Improving on Vanilla
Features
JWiley129
1 March 2016
3565 views
1 March 2016
3565 views
Hello all!
My name is JWiley129 and you might recognize me around these here parts. I don't post many decks, but I enjoy engaging in discussions all over the site. Here, I am going to talk about an idea that I've had brewing in my head for some time now about a favorite format of mine, Limited. So bear with me as we delve into some Limited strategies.
What is Limited?
Limited is a style of play where instead of constructing a deck from your entire collection, you build a deck from the cards you open from a limited (heh, get it?) card pool. Limited formats include Draft, where you select cards from packs and pass those you haven't chosen to your opponents, or Sealed, where you open a set number of packs and build a deck from your pulls. The common thread is that you have no control over what you open. This presents the challenge for Limited deck building: determining which of the cards you select or open are good enough to play.
Before we go on, let's quickly discuss how Limited decks are different than Constructed decks. In Constructed Magic, you have a minimum deck size of 60 cards and you can only play up to 4 copies of any non-basic land card. On the other hand, in Limited your minimum deck size is 40 cards and you can run any number of copies of a card. Of course, it's harder to get 5+ copies of a single card in your deck, but if you drafted 23 Jace, the Mind Sculptors you can play all 23. In those 40 cards you usually want to run somewhere between 17-18 lands and 22-23 spells. Among those 23 spells you should be running 16-18 creatures, which means that creatures are the most important card type in Limited. Many Limited games come down to a key combat step or finding the right removal spell to open up an attack for lethal. So if creatures are the most important card type in Limited, being able to determine which creatures are better than others is a skill that we need to develop. Now it's obvious that Kalitas, Traitor of Ghet is better than Zada's Commando, however it may not be so obvious whether Ondu War Cleric is better than Makindi Aeronaut. That's where we use one of the most popular tools in the Limited players' arsenal: The Vanilla Test.
Vanilla Test? Why not Chocolate?
The Vanilla Test is a way to gauge how effective a creature is, if it has no rules text. It gets its name from the slang that a "vanilla creature" is one with a name, mana cost, power, toughness, and a whole bunch of flavor text but no rules text. Some popular examples are Grizzly Bear, Nessian Courser, and Broodhunter Wurm. How the Vanilla Test works is you compare the Power and Toughness of the creature to its Converted Mana Cost (CMC). So a Grizzly Bear is a 2/2 for , a CMC of 2, so our Grizzly Bear passes the Vanilla Test since we’re getting 2 power and toughness for 2 mana. On the other hand a creature like Catacomb Slug would not because it's a 2/6 for , a CMC of 5, where 2 power or 6 toughness for 5 mana is not a great deal.
This test lets us compare two creatures to see which one might be better than the other, however it fails in some respects. In triple Battle for Zendikar draft, an odd format for those who have played it, I could present two creatures and ask which one does better on the Vanilla Test and then which one is better in the format: Snapping Gnarlid and Kalastria Healer. I don't think anyone would argue that a 2/2 is better than a 1/2 on the Vanilla Test. However the Healer was much better than our Gnarlid friend in the format. Part of it was with how poor Green performed in BFZ draft, but also the abilities of each creature have to be taken into account. And that is where I came up with a better way to compare these creatures.
Vanilla is Boring, Let's Modify It!
I have a Bachelor's degree in Math, and I even have a published paper to my name (not so humblebrago), so forgive me if I think in terms of numbers and formulas. To start improving on it, I'd like to quickly formalize what the Vanilla Test does:
In this formula V is the Vanilla value, P is the Power of the creature, T is its Toughness, and CMC is the creature's converted mana cost. We then take this Vanilla value and compare it to a favorite number of mine, 1. With this formula, any creature with the same power and toughness as its CMC will score a 1. That way it makes sense for us to make 1 our tentpole to compare creatures to. There are some weird cards like Ancient Crab with a 1 Vanilla value despite being a 1/5. But still this doesn't explain why Snapping Gnarlid with its 1 Vanilla value is worse than Kalastria Healer with its 0.75 Vanilla value. This is where I started with how to improve the Vanilla Test, and where the initial question came from: "How can I take into account a creature’s abilities into the Vanilla Test?"
Before we make any changes, the Vanilla Test does help us predict what sort of rules text a non-Vanilla creature might have. Let's take two rares as examples: Inverter of Truth has an absurd 1.5 Vanilla value whereas Munda's Vanguard has a 0.6 Vanilla value. The Inverter's Vanilla value draws our attention and tells us that there should be some downside, which the rules text does show us. The Vanguard's lower Vanilla value of 0.6 should be weaker than the Inverter, however the ability on the Vanguard is an upside. So what I hope to achieve with my new Vanilla Test is a way to capture how an ability impacts its power level.
And that is what led me to the following formula:
Here the only change is the A variable, which I call the Ability Modifier. Whilst I'd love for this to be purely objective, the Ability Modifier is quite subjective. But in a game where many decisions we make, from gameplay to deckbuilding, is subjective I find this to be an acceptable use of some subjective variables.
The Ability Modifier attempts to give a value to the abilities a creature has. My inspiration for it came from an old article by Gavin Verhey, Magic designer, where he mentioned the scale that R&D uses to compare a card's strength in Limited. In the article Blind Phantasm is a B creature while Wind Drake is an A. So despite the fact that the Phantasm is a 2/3 for 3 mana, and a Vanilla value of 0.83, the Wind Drake is considered better despite the fact that its Vanilla value is less, at 0.67. So what's the difference? The Wind Drake has flying! This tells me that R&D values Flying as equivalent to adding 1 mana to the Vanilla value of the creature. Just as we expect power and toughness to have an impact on CMC, we also expect abilities to have an impact on CMC too! So in the example of Wind Drake you’re paying for the 2/2 and an additional for the ability to fly. This leads me to put the Ability Modifier of Flying at 2. With this in mind, here is a table of all my Ability Modifiers for the Evergreen creature keywords:
I came to these numbers through looking at many different sets and creatures to determine which values make the most sense. So with the Modified Vanilla Test (MVT) our Wind Drake from before has a MVT of 1, which is the threshold we needed for the Vanilla Test (VT) to equal 1, as seen on a 3/3 for 3 like Nessian Courser. Some other often-used abilities I also apply an Ability Modifier to include: -N/-N (decreasing power and toughness) is 2N, +N/+N (increasing power and toughness) is N, Life gain of N is N, Life loss of N is N, Tapping a creature on ETB is 2, Untapping a creature on ETB is 1, Drawing a card is 1 for each card drawn, and Discarding a card is 1 for every 2 cards discarded.
So let's see how some popular creatures fare under the MVT.
Gray Merchant of Asphodel
Ah, Gary. The first creature I applied the MVT to. Under the VT, Gary has a measly value of 0.6. However under the MVT alone he causes a life drain of 2 (2 lifegain, 2 lifeloss), so his MVT is 1.0 if he's the only black permanent in play. Gary is the first instance we have (in this article) of a variable ability modifier that scales with the devotion to Black. That is, an ability which scales up. If you have 3 devotion, then Gary’s MVT improves to 1.2, 4 devotion is 1.4, and so on. This is what the MVT is supposed to do, to quantify how much better a creature is than vanilla! And this doesn't take into account multiplayer play!
Thragtusk
The Thragmother, bane of Standard of ole. On the VT, Thragtusk sits at 0.8, not quite vanilla. However if we add in the 5 lifegain, Thragtusk has an MVT of 1.3. But that’s not all. When Thragtusk leaves play you get a 3/3 back! Because of this, we need to add in the 3/3 to the MVT. This means that Thragtusk's final MVT is 1.9!
Siege Rhino
Let's talk about our favorite pachyderm, Dr. Siegeman Rhinocerous. Siege Rhino is already above the curve with a VT of 1.13 without the rules text. But then Siege Rhino drains our opponents' for 3 and has trample! So the MVT for Siege Rhino is 2!
Kor Sky Climber
Back to the commons for a key creature in many White decks in OGW-OGW-BFZ draft. The Sky Climber already sits at a reasonable 0.83 VT, so you'd play her anyway. But the ability to gain flying, even with a mana payment, fits the 2 of Flying. You can reduce the Ability Modifier to 1 if you value being able to maximize your mana, however either way Kor Sky Climber has a MVT of 1 at worst or a MVT of 1.17 - well above Vanilla.
Kalastria Healer
For our last card, let's consider Kalastria Healer. On its own, without any Ally support, its MVT improves from the earlier VT of 0.75 to 1.25. Then, with each Ally you play after it, you improve the MVT by 0.5. This allows for the Healer to snowball out of control and be a Must Kill.
Well That Was a Thing
I'm not going to claim that this is perfect, it could use some iteration and refining. But I feel like the MVT does a better job of assessing a creature’s power level than the Vanilla Test by also taking into account the abilities of the creature. Hopefully this can be informative, and perhaps shape your picks, at your next Limited event!
If you have any questions, you’re free to comment below. I'm also available on Twitter @jwiley129 and on MTGO at the same handle.
It looks like some apostrophes got edited out in some places.
March 1, 2016 9:01 a.m.
lemmingllama says... #3
I'd definitely be interested in reading some other articles about this. I already have my own way of working through the vanilla test, but it's fairly similar to your own process. Maybe your next article should talk about cards with alternate casting costs like Mulldrifter or Grave Scrabbler and how to assess them.
March 1, 2016 9:07 a.m.
Would be nice that especially bad creatures would drop significantly in the V score...perhaps even negative numbers?
Perhaps an additional subjective modifier for the creature's rarity, to compensate when you drew them in a booster.
Yes, looking at you Ruby Leech and Trained Orgg.
March 1, 2016 9:20 a.m.
lemmingllama - Thanks for the feedback! I wanted to start with the basics, evergreen keywords, because those are the easiest to evaluate. Once you get into block mechanics it gets a little hazy, as I hope Gray Merchant of Asphodel and Kalastria Healer showed.
Alternate casting costs are certainly an interesting conundrum, as you essentially have to ask "Is this extra hoop worth it?". With Evoke, they costed the evoke costs as though they weren't creatures. Therefore I wouldn't include the Evoke part in the MVT. However Mulldrifter is great by all metrics. Madness has its own issues, which I'd be willing to look into more once we get to SOI.
GobboE - The VT is used across rarities, so I would be hesitant to add in "oh, this card gets a -1 for being a rare" etc. And being close to 0 already indicates having a poor VT score, so I don't see how negatives help in any way. I quite like the scale being essentially between 0 and 2, with a few exceptions for variable Ability Modifiers.
March 1, 2016 9:25 a.m.
GeminiSpartanX says... #7
Great Article! I enjoyed reading it. It's unfortunate that we often don't have time to do a bunch of math while drafting, but good rounding skills will help here. ;)
March 1, 2016 9:26 a.m.
JWiley129: Did you ever try subtracting the A from the denominator instead of adding to the numerator?
=(P + T)/(2C - A)
Also, what value did you use for A in regards to each point of life gain/ life loss caused?
Would gain 1 life = 1?
Would opponent loses 1 life = 1?
And would you add them together for drain?
March 1, 2016 9:52 a.m.
DruneGrey - I never considered subtracting from the denominator, I liked the idea of seeing how the abilities brought the creature up to pass the VT as opposed to lowering the CMC to match their VT. I also go into the lifegain/drain bit with the analyses of Thragtusk, Siege Rhino, Gray Merchant of Asphodel, and Kalastria Healer. But broadly, I consider 1 life gained equivalent to 1 toughness added and 1 life lost equivalent to 1 power added.
March 1, 2016 10:15 a.m.
I find it interesting that you rate +X+X as equivalent to -X-X. -X-X seems MUCH more powerful but I haven't looked at many cards so I don't have data to back that up.
March 1, 2016 10:18 a.m.
I think I agree that -X/-X should be a little higher as +X/+X helps you win a combat, -X/-X/ helps there be no combat.
Also, really like this article. I would love to try and work with you on this! Currently building an excel that can automate a set evaluation quickly.
Thoughts on lowered costs; i.e. Delve or Surge? Would you use C=(CMC+Lowest Cost Possible)/2? So C for Tasigur, the Golden Fang would be 3.5 and for Reckless Bushwhacker it would be 2.5?
March 1, 2016 10:30 a.m.
I guess it wouldn't work for Surge costs as they'd have possibly have something extra if surged...Do this for delve (though we probably won't see this again), or any other cost reductions that don't affect the effect. Then have a range for Surge cards?
March 1, 2016 10:41 a.m.
DruneGrey & ChiefBell - You're correct that -X/-X is more powerful than +X/+X, so massaging those numbers a bit would be something I would do. However I'd like to look at a few examples of creatures that do so on ETB w/o any riders. Silumgar Butcher and Wasteland Strangler both come to mind as creatures that do this, but both require additional setup.
As far as cost-reduction mechanics. In these I would add a Cost Reduction Modifier to the denominator of the MVT. They're already pretty hard to evaluate, but I think I have a solid idea for a few.
Surge - If you notice, the Surge cost of some of the commons is about 1-2 more mana than you would expect to pay for that spell. However in all cases the least amount of mana you can pay is 0, which is great, but the real-world-case is at least 1. This would make me want to adjust the CMC of a Surge card to become the average of its Surge Cost and it's actual CMC. So Jwar Isle Avenger would have an Adjusted CMC of 4 since (3+5)/2=4. And 4 mana for a 3/3 flier is an MVT of 1. Reckless Bushwhacker is harder to evaluate there, since he gives +1/+0 if surged. This leads me to want two different MVTs for Surge cards: A non-Surged MVT and a Surged MVT. Then I would take the average of the two.
Delve - Looking at the "vanilla" delve creatures I'd wager that 
in Delve is equivalent to 
paid normally. So Gurmag Angler would normally cost 
in a regular set, but because of Delve gets added on. This mans that Tasigur, the Golden Fang's non-Delve CMC would be 
, which makes me uneasy. So I'd want to adjust a Delve spell's CMC by 1.
March 1, 2016 11:31 a.m.
miracleHat says... #16
Hmm, do you plan on continuing this? I am seeing this going into a very advanced analysis of limited if continued to be published.
March 1, 2016 12:35 p.m.
miracleHat - I certainly could. The directions I could go in vary, and would depend on the draft environment. E.g. # of Ingesters vs # of Processors in 3xBFZ draft. I could also write an article where I put my evaluations of block mechanics for the creatures. But a lot of the deep dives have already been done. Plus Frank Karsten could do what I'm doing a lot better, but I'm sure he has bigger fish to fry :)
March 1, 2016 12:55 p.m.
miracleHat says... #18
That is one of the great things about writing. You can write about things that people have already put out there, but you can rewrite it from how you perceive it and it turns into something completely different and new.
On the plus side, if you do continue this: the new set(s) are coming out soon. Preliminary strategy suggestions and tips would be delightful...!
March 1, 2016 1 p.m.
FAMOUSWATERMELON says... #19
Very very interesting. It's nice to see some numbers around here other than average CMC and number of copies :)
As for the formula itself, I do think that there should be a way to factor in downsides, such as adding to the CMC. So if we were to do Inverter of Truth, we would have:
MV = (P+T+A)/(2x(CMC + D)) = (6+6+2)/(2x(4+3)) = 1, where D is the downside.
This would be more accurate, as the Inverter has a downside that almost loses you the game on the spot, unless you got some very good synergy while drafting. The 3 could be changed to a 4, 2, or 1 depending on the downside, of course.
Also, it's Grizzly Bears, not Grizzly Bear, the card currently links to a deck.
March 1, 2016 1:27 p.m. Edited.
FAMOUSWATERMELON - That's a fair point, I never actually discussed the Inverter of Truth again. For me its Ability Modifier would be negative. I should probably write a follow up with more examples of some of my evaluations through the MVT.
Inverter of Truth has flying, so the A is going to start at 2. Now how much you rank the downside of "exile your library, then shuffle your graveyard into your library" is will be up to interpretation. I would probably put it at somewhere between -4 and -6. At the end of the day, I think that Inverter of Truth's MVT is somewhere <1.
March 1, 2016 2:39 p.m.
If you set it at -6 he's a 1. I'd also say he should get an additional -.5 to -1 for having double color in his cost, which does take him below 1.
I'm trying to run through Khans now with the following thoughts;
I'm running it with both A adding to the top for your version and A subtracting from the bottom for mine, the values are essentially the same through 20 creatures.
Morph +1; as its another option that doesn't negatively affect the card
Outlast +.5; it takes mana and a tap for a single +1/+1
+1/+1 counter lord effect (A score of granted ability)/2; it needs to either be outlasted or have counters from something
Double (of one) Color in CMC -.5; this makes it harder to cast outside of a mono colored deck which isn't usually what happens in limited
Triple Color in CMC -1; this makes it much harder to cast in limited
Two Colors in CMC; even in limited this doesn't seem bad, so I've got it tentatively at 0 but maybe should be -.25 to -.5
March 1, 2016 3:32 p.m.
Gurmag Swiftwing has shown me the error of my ways. If you remove the Modifier from the Costs in the denominator you can sometimes get to a denominator of 0...which obviously doesn't work too well...
As the results through the first ~35 creatures were nearly the same, I will concede to your method good sir.
March 1, 2016 3:53 p.m.
Coinman1863 says... #23
JWiley129, this a more of a food for thought kind of thing, but MaRo just confirmed that Dryad Arbor was a "Vanilla Creature". So, I have no idea how this effects your thinking but I thought I would share.
March 1, 2016 4:29 p.m.
This MVT to me almost seems more like an efficiency metric than anything else. where a bear is a 1. Higher than that you are more efficient, lower you are less efficient.
MaRo means simply P/T/CMC full stop. This is a more in depth look at efficiency using vanilla as a baseline. Dryad Arbor would be a case to subtract ability from denominator (the tap to add mana ability) to prevent a 0 there. But, getting anything for what is essentially nothing means it is one of the most efficient creatures possible.
March 1, 2016 4:50 p.m.
TheMuffinMTG says... #25
Sorry that this was the main thing I got out of this... you can run more than a playset of cards in limited formats?! MIND BLOWN. How have I never known?
March 1, 2016 5:16 p.m.
GreenGhost says... #26
Tranple and First Strike should have a value of 1.5 on the A scale because both of these offer pseudo evasion. I would even consider adding Menace to the list of 1.5 and possibly bumping down Vigilance to the 1.5 realm because it is nowhere near as good as flying. I also think that the N in Scry N should be treated as N/2.
March 1, 2016 5:48 p.m.
Seems a lot of people don't know this. I drafted 5 copies of Oblivion Strike a couple weeks ago and drew them all in one game, much to my opponents dismay. When I cast the 5th he was very confused and thought I was cheating. Had to explain that the playset rule doesn't apply to limited formats.
March 1, 2016 6:35 p.m.
DruneGrey - I certianly could add that level of granularity, but I prefer keeping the Ability Modifier values to what shows up in the rules text and not the mana costs per se. Also, this applies to GreenGhost, I'd like keeping the Ability Modifier values to only integers. This keeps the math (mostly) easy.
Coinman1863 - Dryad Arbor is an interesting case. The forumla as I've made it does prevent us from analyzing the 0-cmc creatures: Ornithopter and Memnite. In the case of Dryad Arbor I would say it's CMC is essentially 1 since it can't tap for mana the turn it comes into play, so it's MVT is basically 1. There would be a similar analysis for Ornithopter and Memnite as well.
March 1, 2016 7:20 p.m.
This idea has a lot of potential if you streamlined it. A simple test to see if a creature is good is a great way to improve draft or cube.
Unfortunately, I'm not up to doing algebra in my head during draft, or even sealed. Perhaps if you altered this to a star rating.
Vanilla Rating: 0 to 2 stars (0 being below vanilla, 2 being above it.)
Ability Rating: 0 to 2 stars (0 being no abilities, 2 being very good abilities)
Synergy Rating: 0 or 1 stars. Does this card synergize with cards you've already got?
March 1, 2016 8:26 p.m.
Egann - I'm not advocating you do all these calculations in your head, instead I'm suggesting do these before you head to the draft. The purpose here is to improve the VT, which I feel like I have done. The purpose is not necessarily to have the MVT be that easy to use. Streamlining it beyond this would involve assigning values to all creature abilities, which I don't have the time nor temperament for. I'll happily do so for future set mechanics, and other creatures, but as it sits I'm pleased with how it is. Not to say there won't be changes, the discussion we've had here certainly helps that along. But something like what you suggest is not where I see the future of the MVT.
March 1, 2016 9:37 p.m.
JohnnyBaggins says... #31
For the Vanilla test on 0 Mana creatures, there's a very good chance your verdict for those creatures should be "Never play them". Unless, obviously, they are essential for your synergies - but then the vanilla test doesn't apply anyways.
As Limited is my main thing about Magic, I really like the idea of this Chocolate Test. Here's some things I'd change though:I'd upgrade Menace and Deathtouch to at least 1.5. In addition, combining fist strike and Deathtouch is absolute nuts and needs upgrading. Admittedly, this is a combination that you probably exclusively see on Rares, but still. Same somewhat goes for Flash and Deathtouch in a package. I think Scry N should also be different. I would do it like this:
Scry N
if N == 1
value = 1
else
value = N/2
Also, as a mathematician you're supposed to be lazy. I'm somewhat offended that your list names "Scry 1 = 1" and "Scry N = N" separately. That makes no sense.
Lastly, I'm pretty interested on how this carries out. How would one assert a card like Heir of the Wilds? Its power changes. What's with Prognostic Sphinx? It doesn't always have Hexproof. What's the difference between Scries on ETB; Scries on Damage and Scries on Attacks? What's with Scries on Upkeep? What if the card isn't always a Creature? Imagine the Gods from Theros or Glint Hawk Idol. Obviously, Cards like Ojutai, Sphinx and the Gods are Straight-Up Bombs and don't really need a Vanilla Test, but still. There's a lot of questions and I'd love to see them somehow answered.
Your entry is awesome, keep the good work up!
March 1, 2016 10:29 p.m.
this is something really interesting and for a noob player like me (been playing for a short time, 1 year at most; and only with a few friends. This is something that would definetely help me and other new players out when going to a game store to play for draft/pre release. hope to see more of this!
March 2, 2016 2:57 a.m.
JohnnyBaggins - The Scry N could be worked on some, I remember I thought of the scale for Omenspeaker but I forgot my original templating. But yes, the other creatures/abilities you mention are very tricky to evaluate using the MVT. Hence the subjectivity of the Ability Modifier.
Now I wouldn't include a card like Glint Hawk Idol or any non-creature card that can turn into a creature. That is because they can't, and shouldn't, be evaluated in the same manner. The Gods would work under the MVT, but they do have a larger drawback of the Devotion requirement to get them active.
March 2, 2016 1:19 p.m.
Cards like Hangarback Walker, Hooded Hydra and Clever Impersonator all appear a little difficult to evaluate because of how far from Vanilla they are. :\
March 2, 2016 1:25 p.m.
DruneGrey - I never said the MVT made everything easy to evaluate :p
That said, Hangarback Walker is actually quite easy. It's CMC is 2X and enters as an X/X and dies into X 1/1's with Flying. So the MVT for it would be
MVT= (X+X+2X)/(4X)=1
And that is ignoring the tap ability and the fact the tokens have flying.
Hooded Hydra, and other morphs, are similarly difficult. However using the variable CMC route like we did with Hangarback Walker should give you a good picture.
Clone effects are always rough, and as such should probably be considered an outlier.
March 2, 2016 1:35 p.m.
JWiley129 - Yeah, and I guess I've been trying to get this to a single relatable number per card, but that really would take something more in-depth than what I believe you are intending with this. Probably need much more discussion around the actual value of abilities, particularly the costed ones like morph/outlast/pay X to do thing, to get this to be more detailed. Definitely like this though and will be using it moving forward, trying to tweak things here and there.
March 2, 2016 1:43 p.m.
DruneGrey - Finding an MVT for every creature in Magic is probably a losing battle. That's why I'm going to focus my efforts on current Creatures and upcoming ones.
March 2, 2016 2:15 p.m.
ToolmasterOfBrainerd says... #38
I think something similar to the MVT for noncreature spells would be interesting. Obviously it wouldn't be perfect or work for all noncreature spells, but for pump spells, burn spells, and nonbasic lands, I think it could translate quite well.
My math mind keeps reading MVT as 'Mean Value Theorem', and then I have to think hard every time I come to this thread to remember what it actually stands for.
March 2, 2016 2:47 p.m.
I toyed with the noncreature version ToolmasterOfBrainerd, but it's very hard to establish a baseline for what "vanilla" means for noncreature spells. Especially considering that "Vanilla" means "no rules text" and noncreature spells all have rules text.
You can extend this to creature lands a bit, but the analogies get put off a bit.
March 2, 2016 2:58 p.m.
Haha, I cast Instant. It's U for an instant that triggers prowess and sets up surge...end text...
March 2, 2016 3:29 p.m.
This would be something that WotC and Star City article columns might like to see, however it would be shameful if somebody tried to take credit for it.
Sharing this with my local area facebook group. Great analysis Wiley.
-fist bump-
March 2, 2016 4:10 p.m.
Very nice! I enjoyed this read quite a bit!! I will say it isn't the easiest thing to calculate for 15 cards in the time a draft takes, but if you apply this theory when stuck between picking one card or the other, it can really be a ninch tool! Not to mention the actually deck construction parameters where this gains a whole new sleuth of worth.
March 2, 2016 5:43 p.m.
I would just like to add that an effect upon leaving the battlefield is generally strictly worse than the same effect upon entering the battlefield. Therefore, Thragtusk's 3/3 can't actually be added to Thragtusk's MVT at the normal multiplier. Maybe x2/3 for triggering on exit?
March 2, 2016 10:38 p.m.
BloodoftheBloodMoon says... #46
I would love a 0 cost instant/sorcery spell that did absolutely nothing. i'd play that in a storm deck any day! I'm also just commenting to sub. this is very interesting. I geek out on math and magic alike :)
March 2, 2016 11:55 p.m.
JohanEoghan says... #47
Just a thought. A mana cost of 2 colours compared to a mana cost of 1 colourless and 1 colour? 1 colourless and 1 colour should be easier to cast, but 2 colours are far better when it comes to devotion. What is the better card there, if both are 2/2 creatures?
March 3, 2016 3:59 a.m.
Very interesting! Got my mind turning.
Consider having 0 power applying -2, and 0 toughness as -4.
As for 0 cost creatures I would suggest using .5 in place of CMC. So...
Ornithopter MVT:2 (0+2+2-2/2*.5=2)
Memnite MVT:2 (1+1+0/2*.5=2)
Force of SavageryMVT:.83 (8+0+1-4/2x3=.83)
March 3, 2016 6:29 a.m.
Souljacker says... #49
I think it's funny you make a test that's powerful because its simplicity more complicated. I wouldn't see this as a substitute for the vanilla test, but more as a separate test.
As you see, it's hard to put numbers to the value of the rules text. This is exactly why we have the vanilla test; to determine whether you'd play the card without the rules text; that's a simple question and when answered yes, if the rules are favorable, you will definitely play the card.
This helps mainly with preventing selecting bad cards that have an upside that cannot be achieved consistently, something many newer players do. If you use the vanilla test on all cards; you are sure you will not have a bad deck. You might miss out on potentially good cards, and to prevent that I think you can run the MVT as 2nd test to see if a card does pass that one.
If you use it this way, I do think you need to lower the values on a couple of abilities, as with this process you rather not select a good card than to select a bad card. A couple that come to mind that are too high to me:
Lifelink, Scry, Vigilance, Reach, Flash, lifegain.
Then you also need to account for the power of the creature that has influence on these abilities; it's much better to have lifelink on a 5/5 than on a 1/3.
March 3, 2016 6:39 a.m.
But equally there are lots of cards that are bad from a vanilla perspective but great from an abilities perspective. Spark Trooper was an absolute beast in limited, for example. So was Silverblade Paladin.
And secondly there are many limited format where decks of good vanilla cards actually aren't that favourable. INN draft for example favoured tricks, card advantage and special rules over vanilla beaters. Whereas on the other hand AVR draft favoured vanilla beaters over special rules and tricks.
So I think the MVT is a very necessary part of the game. It seems like recent limited formats such as khans, and zendikar have favoured very vanilla centric construction but that's not necessarily always the case. And in these more complicated limited environments the MVT is a necessary tweak to the VT.
March 3, 2016 7:31 a.m.
Souljacker says... #51
What I'm trying to say that you should handle and design the test in such way that it either needs to pass the vanilla test, or otherwise the MVT. Of course it's best to have a test that would pass all and only good creatures, but at the same time I think it's necessary to have a simple rule of thumb for new players to construct a somewhat decent deck.
March 3, 2016 7:58 a.m.
I think you apply the VT first. If the card passes you may be happy or not, depending on the limited environment and what your deck needs.
Then afterwards you factor in the abilities for a rough MVT calculation. Once again you get a figure and you factor in the usefulness depending on the limited environment and what your deck needs.
The VT is definitely the first pass I make for any card BUT if a card fails I don't just bin it, I give it a quick MVT checkover just to definitively confirm whether it is unplayable or not.
Really picking in a limited environment is a two-step process. Power/Toughness then abilities can be broken down into stages. Then if you want to consider what others are picking and what your deck needs it becomes a 3 step process.
March 3, 2016 8:13 a.m.
Souljacker - After carefully reading your comments, you're mostly correct. The MVT has its pros and cons, the main con being it's not as easy to compute as the VT is. Also, you've hit on why finding the Ability Modifier values is difficult, because it's mainly a subjective idea. That subjectivity hurts the overall value of it, since people will disagree with what I've chosen. Much like you think some of my choices are overvalued.
I also see the MVT less as a "Hey new player, look at this thing to help you in draft!" but more as a "Hey experienced player, here's something to help you get an edge!". Certainly a new player could use the MVT to get a solid baseline for a format. But I feel the real use is going to be as a tool experienced players can use.
I also agree with ChiefBell. In an event, you won't necessarily have time to consider the MVT, while the VT is easier to grok. However prior to the event, you can apply the MVT to the whole set and make a list of MVT values for each creature. That way you have a leg-up when it comes to construction and selection of your deck.
A good analogy is: The VT is for first impressions, the MVT is for the job interview.
March 3, 2016 11:31 a.m.
MurderHood says... #54
Hey JWiley129 what would protection from (color) be worth on that point scale?
March 3, 2016 noon
I generally put protection somewhere in the 1-2 range. Also, the broader the protection, the more it's worth. So Protection from Dragons is probably a 1 while Protection from Multicolored is closer to a 2.
March 3, 2016 12:07 p.m.
JWiley129,
This is really cool. Like others have already suggested, please keep writing!
I'd really love to see an analysis of SOI after the full spoiler comes out, and then a comparison again after a nmonth or so, taking a hard look at the cards you got "right" vs the cards you got "wrong"
Thanks for giving me a new way to look at Limited!
Draft well!
March 3, 2016 2:06 p.m.
clayperce - Thanks for the feedback! A SOI analysis is on my to-do list for the MVT at least.
March 3, 2016 2:22 p.m.
I went through and assigned values to different abilities in the Khans of Tarkir set. Below are what came up as my top 20 with MVT, below that is the list based purely on VT. These are also all based off of my personal thoughts on the value of outlast/conditional abilities/etc...
MVT List;
Name (MVT #)
Monastery Swiftspear (2.5)
Kin-Tree Warden (2)
Ruthless Ripper (2)
Siege Rhino (1.875)
Herald of Anafenza (1.75)
Gurmag Swiftwing (1.75)
Icefeather Aven (1.75)
Mantis Rider (1.6)
Bloodsoaked Champion (1.5)
Master of Pearls (1.5)
Rakshasa Deathdealer (1.5)
Savage Knuckleblade (1.5)
Surrak Dragonclaw (1.4)
Chief of the Edge (1.375)
Chief of the Scale (1.375)
Seeker of the Way (1.375)
Temur Charger (1.375)
Butcher of the Horde (1.375)
Sagu Mauler (1.33)
Pearl Lake Ancient (1.32)
VT List;
Name (VT #)
Disowned Ancestor (2)
Monastery Swiftspear (1.5)
Herald of Anafenza (1.5)
Bloodsoaked Champion (1.5)
Dragon's Eye Savants (1.5)
Embodiment of Spring (1.5)
Meandering Towershell (1.4)
Savage Knuckleblade (1.3)
Anafenza, the Foremost (1.3)
Mardu Blazebringer (1.3)
Chief of the Edge (1.25)
Chief of the Scale (1.25)
Archers' Parapet (1.25)
Surrak Dragonclaw (1.2)
Bear's Companion (1.2)
Ivorytusk Fortress (1.2)
Siege Rhino (1.125)
Butcher of the Horde (1.125)
Bellowing Saddlebrute (1.125)
Kin-Tree Warden (1)
March 3, 2016 3:11 p.m.
JohnnyBaggins says... #59
Hm. One might have to work on this. Neither Swiftspear nor Kin-Tree Warden are cards I really want to have in my deck.
March 3, 2016 3:18 p.m.
JohnnyBaggins - I think the real question is how much does DruneGrey value Morph, Outlast, and the other abilities from Khans. Personally, I wouldn't rate Morph in the Ability Modifiers unless there is a quantifiable benefit to flipping it face up, for example Icefeather Aven. We really can't get a good feel for these values we really need to know what each ability is valued.
March 3, 2016 3:27 p.m.
JWiley129 - I have Morph as a 1 across the board. If it has a relevant ability that gets added separately. Morph always gives some as being able to play this card as a 2/2 for 3 is an ability with merit in its own right. Icefeather Aven is 2/2 for 2, no penalty as the two colors are different in the cost, 1 for being able to morph at all and a 0 for the ability as it isn't worth much at 6 mana total for a bounce effect, and sometimes you don't get it.
JohnnyBaggins - As far as Monastery Swiftspear and Kin-Tree Warden both are extremely efficient creatures. Swiftspear sees its efficiency in the early game (Haste + Prowess), while Kin-Tree's is in the late game (Block anything w/o trample for 2). I understand that they aren't abilities that win you a game of limited, but they are very good at what they do for their cost and certainly don't hurt your deck.
I don't believe that MVT is a tool for what to "Pack 1 Pick 1" but more a way to see that cards have more value that just the VT would indicate. Also, as mentioned, a lot of the A score is subjective, and your own MVT would shift during a draft as you determine what direction you want to go.
March 3, 2016 4:11 p.m.
JohnnyBaggins says... #62
Icefeather Aven is an absolute killer. I'd firstpick it any day. But Swiftspear is pertty abyssmal in limited and Kin-Tree Warden, even taking two slots on the curve, just isn't good. I get how you can prolong your progress with dying, but don't tell me you've ever won a game on the back of Kin-Tree Warden. Not unless you had some crazy counter-action going on.
Generally, my idea about morphs: If I have a card that costs X and also has morph, it's strictly better than the same card without morph. Therefore, Morph is a slight upside. Really relevant is when it gets the following: You might cast the card on turn 3 as a morph. It's then a slighly off-curve bear. You can then morph the card up. If the morph cost's below the normal casting cost, that's another plus. As such, the card can be relevant at three different spots on your curve and, therefore, is quite good. If I get some stupid effect like Efreet Weaponmaster, the card also suddenly has first strike and ruins combat, or like Ruthless Ripper that suddenly has deathtouch and got you fooled, it's even better. Morph, in general, is an upside.
March 3, 2016 4:12 p.m.
Souljacker says... #63
JWiley129 I applaud you for setting this up, I'm just trying to point you were I think you can improve your idea. Thanks for taking my feedback serieusly. As I said, some abilities get better or worse as the P/T changes, so you might need to adapt your formula or ability points for that. For example, the A for lifelink could be 0,5*P(ower).
March 3, 2016 5:02 p.m.
Souljacker - And you aren't wrong, however I place a very high premium on keeping the values I plug in to be integers. This reduces the computational complexity, while losing a bit on the level of accuracy.
Also, as per lifelink on creatures, obviously lifelink is better on a 5/5 than a 1/1. However there are several creatures who were high picks in their formats as 1 power lifelinkers (Hopeful Eidolon, Sungrace Pegasus, etc.). So I feel quite confident in my valuation with lifelink's Ability Modifier.
March 3, 2016 5:51 p.m.
JWiley129 This article is extremely amazing and useful. Big ups for you! As a Cube and Standard player, I have no words to describe how good is this!
Although I completely agree with your formula, I believe that, in reality, things tend to differ a little.
I always have a discussion with some friends regarding lifelink. Practical Magic tells us you're not going to win a match by gaining life. Yes, more life is more time to stay alive, which is more time to draw what we need, and change the course of a match. Yes, it has already been proved how a game was won if a Sphinx's Revelation for 5 did resolve. Still, a 2/2 deathtouch for 
shouldn't be below a 2/2 lifelink for 
. It's not the same to block any of these creatures with a simple 1/3 vanilla. I believe you get my point :)
PS: In the article, Grizzly Bears is linking to a deck, not to the most beautiful card Magic has ever given us ;)
March 3, 2016 10:40 p.m.
Souljacker says... #66
If you want integers I can understand, but then you could just up all the numbers. If you goal is to get as good an estimation of the value of a creature, such adaption is necessary.
In your examples of the low power lifelinkers, those had value because of their other abilities (flying and bestow), that made the lifelink better too. By bestowing the Eidolon on a higher power creature, you got more value out of your lifelink, and lifelink on a flyer is a reliable source of damage and lifegain while a 1 power ground creature cannot attack as freely and thus the lifelink has less value on that.
March 4, 2016 3:15 a.m.
People need to understand that this is a primitive attempt at improving the vanilla test. You need to start with very simple modifications and test them out before you can go into more complex equations. I'm pretty impressed with what you have done, JWiley129. You introduced a simple adjustment that considerably improved the vanilla test. Good job.
The only problem I see is the ability modifiers, while it might be true that Deathtouch, Trample, and Reach have an ability modifier of 1 from a design perspective, they are definitely not equal in terms of functionality. I don't think anyone would deny that, in most cases, Trample and Deathtouch are better than Reach. Anyone would pick Trample over Reach even if the resulting MVT was the same. This is probably where you want to start improving on the design.
Now if anyone wants to improve on the MVT, they would have to start with ability modifier. Once they have a solid idea of how the abilities compare to each other they can manipulate the equation into a power, log, or exponential function. Another thing that could probably further improve on this would be to introduce an "interaction coefficient" that measures how abilities interact with each other and, say, ranges from 0-2 (<1 unfavorable, 1 nuetral, >1 favorable). For example: Flying + Reach would be <1 because if makes no sense to have both, Flying + Flash would be 1 because they don't interact, and Flying + Haste would be >1 because of the positive interaction between the abilities.
March 4, 2016 4:24 a.m.
JohnnyBaggins - Again, I say in my response that neither of those creatures win you a game. All I'm advocating for is that for their mana costs they are very efficient. Also the reason I'd take the mana cost of morph into effect is simplicity. A 3 cost creature with morph plus flip for 2 is actually paying 5 for the creature plus some effect. I'd argue in that case it's a waste of mana if though the flip cost is less than the true casting cost, even the larger morph guys say a 5 that flips for 4, sure you're using all of your mana on curve but you aren't significantly advancing the board. I'm trying to look at this from the perspective of, this will be the only creature on the board. If we start to get off in the weeds and take the individual creatures out of this vacuum, we end up with a much more complicated measure.
I also would like to emphasize my agreement with JWiley129, and forgive me if I misunderstand your opinion, that critical thinking is still needed with evaluating cards and this isn't intended to take the thought out of drafting, but just to help you form your own opinions. The value assigned to block specific mechanics and even evergreen ones should be constantly shifting based on how the limited format actually plays. This is just a starting point, not a destination. Again JW, I don't want to put words in your text box, this is how I believe you intended this MVT.
March 4, 2016 8:15 a.m.
ComixWriter says... #69
How do we evaluate tokens, especially tokens that may immediately produce mana, like Edlrazi Scions?
Is Scion Summoner different than Eldrazi Skyspawner or Ghirapur Gearcrafter?
Does this formula also hold value for non-creature spells that create token creatures? For example, Dragon Fodder is CMC2 for two 1/1 goblins. While 'goblins' may have value outside of the spell, can we correctly evaluate creature-making non-creature spells in similar fashion?
Could we evaluate lands (CMC=0) that generate creatures, too? For example, how might we look at Khalni Garden, or even Elemental Uprising?
Thanks in advance, and I appreciate your hard work!
March 4, 2016 12:36 p.m.
This is to evaluate creatures. Non-creature spells get a little more off in the weeds. As for the three you mentioned, using the above ability values we get:
Scion Summoner: P=3 (2+1 from token) T= 3 CMC=3 A=.5 (this part is subjective, you are trading one creature for one mana)
7/6 or about 1.12
Eldrazi Skyspawner: P=3 T=2 CMC=3 A=2.5 (2 from flying)
7.5/6 or 15/12 or 1.25
Ghirapur Gearcrafter: P=3 T=2 CMC=3 A=2
7/6 or about 1.12
March 4, 2016 1:50 p.m.
Scion Summoner: Whoops
6.5/6 or 13/12 or about 1.08
Making it the worst of them.
March 4, 2016 1:52 p.m.
For me I treat the tokens as if they are part of the original payment. So,
Eldrazi Skyspawner = (2+1+2+2)/4=1.75 2 for flying and 2 for the 1/1 Scion
Scion Summoner = (2+2+2)/6=1 An additional 2 for the 1/1 Scion
Ghirapur Gearcrafter = (2+1+2+2)/6=1.17 Additional 2 for the 1/1 Thopter and 2 more for the Thopter's flying.
As for spells that make tokens, if that's all they do, you can evaluate them as if they were creatures. So Allied Reinforcements is a 4/4 for 4, Dragon Fodder a 2/2 for 2, Raise the Alarm a 2/2 for 2 with Flash, etc.
March 4, 2016 2:21 p.m.
JohnnyBaggins says... #73
I fully get it. I just think the card seems too good looking at the rating. A guy sits with a pack of Khans and has Swiftspear, Warden and Rhino... picks Swiftspear. Would pick Warden over Rhino. Abviously, nobody in their right mind would do that, but still. Assume a total newbie who refers to things like the (God-Awful) BREAD or test like these. Vanilla-Test would tell them: Yeah, all three are decent to play on curve. All "pass". This test would somehow just tell them "Yeah, X is better" - even though that's not what this test actually says. It actually says "This is absolutely nuts to cast on curve" and "This is nuts to cast on curve". A newbie would, however, never read this test. I'm just thinking out loud.
March 4, 2016 3:33 p.m. Edited.
TheFoilAjani says... #74
Wow, this was a really, really well written article. It put to words what I personally couldn't (I use a similar process in my head, but I can't articulate it). I think I want to start a GDoc analyzing recent sets. Would you be interested in taking part in that? Again, congrats on this breakthrough Wiley.
Also, idea; conditional upsides that trigger fine enough are 1, conditional upsides that trigger rarely are 0.5. I know you would prefer they all be integers, but I think it might work fine.
March 5, 2016 4:15 p.m.
TheFoilAjani - Thanks for the feedback! As far as a google doc, I'm leaning more towards using a Spreadsheet than a Doc so it's easier to search through. But that would be something I would want to participate in.
And I still want to stand firm on my "integers only" clause. I'd be willing to modify it somewhat, but the numbers you listed seem fine.
March 5, 2016 11:47 p.m.
TheFoilAjani says... #76
JWiley129 Herp, of course a GSheet. My bad. But yeah, awesome. I have a lot of free time this weekend, so I might be able to start the project.
March 6, 2016 1:38 a.m.
AwesomeSean says... #77
This is probably a dumb question but how would the MV equation apply for alternate/other casting cost like Delve or madness? Would you just change the CMC part? I feel like it would have to lead to adding more variables such as land in play compared to how many cards are in the graveyard.(?)
March 6, 2016 12:58 p.m.
JonathanSamurai says... #78
Curious how you would evaluate Bushido, Flanking, Wither and Infect? This is an amazing article! I am a mostly casual player, as most of my decks are modern but the FNM's in my area are dominated by Standard. I have played a few drafts but always had really bad luck/skill at building competent decks. I think this article helps people like myself who lack experience in Limited to get better at evaluation. Thank you!
March 6, 2016 1:54 p.m.
AwesomeSean - Cost reducers (e.g. Delve) and Alternate Casting Costs (e.g. Surge, Madness) are very hard to evaluate. I think I've come to a decision about alternate casting costs which I'll probably put in my next article. Cost Reducers require a little more work, but I've put my thoughts on Delve above.
JonathanSamurai - Bushido I put as 2x where X is the Bushido number, Flanking seems like a 2 on the Ability Modifier, Wither & Infect are harder to evaluate but I'd put them at 1 for now.
And wow guys, I'm still blown away at the comments and feedback you've given. I'm leaning towards the next article being an analysis of the Shadows over Innistrad creatures, but we'll see.
March 6, 2016 3:19 p.m.
DragonFaceEater says... #80
So I was really interested by your article (though my friends thought it was stupid), coming up with a math equation to determine whether a card is good or not. I decided to find some cards with abilities you have not yet covered, and rate them.
Abattoir Ghoul: 0.875 (I decided his "whenever he kills a creature" ability to be worth half lifelink, so 1)
Accursed Spirit: 0.875 (I decided Intimidate was worth 2)
Araba Mothrider: 1.375 (I decided Bushido is worth it's value times two minus 0.5 [2V-0.5 if you want to get compacted])
Arashin Foremost 1.25 (I decided that the last ability is worth 1.5.)
Here are some downsides that I rated (the score is how much you subtract, not add).
At the beginning of the end step, sacrifice (Ball Lightning) = 3
Vanishing X (Aven Riftwatcher) = 5 - X
When ETB/As an additional cost, sac X creatures (Accursed Centaur) = 2X
Whenever it becomes the target of spell or ability, sacrifice it (Crystalline Nautilus) = 1.5
Echo (Goblin War Buggy) = 2
Legendary (Nicol Bolas) = 1.5
Attacks each turn if able (Riot Piker) = 1
Hope this helped!
March 6, 2016 9:50 p.m.
ToolmasterOfBrainerd says... #81
I don't think legendary is a drawback in a format where it's almost impossible to get 2 copies of anything beyond an uncommon. Outside of some crazy cube, I'd be impressed if legend rule ever came into effect in limited.
March 6, 2016 9:57 p.m.
So DFCs? Do you just count the flip + flip abilities as a bonus A of some value?
March 7, 2016 9:23 a.m.
Let me know if you need any help. I'll certainly be thinking about it.
March 7, 2016 9:30 a.m.
I think Goblin Guide basically breaks the vanilla test...
It gets at least a 2.5, if not 3 because of its other ability.
March 8, 2016 8:16 a.m.
DragonFaceEater says... #87
Follow-up for my last comment about downsides:
When enters the battlefield, lose X life = 0.5X
At the beginning of upkeep, deals X dmg to you = X1.5
When enters the battlefield, opponent gains X life = X1.4
Cumulative Upkeep X = 3.3X
Defender = 2
Now for upsides:
: +1/+0 until end of turn = 1.5
When attacks, gets +X/+Y until end of turn = X + Y0.6
Renown X = X
Tap: add X mana to mana pool = 1
Landwalk = 1
Heroic - Put X +1/+1 counters on this creature = X
Graft X = X2 + 1
Protection from color = 1
Tap: deals X damage to creature/to target player = X1.5 + 1 if creature or player (as opposed to just target player)
X mana: Regenerate = If X is 1, Value 2. X is 2, V 1.5. X is 3, V 1. X is 4, V 0.5. X is 5, V 0.2. X is 6 and over, V 0.1.
Exalted = 1
Extort = 2
Well, that's all I could find that you would see on more that two creatures! Hope it helped!
March 8, 2016 8:17 a.m.
DragonFaceEater says... #88
sorry, for deals damage to you and opponent gains life is supposed to be X0.5 and X0.4
ChiefBell says... #1
JWiley129
March 1, 2016 8:42 a.m.