I initially misread this thinking that the loss meant it returned to your hand after resolving, but this still might be good enough with cheap burn spells, discard spells, or of course Time Walk effects.
If it just had the copying upside if I "won" the flip, I'd be interested. But with the drawback of Unsubstantiating your own spells when you "lose" flips, I think I can pass.
It's good with super cheap spells though, and you still get "cast" triggers even if your spell is bounced, so there's hope that the card could be good still. But my gut says it's a miss.
Krark's Thumb in case anyone was wondering where the name came from.
It's pretty good with bolt variants, but I don't think I want a card that basically exists only to make bolts better.
Edit - to clarify, I think any red deck would happily play a card that said you had to pay RR to make all its bolts to 6 split 3 and 3 and 2RR to make all its lightning strikes do the same, which is what this card is on average. But that's the only deck that would play a card like this and there are both much better aggro cards and much better spells matter cards.
Just to clarify the math in case someone was wondering. The first time you cast the spell, you have a 50% of succeeding. If you fail, you cast again, you now have a new chance at failing or succeeding. So effectively, for each spell you cast you now have the following probabilities:
50% chance of paying the spell's cost once
25% chance of paying the spell's cost twice
12.5% chance of paying the spell's cost thrice
6.25% chance of paying the spell's cost four times
Etc
Now, on average this plays out that you will pay the spell's cost exactly twice. Again, this is the "average", i.e., if you cast a hundred spells with this guy in play over many games, it will average out to twice the cost in the long run. Effectively, this card is "pay twice, get twice the effect", which seems great. Of course, there is a huge variance tacked on that will greatly impact how the card plays in practice.
This guy seems pretty great with CMC 1/2 spells, especially with cards like Young Pyromancer or Shaeeli, Sublime Artificer. However, he seems worse with reactive spells. Copying a removal spell with only one target nullifies the advantage, and having to cast a removal twice or thrice at an unopportune time can loose you the game. Similarly, he is terrible with counterspells as you pretty much have all the drawback, none of the advantage.
Due to the variance and drawbacks, I can't see him go into that many different decks, pretty much only spell heavy burn decks, and it probably won't find a home in my cube. I also don't play power, so no 2xRecall or 2xTime Walk dream scenarios.
Private Mod Note
():
Rollback Post to RevisionRollBack
Cubing in Oslo, Norway --- contact me if you want to join in!
While he might not be the best card, the average of "double for double" doesn't seem all that bad. Also I like the thought of people repeatedly saying "Double or nothing!" Krark might be just worth it for the fun while having a solid average.
I think I will flip a coin to decide if I include 2 or none.
Yes it would make dark ritual on average BB for BBBBBB
So if I understand this correct:
The computation will stop once it wins, on proceeding iterations, assume all previous flips were loses
Dark Ritual One, add 3 Black, if win, add 6 black, else, return dark ritual to hand. (50% chance to Win - add 6, lose add 3
If Win - add 6, copy = 6 (Net 5)
If Loss - add 3, return ritual to hand = 3
- Lost Coin Flip Cont -
Dark Ritual, add 3 black, if win, add 6 black - 1 for recast, else return dark ritual to hand (25% chance to Win - add 9, lose add 6)
If Win - Coin Flip 1 was loss - add 3, Coin flip 2 copy and add 6 = total 9 (Net 7)
If Loss - Coin Flip 1 was loss - add 3, Coin flip 2 lose, add 3 + return Dark Ritual to hand = total 6
- Lost Coin Flip Cont -
Dark Ritual Recast 2 - Add 3 black, if win, add 12 black - 2 for recasts, else return dark ritual to hand (12.5% chance to Win - add 12, lose add 9)
If Win - Coin Flip 1 was loss, Coin Flip 2 was loss, Coin flip 3 was win = total 12 (Net 9)
If Loss - Coin flip 1 was loss, coin flip 2 was loss, coin flip 3 was loss + Return dark ritual to hand = total 9
.....
This is an infinite geometric sequence. Calculate the expected value at each win -
I brute forced this geometric sequence using a python program and it evaluates to 7B as expected value, which is Net 7 Black for 2 Cast Cost in the average case.
I think your 2B for 6B as the average case is incorrect.
I'm actively maintaining a comprehensive article to help explain to new cube players how some complex vintage level cards work in a cube environment. Vintage Cube Cards Explained
You also run the risk of attempting to cast a spell multiple times, failing both times, and not having the colored mana required to attempt to cast it again ...netting you nothing at the end of your exchange because you can't attempt to cast the spell again at all.
Yes it would make dark ritual on average BB for BBBBBB
So if I understand this correct:
The computation will stop once it wins, on proceeding iterations, assume all previous flips were loses
Dark Ritual One, add 3 Black, if win, add 6 black, else, return dark ritual to hand. (50% chance to Win - add 6, lose add 3
If Win - add 6, copy = 6 (Net 5)
If Loss - add 3, return ritual to hand = 3
- Lost Coin Flip Cont -
Dark Ritual, add 3 black, if win, add 6 black - 1 for recast, else return dark ritual to hand (25% chance to Win - add 9, lose add 6)
If Win - Coin Flip 1 was loss - add 3, Coin flip 2 copy and add 6 = total 9 (Net 7)
If Loss - Coin Flip 1 was loss - add 3, Coin flip 2 lose, add 3 + return Dark Ritual to hand = total 6
- Lost Coin Flip Cont -
Dark Ritual Recast 2 - Add 3 black, if win, add 12 black - 2 for recasts, else return dark ritual to hand (12.5% chance to Win - add 12, lose add 9)
If Win - Coin Flip 1 was loss, Coin Flip 2 was loss, Coin flip 3 was win = total 12 (Net 9)
If Loss - Coin flip 1 was loss, coin flip 2 was loss, coin flip 3 was loss + Return dark ritual to hand = total 9
.....
This is an infinite geometric sequence. Calculate the expected value at each win -
I brute forced this geometric sequence using a python program and it evaluates to 7B as expected value, which is Net 7 Black for 2 Cast Cost in the average case.
I think your 2B for 6B as the average case is incorrect.
And also: How are you supposed to net 7 black mana from 2 black rituals!?
The outcome is always the effect of the card twice. So Dark Ritual will always net BBB BBB with Krark. The only variance is how many coin tosses you will have to do to get the result, which is two on average.
Yes it would make dark ritual on average BB for BBBBBB
So if I understand this correct:
The computation will stop once it wins, on proceeding iterations, assume all previous flips were loses
Dark Ritual One, add 3 Black, if win, add 6 black, else, return dark ritual to hand. (50% chance to Win - add 6, lose add 3
If Win - add 6, copy = 6 (Net 5)
If Loss - add 3, return ritual to hand = 3
- Lost Coin Flip Cont -
Dark Ritual, add 3 black, if win, add 6 black - 1 for recast, else return dark ritual to hand (25% chance to Win - add 9, lose add 6)
If Win - Coin Flip 1 was loss - add 3, Coin flip 2 copy and add 6 = total 9 (Net 7)
If Loss - Coin Flip 1 was loss - add 3, Coin flip 2 lose, add 3 + return Dark Ritual to hand = total 6
- Lost Coin Flip Cont -
Dark Ritual Recast 2 - Add 3 black, if win, add 12 black - 2 for recasts, else return dark ritual to hand (12.5% chance to Win - add 12, lose add 9)
If Win - Coin Flip 1 was loss, Coin Flip 2 was loss, Coin flip 3 was win = total 12 (Net 9)
If Loss - Coin flip 1 was loss, coin flip 2 was loss, coin flip 3 was loss + Return dark ritual to hand = total 9
.....
This is an infinite geometric sequence. Calculate the expected value at each win -
I brute forced this geometric sequence using a python program and it evaluates to 7B as expected value, which is Net 7 Black for 2 Cast Cost in the average case.
I think your 2B for 6B as the average case is incorrect.
And also: How are you supposed to net 7 black mana from 2 black rituals!?
The outcome is always the effect of the card twice. So Dark Ritual will always net BBB BBB with Krark. The only variance is how many coin tosses you will have to do to get the result, which is two on average.
Okay I get it now.
If the spell bounces back to your hand, it doesn't leave behind a copy ...
Assume we're always going to cast the dark ritual, the end result is we get a dark Ritual + Copy = 6 Black and we should look for the expected value in the number of casts required to get to the end result.
As you have corrected pointed out, this should be the same as the expected value for the number of coin flips which is 2.
Therefore, yes, the average case is 2B for 6B.
Private Mod Note
():
Rollback Post to RevisionRollBack
I'm actively maintaining a comprehensive article to help explain to new cube players how some complex vintage level cards work in a cube environment. Vintage Cube Cards Explained
I initially misread this thinking that the loss meant it returned to your hand after resolving, but this still might be good enough with cheap burn spells, discard spells, or of course Time Walk effects.
It's good with super cheap spells though, and you still get "cast" triggers even if your spell is bounced, so there's hope that the card could be good still. But my gut says it's a miss.
My 630 Card Powered Cube
My Article - "Cube Design Philosophy"
My Article - "Mana Short: A study in limited resource management."
My 50th Set (P)review - Discusses my top 20 Cube cards from OTJ!
It's pretty good with bolt variants, but I don't think I want a card that basically exists only to make bolts better.
Edit - to clarify, I think any red deck would happily play a card that said you had to pay RR to make all its bolts to 6 split 3 and 3 and 2RR to make all its lightning strikes do the same, which is what this card is on average. But that's the only deck that would play a card like this and there are both much better aggro cards and much better spells matter cards.
375 unpowered cube - https://cubecobra.com/cube/list/601ac624832cdf1039947588
Also Krark's Other Thumb
If nothing else, the card title did make me chuckle
375 unpowered cube - https://cubecobra.com/cube/list/601ac624832cdf1039947588
This guy seems pretty great with CMC 1/2 spells, especially with cards like Young Pyromancer or Shaeeli, Sublime Artificer. However, he seems worse with reactive spells. Copying a removal spell with only one target nullifies the advantage, and having to cast a removal twice or thrice at an unopportune time can loose you the game. Similarly, he is terrible with counterspells as you pretty much have all the drawback, none of the advantage.
Due to the variance and drawbacks, I can't see him go into that many different decks, pretty much only spell heavy burn decks, and it probably won't find a home in my cube. I also don't play power, so no 2xRecall or 2xTime Walk dream scenarios.
I think I will flip a coin to decide if I include 2 or none.
Cubecobra
---
Magic Cardmarket Profile
---
DCI Rules Advisor
How much time I could save by implementing this decision-process in Cube-building
So if I understand this correct:
The computation will stop once it wins, on proceeding iterations, assume all previous flips were loses
Dark Ritual One, add 3 Black, if win, add 6 black, else, return dark ritual to hand. (50% chance to Win - add 6, lose add 3
If Win - add 6, copy = 6 (Net 5)
If Loss - add 3, return ritual to hand = 3
- Lost Coin Flip Cont -
Dark Ritual, add 3 black, if win, add 6 black - 1 for recast, else return dark ritual to hand (25% chance to Win - add 9, lose add 6)
If Win - Coin Flip 1 was loss - add 3, Coin flip 2 copy and add 6 = total 9 (Net 7)
If Loss - Coin Flip 1 was loss - add 3, Coin flip 2 lose, add 3 + return Dark Ritual to hand = total 6
- Lost Coin Flip Cont -
Dark Ritual Recast 2 - Add 3 black, if win, add 12 black - 2 for recasts, else return dark ritual to hand (12.5% chance to Win - add 12, lose add 9)
If Win - Coin Flip 1 was loss, Coin Flip 2 was loss, Coin flip 3 was win = total 12 (Net 9)
If Loss - Coin flip 1 was loss, coin flip 2 was loss, coin flip 3 was loss + Return dark ritual to hand = total 9
.....
This is an infinite geometric sequence. Calculate the expected value at each win -
E(X) = SUM((1/2)^x * (3 + 2x)) = 0.5*5 + 0.25*7 + 0.125*9 ...
I brute forced this geometric sequence using a python program and it evaluates to 7B as expected value, which is Net 7 Black for 2 Cast Cost in the average case.
I think your 2B for 6B as the average case is incorrect.
Vintage Cube Cards Explained
Here are some other articles I've written about fine tuning your cube:
1. Minimum Archetype Support
2. Improving Green Archetypes
3. Improving White Archetypes
4. Matchup Analysis
5. Cube Combos (Work in Progress)
Draft my Cube - https://cubecobra.com/cube/overview/d8i
My 630 Card Powered Cube
My Article - "Cube Design Philosophy"
My Article - "Mana Short: A study in limited resource management."
My 50th Set (P)review - Discusses my top 20 Cube cards from OTJ!
On average two is correct. Here is an example of "How many coin flips until you get a head?" https://diego.assencio.com/?index=6c1cfa313064329046317358d2aa22c0
See also here: https://stats.stackexchange.com/questions/136806/expected-number-of-tosses-till-first-head-comes-up
And also: How are you supposed to net 7 black mana from 2 black rituals!?
The outcome is always the effect of the card twice. So Dark Ritual will always net BBB BBB with Krark. The only variance is how many coin tosses you will have to do to get the result, which is two on average.
Cubecobra
---
Magic Cardmarket Profile
---
DCI Rules Advisor
Okay I get it now.
If the spell bounces back to your hand, it doesn't leave behind a copy ...
Assume we're always going to cast the dark ritual, the end result is we get a dark Ritual + Copy = 6 Black and we should look for the expected value in the number of casts required to get to the end result.
As you have corrected pointed out, this should be the same as the expected value for the number of coin flips which is 2.
Therefore, yes, the average case is 2B for 6B.
Vintage Cube Cards Explained
Here are some other articles I've written about fine tuning your cube:
1. Minimum Archetype Support
2. Improving Green Archetypes
3. Improving White Archetypes
4. Matchup Analysis
5. Cube Combos (Work in Progress)
Draft my Cube - https://cubecobra.com/cube/overview/d8i