Well, you're not wrong that the best blue cards are pre-modern. There's a reason blue dominates legacy/Vintage and is much less represented in Modern. It's not quite the same thing in cube, of course, because the specific cards that help blue in Legacy (most importantly Force of Will) aren't quite as important to the color in cube, but the fact is blue used to be super overpowered and other colors have had their time in the limelight since then.
That said, you have to remember that draft is a self-correcting format. If each color is even remotely playable (which is hard not to be the case in cube), then it will be played, but simply by fewer players, and it should help to even out the winrates. As such, it doesn't matter that much that colors aren't perfectly balanced. They certainly aren't in most powered (or unpowered but legacy) cubes, where blue is clearly the best color.
If they're too unbalanced for your tastes, here's a few ideas.
First, since you're still building your cube, you might want to concentrate on blue upgrades for a while. If you ever get to a point where you're satisfied with balance, you can then try to maintain it, which should be easier.
Second, perhaps the color has issues beyond just card quality. For instance, consider what kind of deck your blue cards go into, and whether you can cut some good cards that are "orphaned". Consider something like Chasm Skulker. It's a perfectly fine midrange or aggro card, but if the rest of your blue section is really geared towards hard control and combo, perhaps it doesn't fit.
Third, if your issue is that it isn't enough of a main colour, you could try to include cards that really reward for going really heavy on blue Cryptic Command is an obvious one you might already have, but Thassa, God of the Sea and Master of Waves are huge blue devotion rewards (though I should add that you should also support them with other heavy blue permanents, but you get the idea).
And finally, here's the big thing: your job as a cube designer is to make a fun draft environment. That's it. If you feel blue is missing something only available from older cards, you should feel free to add cards as you wish. Especially those cards that you feel would improve the color, but wouldn't really be weird in modern. Consider a card like Memory Lapse. It's obviously a strong piece of blue interaction, and while it's not legal in modern, similar cards exist there (Remand, Lapse of Certainty) so it wouldn't feel out of place.
Making a cube is kind of like dungeon mastering a game of DnD, if you've ever done that: rules are cool, but ultimately they're just a framework, and the real game is making a fun experience, which sometimes means cheating a little.
I tried the card. It's not bad, but it's not good, either. Here's why.
In absolute magic terms, 1 mana for a 4/3 is much better than one mana for 4 damage. In limited, the first would be format destroying but the second would still be pretty bad. So let's assume that your opponent is a robot that always chooses 4 damage, even if they're at 4 with a lightning bolt in hand.
Now, the question I want to ask is, how often does the extra point of damage as compared to lava spike make you win games you would have lost with lava spike? Asked differently, how often, were you playing lava spike in the place of VD, would your opponent finish at exactly one life¹? And don't forget, that doesn't mean that your last burn spell that specific game made him go to 0, but your last burn spell you could have ever cast in the game (meaning if you weren't dead the turn after and drew another burn spell that then killed your opponent, VD didn't actually help you) In my experience, this doesn't happen very often. The vast majority of the time I play both with and against burn, the margins are bigger than that. You might say I don't have numbers, but than again I don't think you have detailed statistics on final life totals in burn matchups either.
Next, consider opposite question: how often will the opponent's choice of it being a creature, cost you the game? The most obvious scenario for this, of course, is you're playing against a combo deck, they're at 3 and VD is your last card in hand, they let it stay in play, and then they go off the next turn. In my experience with vexing devil, this happened reasonably often. Not just against combo decks, but also against red white decks that drew Lightning Helix in that spot, or heck, just had any removal spell and eventually killed you because the next turn you happened to draw a land.
This, on the other hand, has happened to me quite a few times playing both with and against the card.
But that's not even the only way VD can bite you. For instance, imagine your specific draw of burn involved either only one creature (say, a goblin guide, but anything will do) and a bunch of spells, or two creatures, including a VD. If your opponent drew two removal spells that game, and had spare mana to cast it, you've effectively discarded a card for no effect when you could have had a lava spike instead. How often will that happen? Not very often, you're right. Most of the time killing it will be at least a small tempo loss. But when it does happen, you've effectively mulliganed to 6 that game, which is absolutely huge. One less card means something like 10 to 15% less win rate.
And there's plenty more scenarios like that where your opponent having the option of giving you a 4/3 will be better for them.
Now, none of them or deal-breakers, individually. The percentage chance of each of these happening is fairly low. But if you're rebutting each of those scenarios individually, you're not seeing the full picture. The important equation is this:
Is (% Chance that your opponent ends the game at 1 because you played lava spike instead of VD) > (% Chance you lose the game because of all the different reason outlined above)
The answer to that question is no, and that's why you shouldn't play VD.
One last thing. You're going to say "But I already play Lava Spike!", which is true. But lava spike isn't exactly Burn's best card, either. In fact it's probably it's worst, with Rift Bolt; you just need it because it's cheap. And so that's probably what you're going to replace with VD. Even if you disagree, there's more cards close to Rift Bolt you could cram in burn that aren't being played right now, like Shard Volley or Bump in the Night. Put another way, I don't think that if they printed a new lava spike with a different name you'd see a lot of it in lists.
1: Or more than that, if you drew more than one devil, but your odds of that are lower than not drawing a devil at all
I've actually been impressed over time how much those cards (mostly Kiki, but sometimes also Twin) do great work in normal decks with one or no combo pieces.
But if I were to fit this card, I would have to take into consideration that a few people complain about Bribery, Aetherling, and True-Name Nemesis for being un-fun. Which would be the most appropriate cut?
I think I dislike Bribery more because of how variable it's power level is. Against mono-Red it's extremely underwhelming, while against grenn decks it's completely game-breaking. I don't run it for that reason alone.
On topic: this card looks absurd and I'm running it.
Yeah, super excited for this as well. It's unfortunately not in the right colors for the kiki/twin package, but sometimes those decks splash for Restoration Angel, so it might fit. Also an all-star in the blink decks, as has been said before.
No. there is a reason nobody ran Kiki anymore: it's both too slow and too fragile. Hell, it even makes your mana more inconsistent.
And by "too slow", I don't mean just killing a turn later, but also that it's that much harder to hold up protection while going for it, which is a huge deal.
You could just limit the amount of fixing players have access to. Looking at your cube, you currently have fetches, shocks, painlands, checklands, BFZ duals, fastlands, manlands and tri-lands. The latter are basically only good at supporting slow, greedy decks; cutting those might be a good start. You could also replace your 10 signets/talisman by other rocks (Diamonds, Coldsteel Heart, Mind Stone) that are less good at fixing.
Heck, in your green section, Harrow's only real upside compared to good ol' Rampant Growth is that it fixes more; that could make for an easy swap.
For what it's worth, I don't think it's very much fun to build a 5c monstrosity only to be randomly blown out by a price of progress or a blood moon; I think I'd rather just not be able to amass so much fixing that the triple splash becomes attractive. But hey, it's your draft environment, so take this with a grain of salt.
Yes, holding up 4 mana every turn to try and prevent an opponent from attacking with his 'Goyf or something is bad.
But come on.
You never activate manalands when you have something better to do anyway. This is true even for the best ones like colonnade. The fact is, playing a spell is almost always better than either threatening a block or getting some damage in. So what Neck Snap is a bad card? you're getting it from one of your lands on turn 1000.
Also, attacking for 2 really isn't that bad. It'll still kill your opponent just the same as anything else.
I don't know, I really like this. It seems great in abzan v abzan (both in standard and modern). In jund, it's worse than ravine, but much better than Treetop Village, and fixing for B/G is more valuable than R/G.
It's worse than "works poorly" with GSZ and Natural order: since it's devoid, it's not green even the library, and so can't be fetched. Granted, not that you'd want to very often anyway, but there you go.
That being said, now that we have three decent green 6-drops, I don't think I need more top end, especially since this is much worse in stax-ish builds.
That said, you have to remember that draft is a self-correcting format. If each color is even remotely playable (which is hard not to be the case in cube), then it will be played, but simply by fewer players, and it should help to even out the winrates. As such, it doesn't matter that much that colors aren't perfectly balanced. They certainly aren't in most powered (or unpowered but legacy) cubes, where blue is clearly the best color.
If they're too unbalanced for your tastes, here's a few ideas.
First, since you're still building your cube, you might want to concentrate on blue upgrades for a while. If you ever get to a point where you're satisfied with balance, you can then try to maintain it, which should be easier.
Second, perhaps the color has issues beyond just card quality. For instance, consider what kind of deck your blue cards go into, and whether you can cut some good cards that are "orphaned". Consider something like Chasm Skulker. It's a perfectly fine midrange or aggro card, but if the rest of your blue section is really geared towards hard control and combo, perhaps it doesn't fit.
Third, if your issue is that it isn't enough of a main colour, you could try to include cards that really reward for going really heavy on blue Cryptic Command is an obvious one you might already have, but Thassa, God of the Sea and Master of Waves are huge blue devotion rewards (though I should add that you should also support them with other heavy blue permanents, but you get the idea).
And finally, here's the big thing: your job as a cube designer is to make a fun draft environment. That's it. If you feel blue is missing something only available from older cards, you should feel free to add cards as you wish. Especially those cards that you feel would improve the color, but wouldn't really be weird in modern. Consider a card like Memory Lapse. It's obviously a strong piece of blue interaction, and while it's not legal in modern, similar cards exist there (Remand, Lapse of Certainty) so it wouldn't feel out of place.
Making a cube is kind of like dungeon mastering a game of DnD, if you've ever done that: rules are cool, but ultimately they're just a framework, and the real game is making a fun experience, which sometimes means cheating a little.
I tried the card. It's not bad, but it's not good, either. Here's why.
In absolute magic terms, 1 mana for a 4/3 is much better than one mana for 4 damage. In limited, the first would be format destroying but the second would still be pretty bad. So let's assume that your opponent is a robot that always chooses 4 damage, even if they're at 4 with a lightning bolt in hand.
Now, the question I want to ask is, how often does the extra point of damage as compared to lava spike make you win games you would have lost with lava spike? Asked differently, how often, were you playing lava spike in the place of VD, would your opponent finish at exactly one life¹? And don't forget, that doesn't mean that your last burn spell that specific game made him go to 0, but your last burn spell you could have ever cast in the game (meaning if you weren't dead the turn after and drew another burn spell that then killed your opponent, VD didn't actually help you) In my experience, this doesn't happen very often. The vast majority of the time I play both with and against burn, the margins are bigger than that. You might say I don't have numbers, but than again I don't think you have detailed statistics on final life totals in burn matchups either.
Next, consider opposite question: how often will the opponent's choice of it being a creature, cost you the game? The most obvious scenario for this, of course, is you're playing against a combo deck, they're at 3 and VD is your last card in hand, they let it stay in play, and then they go off the next turn. In my experience with vexing devil, this happened reasonably often. Not just against combo decks, but also against red white decks that drew Lightning Helix in that spot, or heck, just had any removal spell and eventually killed you because the next turn you happened to draw a land.
This, on the other hand, has happened to me quite a few times playing both with and against the card.
But that's not even the only way VD can bite you. For instance, imagine your specific draw of burn involved either only one creature (say, a goblin guide, but anything will do) and a bunch of spells, or two creatures, including a VD. If your opponent drew two removal spells that game, and had spare mana to cast it, you've effectively discarded a card for no effect when you could have had a lava spike instead. How often will that happen? Not very often, you're right. Most of the time killing it will be at least a small tempo loss. But when it does happen, you've effectively mulliganed to 6 that game, which is absolutely huge. One less card means something like 10 to 15% less win rate.
And there's plenty more scenarios like that where your opponent having the option of giving you a 4/3 will be better for them.
Now, none of them or deal-breakers, individually. The percentage chance of each of these happening is fairly low. But if you're rebutting each of those scenarios individually, you're not seeing the full picture. The important equation is this:
The answer to that question is no, and that's why you shouldn't play VD.
One last thing. You're going to say "But I already play Lava Spike!", which is true. But lava spike isn't exactly Burn's best card, either. In fact it's probably it's worst, with Rift Bolt; you just need it because it's cheap. And so that's probably what you're going to replace with VD. Even if you disagree, there's more cards close to Rift Bolt you could cram in burn that aren't being played right now, like Shard Volley or Bump in the Night. Put another way, I don't think that if they printed a new lava spike with a different name you'd see a lot of it in lists.
1: Or more than that, if you drew more than one devil, but your odds of that are lower than not drawing a devil at all
I've actually been impressed over time how much those cards (mostly Kiki, but sometimes also Twin) do great work in normal decks with one or no combo pieces.
I think I dislike Bribery more because of how variable it's power level is. Against mono-Red it's extremely underwhelming, while against grenn decks it's completely game-breaking. I don't run it for that reason alone.
On topic: this card looks absurd and I'm running it.
And by "too slow", I don't mean just killing a turn later, but also that it's that much harder to hold up protection while going for it, which is a huge deal.
Heck, in your green section, Harrow's only real upside compared to good ol' Rampant Growth is that it fixes more; that could make for an easy swap.
For what it's worth, I don't think it's very much fun to build a 5c monstrosity only to be randomly blown out by a price of progress or a blood moon; I think I'd rather just not be able to amass so much fixing that the triple splash becomes attractive. But hey, it's your draft environment, so take this with a grain of salt.
But come on.
You never activate manalands when you have something better to do anyway. This is true even for the best ones like colonnade. The fact is, playing a spell is almost always better than either threatening a block or getting some damage in. So what Neck Snap is a bad card? you're getting it from one of your lands on turn 1000.
Also, attacking for 2 really isn't that bad. It'll still kill your opponent just the same as anything else.
I'll try it in modern.
That being said, now that we have three decent green 6-drops, I don't think I need more top end, especially since this is much worse in stax-ish builds.