"Ok, make a Lux Cannon token at the start of your turn, using 4 pristine talismans for mana, gain 4 life. Tap my now three cannons to put a charge counter on each. Unwinding Clock triggers and the copy of Unwinding Clock I made last turn triggers too, untap all artifacts by letting the first one resolve, then respond to the second one. Make a Lux Cannon token using the 4 pristine talismans, gain 4 life. Put a charge counter on all four cannons. Let the other Clock trigger resolve, make a lux cannon token, gain 4 life. Blow up two of your lands with the cannons that have 3 tokens. Put a charge counter on the other two.
...ok, you can draw now."
You can't use unwinding clock like that, you only get one untap during your opponents untap, no matter how many clocks you have.
Except that the "statistics" show a ridiculous scenario where you have 7 cards in hand and are apparently digging because they all suck. What happens when you have 1 card in hand? Suddenly that 12.5% chance to discard a useful card is now a 50% chance.
That math is completely wrong. You can't a priori assume that you'll actually draw your answer. Statistically, casting ravings with no other cards in hand is equivalent to a cantrip in terms of probability of ending up with an answer in hand (which is what's really important). If you didn't find what you were looking for, the flashback gives you a second chance to find it, and with that extra card in hand, it's better than a cantrip.
AKA, Ravings gets worse, and by that I mean much worse, the fewer cards you have in your hand. So, when you need Ravings the most, it is at it's worst. Not the kind of card I'm looking for most of the time. It has it's uses, but claiming stats where you've already got 7 cards in your hand and you're digging doesn't really help your case.
Yes, ravings gets worse the fewer cards you have in hand, but the worst case really isn't bad enough not to have at least 3 in my 60. If I'm getting "desperate", I'd very much rather see ravings than a wrong answer.
Gambit is much more dependent on abusing the proliferate effect to be amazing than ravings is dependent on graveyard shenanigans. In my old pre-rotation deck, gambit effectively read "Draw 3, Rampant Growth 1 colorless, gain 3-6 life" because I'd almost always be proliferating Jace Beleren, Everflowing Chalice, and Tumble Magnet whenever it was cast.
..Why aren't you playing *both*? They're both good cards.
Ravings is an inconsistent dig at best due to the random nature of the discard. You are seeing 2 cards, but you do not get to CHOOSE what is let go. You could very well dig 2, find the answer, and then watch it fall to your graveyard.
That's basic human fear where things you don't control feel more risky than they actually are. Ravings in actuality (as backed up by the statistics) is very consistent at finding what you don't already have, at the cost of stuff that doesn't matter.
If you lack Graveyard interactions, Ravings is not the card to use, plain and simple.
Graveyard interactions make ravings better, but it's still very good, just like alchemy.
However, as a Draw Engine, with no GY interactions, it is inferior due to it's chances of forcing you to discard something you actually need (or JUST drew).
...
In a deck without GY interactions, Ravings is an unreliable dig, as you may actually draw the card you needed, only to discard it a moment later. Additionally, if you are holding cards which are important to your gameplan, it becomes far too risky to utilize for card advantage.
Again, that's a fundamental misunderstanding of what ravings is about. If you've got an answer, then just use it. If you don't, then ravings will give you the best chance of finding it before you die. Cards that aren't useful *right now* might as well be blanks.
if you have only DR in hand, and nothing else then you draw two and discard 1 randomly. Thus, assuming you drew 1 answer and 1 non-answer, the odds of you keeping it in your hand is 50%? because you randomly discard 1 of 2 cards? If you draw 2 answers then the odds of keeping 1 becomes 100%?
And that's only for zero cards other than DR in hand and assuming no flashback?
With A being the number of answers and N being the current size of the library, TT clearly has A/N chance of drawing an answer. For DR, it's rather complicated:
(A/N)*((A-1)/(N-1)) + (A/N)*((N-A)/(N-1))/2 + ((N-A)/N)*(A/(N-1))/2
The first term is the chance both cards draw is an answer, the second term is the chance that only the first card drawn is an answer and it wasn't discarded, and the third term is the chance that only the second card is an answer and it wasn't discarded. simplifying it:
(A*(A-1))/(N*(N-1) + (A*(N-A))/(N*(N-1)
(A*(A-1) + A*(N-A))/(N*(N-1))
(A*(N-1))/(N*(N-1)
A/N
So, the probability that an answer is in your hand after a TT or a DR if you don't already have any cards in hand is exactly the same. The formula could be further generalized for a hand H of dead cards:
(A/N)*((A-1)/(N-1)) + (A/N)*((N-A)/(N-1))*(H+1)/(H+2) + ((N-A)/N)*(A/(N-1))*(H+1)/(H+2)
Which simplifies (slightly) to
(A*(A-1))/(N*(N-1) + (A*(N-A)*(H+1)/(H+2)*2)/(N*(N-1)
TT doesn't even net you a card unless you flash it back, in which case your math on desperate ravings is even worse(i think). When you flash back DR, assuming you got the answer you needed and it's now the only card in hand, you draw two and the odds of you keeping the card you need is 66%(assuming you drew no other/comparable answers)?
If you already have an answer, just use the answer. DR works best when you've got a dead hand. Chapin is spot-on when he says it's the closest thing we have to preordain nowadays, and just like preordain you should try to cast it as late as possible.
Seems hypergeometric to me as well. Breaking it down,
The probability of drawing a 4-of in the first 10 draws is the 1 minus the probability of *not* drawing the card.
The chance of failing to draw the card the first time is 56/60
Given that the first draw failed, the chance of failing to draw the card the second time is 55/59
This gives an overall probability to fail to draw the card in the first two draw steps as (56/60)*(55/59)
In general, the probability that the Nth (starting at 1) draw failed given that all previous draws failed is (56-(n-1))/(60-(n-1))
For the third draw, this overall probability is (56/60)*(55/59)*(54/58)
For the forth draw, this overall probability is (56/60)*(55/59)*(54/58)*(53/57)
This continues on for 10 draws, which gives the overall probability to fail as
(56/60)*(55/59)*(54/58)*(53/57)*(52/56)*(51/55)*(50/54)*(49/53)*(48/52)*(47/51)
Then we ask google to do the basic arithmetic for the answer
My current UR superfriends deck actually started out kind of like this.
Like you said, RSZ is really awesome and wins games out of nowhere. I'd cut a Slagstorm and up the RSZ count to 3. I'm pretty much always happy to see an RSZ and use it almost immediately to remove some annoying threat (bloodghast and vengevine are on top of that list). You don't really need to slagstorm all that often.
I've experimented with both clasps and engines, and never quite happy with the clasps. 4 is so expensive for only one proliferate. Clasp does work a bit better with tumble magnets, though, because by the time you throw down an engine, it's not really relevant to proliferate the magnets anymore.
I also ran 2-3 chandra nalaar, especially with the engine. Before the engine hits she can sort of act like a bad gideon, but with an engine on board and 7RR, you can flame wave for 10 in one shot, which always feels pretty awesome.
Finally, I'd cut some of your huge dudes and put in necropede and/or plague myr. Necropede is just amazing at blocking and letting you survive long enough to really go off, and plague myr makes a turn 3 koth much more consistent. Which you like better depends on your local meta, but it's actually pretty viable to run 4x of both instead of any fatties as well as 4x inkmoth nexus and go straight poison with RSZ as an alternate win-con. Also, kor firewalker is just not scary like it is to most red decks.
Galvanic blast is amazing, it's pretty much bolts 5-8, but even better because you'll always have metalcraft.
I have little jace for card draw, but for mono red culling dais could work quite nicely to refill your hand. You only really need to sac 1 creature to it, and necropede is a fine, fine, target for that. It also protects fatties from threaten or control magic effects. I got sick of getting killed by my own titan.
See Beyond is absolutely amazing in this deck, and is pretty much always an effective draw two. I'm also a big fan of Red Sun's Zenith and almost never unhappy when I see it.
Match 1 - Random Jank
The deck seemed like it was a couple of Draft decks slapped together or something.. pretty much a bye. Nothing much to see. Ended one game by proliferating a Chandra 10(!) times. I don't think I even sideboarded.
2 - 0
Match 2 - WG Allies
I drew into enough red removal to keep the board clear both games. Sided in pyroclasms for the tumble magnets, which made it even harder for them to stick anything. Eventually RSZed for the win.
2 - 0
Match 3 - Naya
Managed to get a rare turn 3 koth on the board and got him down to 8 life, but stalled out and scooped to an Inferno titan I didn't draw an answer for, since I used up all my tumble magnet counters on a vengevine. Game 2 I ended up anti-sideboarding by taking out my koths and a chandra for the pyroclasms.
0 - 2
Match 4 - RDW
Very close games. Game 1 I ended up ultimating all three of my planeswalkers at some point, but the jace ultimate actually got him to the burn he needed in order to finish me off. Game 2 was under control fairly quickly, got plenty of elixer activations and pyroclasm 3-for-1s to seal it. He got a fast start Game 3, but a turn 3 slagstorm took out his damage output enough for me to stabilize.
2 - 1
Landfall doesn't matter if you can't get the Avenger out fast enough in the first place. One battlement in your starting hand is decent, but two is just insane. Also helps tons with aggro if you leave them back when you don't need the mana quite as badly.
Eventually the Terramorphics will become Tarns, but it usually doesn't hurt the deck all that much.
I've experimented a bit with counters, but I find that I want to tap out every turn and my counters sit useless in my hand through the whole game. Stealing stuff with volition can get pretty insane. Steal a Primeval Titan and suddenly you're playing a Valakut deck. Steal a sword and equip it on one of our own dudes for effectively +4/+2.with bonuses.
The 8x poison dudes are the early aggro beatdown and the anti-aggro. With the plague myrs, I have effectively 7 chalices, so getting to 6 mana isn't all that tough. Then the Engine comes down and you proliferate to victory while ramping like crazy. I've also often ultimated Koth or Chandra the same turn they came down.
Metalcraft is fairly reliable, so we've got effectively 8x Bolt++. Also Wurmcoil just wins games all on its own.
The sideboard's very iffy, it's pretty much just a bunch of hate cards mashed together. I'm experimenting with the Gomazoa vs. Hawk/Sword. I could really use more help myself here.
Because of the insane mana production of this deck, I've been thinking about 2-4x Red Sun's Zenith, but haven't gotten around to getting any to really test the idea out. I've also repeatedly put in and taken out the trinket mage toolbox, the mages make the deck more consistent but less aggressive since I have to pull the Necropedes in order to put them in. When I do put it in, I have a full set of mages and 1x each of chalice, collar and elixer, with sparkmages and an extra collar in the sideboard. 1x Voltaic Key might be really good, but I haven't tested it.
You can't use unwinding clock like that, you only get one untap during your opponents untap, no matter how many clocks you have.
That math is completely wrong. You can't a priori assume that you'll actually draw your answer. Statistically, casting ravings with no other cards in hand is equivalent to a cantrip in terms of probability of ending up with an answer in hand (which is what's really important). If you didn't find what you were looking for, the flashback gives you a second chance to find it, and with that extra card in hand, it's better than a cantrip.
Yes, ravings gets worse the fewer cards you have in hand, but the worst case really isn't bad enough not to have at least 3 in my 60. If I'm getting "desperate", I'd very much rather see ravings than a wrong answer.
Gambit is much more dependent on abusing the proliferate effect to be amazing than ravings is dependent on graveyard shenanigans. In my old pre-rotation deck, gambit effectively read "Draw 3, Rampant Growth 1 colorless, gain 3-6 life" because I'd almost always be proliferating Jace Beleren, Everflowing Chalice, and Tumble Magnet whenever it was cast.
..Why aren't you playing *both*? They're both good cards.
That's basic human fear where things you don't control feel more risky than they actually are. Ravings in actuality (as backed up by the statistics) is very consistent at finding what you don't already have, at the cost of stuff that doesn't matter.
Graveyard interactions make ravings better, but it's still very good, just like alchemy.
Again, that's a fundamental misunderstanding of what ravings is about. If you've got an answer, then just use it. If you don't, then ravings will give you the best chance of finding it before you die. Cards that aren't useful *right now* might as well be blanks.
With A being the number of answers and N being the current size of the library, TT clearly has A/N chance of drawing an answer. For DR, it's rather complicated:
(A/N)*((A-1)/(N-1)) + (A/N)*((N-A)/(N-1))/2 + ((N-A)/N)*(A/(N-1))/2
The first term is the chance both cards draw is an answer, the second term is the chance that only the first card drawn is an answer and it wasn't discarded, and the third term is the chance that only the second card is an answer and it wasn't discarded. simplifying it:
(A*(A-1))/(N*(N-1) + (A*(N-A))/(N*(N-1)
(A*(A-1) + A*(N-A))/(N*(N-1))
(A*(N-1))/(N*(N-1)
A/N
So, the probability that an answer is in your hand after a TT or a DR if you don't already have any cards in hand is exactly the same. The formula could be further generalized for a hand H of dead cards:
(A/N)*((A-1)/(N-1)) + (A/N)*((N-A)/(N-1))*(H+1)/(H+2) + ((N-A)/N)*(A/(N-1))*(H+1)/(H+2)
Which simplifies (slightly) to
(A*(A-1))/(N*(N-1) + (A*(N-A)*(H+1)/(H+2)*2)/(N*(N-1)
If you already have an answer, just use the answer. DR works best when you've got a dead hand. Chapin is spot-on when he says it's the closest thing we have to preordain nowadays, and just like preordain you should try to cast it as late as possible.
The probability of drawing a 4-of in the first 10 draws is the 1 minus the probability of *not* drawing the card.
The chance of failing to draw the card the first time is 56/60
Given that the first draw failed, the chance of failing to draw the card the second time is 55/59
This gives an overall probability to fail to draw the card in the first two draw steps as (56/60)*(55/59)
In general, the probability that the Nth (starting at 1) draw failed given that all previous draws failed is (56-(n-1))/(60-(n-1))
For the third draw, this overall probability is (56/60)*(55/59)*(54/58)
For the forth draw, this overall probability is (56/60)*(55/59)*(54/58)*(53/57)
This continues on for 10 draws, which gives the overall probability to fail as
(56/60)*(55/59)*(54/58)*(53/57)*(52/56)*(51/55)*(50/54)*(49/53)*(48/52)*(47/51)
Then we ask google to do the basic arithmetic for the answer
Like you said, RSZ is really awesome and wins games out of nowhere. I'd cut a Slagstorm and up the RSZ count to 3. I'm pretty much always happy to see an RSZ and use it almost immediately to remove some annoying threat (bloodghast and vengevine are on top of that list). You don't really need to slagstorm all that often.
I've experimented with both clasps and engines, and never quite happy with the clasps. 4 is so expensive for only one proliferate. Clasp does work a bit better with tumble magnets, though, because by the time you throw down an engine, it's not really relevant to proliferate the magnets anymore.
I also ran 2-3 chandra nalaar, especially with the engine. Before the engine hits she can sort of act like a bad gideon, but with an engine on board and 7RR, you can flame wave for 10 in one shot, which always feels pretty awesome.
Finally, I'd cut some of your huge dudes and put in necropede and/or plague myr. Necropede is just amazing at blocking and letting you survive long enough to really go off, and plague myr makes a turn 3 koth much more consistent. Which you like better depends on your local meta, but it's actually pretty viable to run 4x of both instead of any fatties as well as 4x inkmoth nexus and go straight poison with RSZ as an alternate win-con. Also, kor firewalker is just not scary like it is to most red decks.
Galvanic blast is amazing, it's pretty much bolts 5-8, but even better because you'll always have metalcraft.
I have little jace for card draw, but for mono red culling dais could work quite nicely to refill your hand. You only really need to sac 1 creature to it, and necropede is a fine, fine, target for that. It also protects fatties from threaten or control magic effects. I got sick of getting killed by my own titan.
3 Koth of the Hammer
3 Chandra Nalaar
3 Contagion Engine
3 Trinket Mage
1 Everflowing Chalice
1 Elixer of Immortality
4 Tumble Magnet
4 Lightning Bolt
3 Slagstorm
3 Red Sun's Zenith
4 Scalding Tarn
4 Halimar Depths
4 Island
12 Mountain
3 Brittle Effigy
4 Negate
4 Shatter
4 Pyroclasm
See Beyond is absolutely amazing in this deck, and is pretty much always an effective draw two. I'm also a big fan of Red Sun's Zenith and almost never unhappy when I see it.
Match 1 - Random Jank
The deck seemed like it was a couple of Draft decks slapped together or something.. pretty much a bye. Nothing much to see. Ended one game by proliferating a Chandra 10(!) times. I don't think I even sideboarded.
2 - 0
Match 2 - WG Allies
I drew into enough red removal to keep the board clear both games. Sided in pyroclasms for the tumble magnets, which made it even harder for them to stick anything. Eventually RSZed for the win.
2 - 0
Match 3 - Naya
Managed to get a rare turn 3 koth on the board and got him down to 8 life, but stalled out and scooped to an Inferno titan I didn't draw an answer for, since I used up all my tumble magnet counters on a vengevine. Game 2 I ended up anti-sideboarding by taking out my koths and a chandra for the pyroclasms.
0 - 2
Match 4 - RDW
Very close games. Game 1 I ended up ultimating all three of my planeswalkers at some point, but the jace ultimate actually got him to the burn he needed in order to finish me off. Game 2 was under control fairly quickly, got plenty of elixer activations and pyroclasm 3-for-1s to seal it. He got a fast start Game 3, but a turn 3 slagstorm took out his damage output enough for me to stabilize.
2 - 1
4x Plague Myr
3x Everflowing Chalice
4x Lightning Bolt
4x Galvanic Blast
2x Volition Reigns
4x Koth of the Hammer
3x Chandra Nalaar
2x Treasure Mage
3x Wurmcoil Engine
2x Contagion Engine
1x Steel Hellkite
3x Valakut, the Molten Pinnacle
2x Scalding Tarn
2x Terramorphic Expanse
3x Island
10x Mountain
4x Pyroclasm
4x Guard Gomazoa
3x Tunnel Ignus
2x Sword of Body and Mind
1x Stormtide Leviathan
1x Hoard-Smelter Dragon
Eventually the Terramorphics will become Tarns, but it usually doesn't hurt the deck all that much.
I've experimented a bit with counters, but I find that I want to tap out every turn and my counters sit useless in my hand through the whole game. Stealing stuff with volition can get pretty insane. Steal a Primeval Titan and suddenly you're playing a Valakut deck. Steal a sword and equip it on one of our own dudes for effectively +4/+2.with bonuses.
The 8x poison dudes are the early aggro beatdown and the anti-aggro. With the plague myrs, I have effectively 7 chalices, so getting to 6 mana isn't all that tough. Then the Engine comes down and you proliferate to victory while ramping like crazy. I've also often ultimated Koth or Chandra the same turn they came down.
Metalcraft is fairly reliable, so we've got effectively 8x Bolt++. Also Wurmcoil just wins games all on its own.
The sideboard's very iffy, it's pretty much just a bunch of hate cards mashed together. I'm experimenting with the Gomazoa vs. Hawk/Sword. I could really use more help myself here.
Because of the insane mana production of this deck, I've been thinking about 2-4x Red Sun's Zenith, but haven't gotten around to getting any to really test the idea out. I've also repeatedly put in and taken out the trinket mage toolbox, the mages make the deck more consistent but less aggressive since I have to pull the Necropedes in order to put them in. When I do put it in, I have a full set of mages and 1x each of chalice, collar and elixer, with sparkmages and an extra collar in the sideboard. 1x Voltaic Key might be really good, but I haven't tested it.