowar Elves being a 15 kinda makes sense-ish, but not flying makes me want to knock off at least a point there too.

Also the Talismans not being the same kinda makes sense, but I think the non-G ones would probably be better than the G ones.

There is *TONS* of quibbling that can and should be done. This doc represents a day of my work and I don't know how much effort from Rumpy, but probably under a day I basically am throwing this at the wall.

I believe I pointed the talismans that way because blue tends to be overpowered with artifacts, but non-green is as valid an argument if not moreso than blue as the determining power. My feeling generally is that blue talismans will more correlate with deck power but that was just a gut instinct.

Elves of Deep Shadow was another one there where I was like...most people do not play this card, and it's probably indicative of a high power level. Again gut instinct. I think you're correct that it's worse than birds. I think I was just trying to be clever in sussing out higher power decks by rarely played but very powerful cards. Since I started this the card has had a bit of a renaissance though

This type of stuff is what needs to happen. My fervent hope is someone will take this and just project manage the hell out of it, go get every major deck synergyset in there and get all the individually powerful cards pointed, and polish off the generic synergies (e.g. selection + selection is a synergy that will capture a *ton* of power level if all the cards are tagged correctly with selection).

There really cannot be more than maybe 1500 cards that are individually powerful enough to deserve points *at the most*. So it's definitely achievable.

[Inkeyes22]

Should I make a copy and edit, as it looks like if I make changes it would affect things. I know there would be a changelog etc. but I would rather not cause more work.

Edit: it feels like ramp has a lot of the points. Mana Crypt is worth 100, Imperial Seal 25 and Purphoros, God of the Forge 15? But then Captain Sisay 125? I would say all four of those cards are closer in power level. I think they should all be in the 20-40 range.

Another Edit: Flash is no more

[pokken]

Should I make a copy and edit, as it looks like if I make changes it would affect things. I know there would be a changelog etc. but I would rather not cause more work.

Edit: it feels like ramp has a lot of the points. Mana Crypt is worth 100, Imperial Seal 25 and Purphoros, God of the Forge 15? But then Captain Sisay 125? I would say all four of those cards are closer in power level. I think they should all be in the 20-40 range.

Another Edit: Flash is no more

If you're going to do edits and own them make a new version. If you want the git library I think it's in the first page and you can check it out and send me a pull request if you want to post a new TSV -- or you can just let me know and link it.

Just my experience talking but I always find fast mana to be the cause of more broken games than tutors; it's possible imperial seal should go up, but I'd argue that Mana Crypt is the individual most powerful card currently legal in commander - barring potential arguments about Demonic Tutor. The percentage of the time people demonic tutor for Mana Crypt or Sol Ring is high enough that I think the net positive fast mana is the best stuff. Generally speaking for tutors to enable a broken game you either need to be tutoring for fast mana, or already have fast mana to enable casting the stuff you tutor for.

Captain Sisay is a special case in that it's 125 *as a commander* and I think the points are just in the wrong column. Data entry error I expect the non-commander points should prob be like 10 or something. if that.

(Re: banned stuff - I see no reason to remove it from the dataset in case someone wants to compare historical decklists for some reason).\

---

Note, one thing to keep in mind is that we added an oracle text parser that's very limited, so you can (and should) do stuff like {text:["shuffle"]} over tags where you can. There are a number of good examples of that in there, e.g. with brago enters/leaves the battlefield. It's pretty rudimentary but it does a decent job of approximating.

For example -

{"card":["melira, sylvok outcast"]}	{"text":["persist"]}	8	0	combo
{"card":["vizier of remedies"]}	{"text":["persist"]}	5	0	combo
{"card":["rhythm of the wild"]}	{"text":["persist"]}	4	0	combo
{"card":["metallic mimic"]}	{"text":["persist"]}	5	0	combo

---

And philosophically speaking, correctly capturing what the synergy is should be the focus I think, over revising points. Even if you had like 2000-3000 synergies you could easily review the point numbers in a couple days and look at gaps and issues by putting decks in.

Spending a bunch of time wound around the axle about whether Talisman of Xyz is 2 points or 3 points is largely a waste, since missing 50 points from an undocumented synergy is a far bigger source of error.

Hope that is helpful. My recommendation would be to capture things that are missing before making changes to points, so you can have a closer to holistic view to help accurately pointing.

[pokken]

A hopefully useful way to illustrate how much more impactful missing synergies are than point errors:

{"card":["Sylvan Library"]}		20	0	advantage, draw, selection
{"card":["Sylvan Library"]}	{"text":["shuffle"]}	1	0

Suppose I pointed Sylvan Library at 25 instead of 20 - this is a 5 point error. You could surely argue for it to be 25 based on Wheel of Fortune being 25 if you really wanted to, and you could surely argue that in a vacuum it's more powerful than Windfall because Windfall (like Wheel) depends on some setup and is fairly situational whether it's good.

You could even argue that Library is, in its own way, closer to the point total of Mystic Remora at 30 points. I would entertain that argument.

However, if you have 20 shuffle effects in your deck now Library is a 40 point synergy vs. Even if you're significantly off (50% in your point total) it doesn't make half as much difference as neglecting a synergy.

That kinda segues into Wheel of Fortune - Wheel of Fortune + Notion Thief effects is a powerful set of synergies that're mossing. Wheel of Fortune + Reanimate and Dread Return effects are probably also missing.

So if I spend a lot of time wrangling Wheel's points to say, oh, really, it's 28 not 25, it won't make near as much difference as capturing its synergies appropriately.

Ultimately, I believe most single card point values should go away and only the absolutely most powerful cards would have point values.

Even Mana Crypt could ultimately have its points replaced with a synergy with [tag:mana rock, cmc<3] if you wanted to get really fancy. Mana Crypt is *way better* when it's powering out other ramp spells. Or you could say Mana Crypt with [tag:ramp] as a shorthand way of illustrating that mana crypt amplifies the power of your ramp spells.

There're a ton of angles and ways to think about that, but ultimately I think it is fine shorthand to have points for single super powerful cards.

But it's probably not right to have 10 points for Talisman of Conviction in a vacuum - just playing that card by itself doesn't really do much to your deck's power. Playing critical mass of 2 cmc ramp is way more impactful.

Sorry, this is very very complicated

Putting key cards in and tagging them appropriately even if they have 0 points is definitely a big powerup to the dataset.

To see that, look at this synergy that has an amplifying effect based on volume of ramp spells:

{"tag":["ramp"]}	{"tag":["ramp"]}	0.75	0	

That synergy is really only as good as the set of cards tagged with Ramp. But on the flipside you can noise the data up by putting too many things in there I guess. Hence my recommendation to focus on actually playable cards for the most part.