General Gold aggro and rethinking colour roles

Dom Harvey

Dom Harvey

Contributor
We've had a lot of debates recently about how much fixing to run and what it should look like, as well as the problems with certain colour pairs.

A long time ago we complained about the Gravecrawler-Griselbrand problem - most Cube designers assign the same number of cards to each colour as if that's the natural dividing line, even though many of those cards rarely go in the same decks. This new Chandra looks sweet - do I cut Jackal Pup or Wildfire? How is that a sensible question?

We've largely accepted that's a bad way to look at things, but I'd argue the same applies to gold cards. Let's say I have 5 slots for each guild. Do I want my BG section to be aggressive? If so, do I play Putrid Leech? Lotleth Troll? Grim Flayer? These are all great in their own ways but each does different things and sends different signals. If I run all of them, I can't also support a more controlling BG deck with Pernicious Deed/Meren or find space for BG's good generic removal in Abrupt Decay/Maelstrom Pulse. Maybe I make this easier by only supporting half the guilds, but that only opens up a few more slots and the problem remains.

I've seen opposite ways to handle this. Either you accept that gold cards are narrow, and fill those slots with cards that were narrow anyway, or you aim for balance by playing gold cards that everyone in that colour pair wants and others might splash for.

In a seemingly unrelated problem, it's hard to push aggro in colour pairs where the pool of good 2-drops (or 3-drops, or...) is shallow or those cards only work in specific archetypes. When they are good, that doesn't make them interesting. Gold 2-drops tend to have a lot more depth as a reward for the restrictive mana costs: compare Borderland Marauder to Voltaic Brawler. Many are less restrictive because they can star in other theatres - Abzan Control decks in Standard and Block happily played Fleecemane Lion when Kalonian Tusker would be booed off stage. The aggressive Delirium deck in Standard plays Grim Flayer, but it's a big draw to the controlling version too.

We've also learned the hard way that fixing is crucial for aggro decks. If your cards lose relevance quickly, being unable to cast them on time is brutal. So we want to play lots of fixing and, as il duce taught us long ago, AB cards are less taxing on a two-colour manabase than AA or BB. By shifting some of the aggro focus into the gold sections, we take two areas of Cube that are largely self-contained and have them overlap, creating more space for less narrow cards and making those aggro decks more powerful/interesting. Meanwhile, between equipment, Vehicles, and good artifact creatures, there are lots of universally playable aggro cards to alleviate concerns about gold cards being siloed.

This approach also lets you explore multiple decks that all require (or would like) support from gold cards. If I get to expand my gold section, GB can support +1/+1 counters and Delirium at the same time, or...

(This can inform control decks too. The pushed gold cards in the various multicoloured blocks led to some really sweet control decks in their time, and control faces the same dilemma as above: why should I choose between Fire // Ice, Izzet Charm, Electrolyze, and Prophetic Bolt?)



TL;DR:
- Aggro is narrow and requires fixing
- Gold cards are narrow but sweet and make aggro better; they also require fixing
- Play lots of fixing and gold aggro cards
 
sigh

sigh

Super interesting thread! I like it :). Its interesting that I've consistently stuck to my guns when it comes to keeping each of the five color sections the same size, but I've been fairly lackadaisical regarding my multicolor balance. I do think that this is a good place and opportunity to provide case-by-case support and anchors for different themes. Some particular stuff that's on my radar/next round of inclusions for filling holes, bolster themes, etc:


I know not all focused on aggro, but it is what it is. I could see keeping my weapons trainer in my RW section (add trenches) to help keep RW aggro more relevant. Could definitely also use another tool for WB D&T/aggro...
 
  • Like
Reactions: dbs
Jason Waddell

Jason Waddell

Administrator
Staff member
Okay, so, Dom linked in the OP an old debate between CML and I that was basically to the effect of, which card is harder to cast between a CC card and a CD card. Let's say we have:



For simplicity, say we have a deck with 11 lands, distributed as follows.
4 Forest
4 Plains
3 Savannah

Consider all of our two land combinations.

If our first land is:

60% of our next lands allow us to cast Anafenza (3 Plains, 3 Savannah)
70% of our next lands allow us to cast Fleecemane (4 Forest, 3 Savannah)

If our first land is:

60% of our next lands allow us to cast Anafenza (4 Plains, 2 Savannah)
100% of our next lands allow us to cast Fleecemane (4 Forest, 2 Savannah, 4 Plains)

If our first land is:

0% of our next lands allow us to cast Anafenza
70% of our next lands allow us to cast Fleecemane (3 Savannah, 4 Plains)

In all cases, the CD card is more castable. Assuming equal power level (ignore the actual cards for a moment) there's an argument that Anafenza is in "higher demand" because "all white decks can cast her" and "only GW decks (objectively least cool guild) can cast Fleecemane, but I don't even know if that's a valid argument. Say you put functionally identical cards in your cube, one costed at CC and costed at CD, which would see more play? If I'm in {R}{W}, I'm still hesitant to play a {W}{W} card. But probably somebody at the table is happy to play the {G}{W} card. I'd guess that the CD card gets more maindeck action, and this is before you even take into account that CD cards tend to be higher power than CC cards. Notable exception: Hymn to Tourach. But Hymn to Tourach would be played soooo much if it was costed {R}{B}. Searing Blaze is nice too, probably similar power level to Lightning Helix, but more narrow.

Anyways, I think it's interesting to think about if you're looking at your cube contents from the perspective of their demand.
 
japahn

japahn

One thing from Jason's old post is worth repeating:

With two random basic lands in a two-color deck (with the same number of basic lands in each color), the possible combinations are:
Two plains.
Plains and mountain.
Mountain and plains.
Two mountains.

{1}{W} has 75% chance of being castable.
{W}{R} has 50% chance of being castable.
{W}{W} has 25% chance of being castable.

Can someone do the math on how the odds change, given 16 lands and n duals (0 <= n < 6)?
 
Miasmir

Miasmir

Okay, so we have here some rough data with some ASSUMPTIONS.
We're a White-Red aggro deck, leaning ever so slightly to White.
We're looking at the odds for any possible 8-cards-to-have-seen-by-turn-2-on-the-play, and finding out which of those hands have:
a) at least 2 lands
b) not 8 lands (so the other card is a 2-drop)
c) the correct colors to cast said 2-drop.
The funniest assumption here is that for, say, the {W}{R} column, every card in our deck is assumed to be either a land or a card costing {W}{R}.

The numbers given in japahn's post assume a 2-land hand.
So our numbers will be different because we assume you are (for whatever reason) keeping hands like Plains, Plains, Plains, Plains, Plains, Mountain, Truefire Paladin. You're completely all-in on playing a 2-drop on turn 2. Or rather, we count that hand as a 'success' because it plays a 2-drop on turn 2. We aren't really incorporating mulligan strategy into this analysis.

For odd numbers of duals, we assume the extra land is a Plains, for our primary color. When we cut a Plains for a dual, we don't expect the chance of a WW to go up by much. This accounts for the 'stepping' seen in the table, where the numbers change every other entry.

Code: 
Duals        1W        WR        WW    This last column is WRONG;
0            83%        74%        51%    see my follow-up post below for details.
1            86%        77%        61%
2            86%        80%        61%
3            88%        82%        70%
4            88%        84%        70%
5            89%        85%        77%
6            89%        87%        77%
7            90%        88%        82%
8            90%        89%        82%

Coming soon: A restriction of this analysis to hands you might ever keep, and a way to incorporate having cards in your deck other than lands and 2-drops.
 
Miasmir

Miasmir

Due to a bug, the previous table actually OVER-represents the probability of hitting a WW 2-drop.

So here's that table again with the bug fixed:
Code: 
Duals        1W        WR        WW
0            83%        74%        51%
1            86%        77%        59%
2            86%        80%        59%
3            88%        82%        66%
4            88%        84%        66%
5            89%        85%        72%
6            89%        87%        72%
7            90%        88%        77%
8            90%        89%        77%

As Jason requests, here are tables assuming a 2-land hand, a 3-land hand, and so forth. We're still assuming every nonland card in the deck is a 2-drop but it's less ridiculous to do so when we have more cards in our hand that could be that 2-drop.

You'll note that some of the chances are higher for specific land numbers than in the above table. That's because we're removing the probability of a 0-, 1-, or 8-land hand, each of which are always failures. At high numbers of dual lands, the chance of failing to hit colors becomes small enough that hitting the right number of lands is a greater concern. So much so, in fact, that we need an extra decimal place to represent just how close to certain we are to hit our colors with many duals in a land-rich hand.

Code: 
Assuming 2-land:
Duals        1W        WR        WW
0            76.7%    53.3%    23.3%
1            82.5%    59.2%    30.0%
2            82.5%    65.0%    30.0%
3            87.5%    70.0%    37.5%
4            87.5%    75.0%    37.5%
5            91.7%    79.2%    45.8%
6            91.7%    83.3%    45.8%
7            95.0%    86.7%    55.0%
8            95.0%    90.0%    55.0%
 
Assuming 3-land:
Duals        1W        WR        WW
0            90.0%    80.0%    50.0%
1            93.8%    83.8%    60.0%
2            93.8%    87.5%    60.0%
3            96.4%    90.2%    69.6%
4            96.4%    92.9%    69.6%
5            98.2%    94.6%    78.6%
6            98.2%    96.4%    78.6%
7            99.3%    97.5%    86.4%
8            99.3%    98.6%    86.4%
 
Assuming 4-land:
Duals        1W        WR        WW
0            96.2%    92.3%    71.5%
1            98.1%    94.2%    80.8%
2            98.1%    96.2%    80.8%
3            99.2%    97.3%    88.2%
4            99.2%    98.4%    88.2%
5            99.7%    98.9%    93.7%
6            99.7%    99.5%    93.7%
7            99.9%    99.7%    97.3%
8            99.9%    99.9%    97.3%
 
Assuming 5-land:
Duals        1W        WR        WW
0            98.7%    97.4%    85.9%
1            99.5%    98.2%    92.3%
2            99.5%    99.0%    92.3%
3            99.9%    99.4%    96.4%
4            99.9%    99.7%    96.4%
5            99.98%   99.84%   98.72%
6            99.98%   99.95%   98.72%
7            100%     99.98%   99.89%
8            100%     100%     99.89%

Those 100%s are not rounding errors; there is literally no way to miss with that many duals in a 5-land hand. If your deck contains 8 Plateau, 4 Plains, and 4 Mountain, you can't make a 5-land hand that doesn't hit {W}{R}.

How's everyone feeling? Any statistics deluge fatigue setting in? Want more spreadsheets?

I just realized I got this far while writing {R}{W} in the wrong order. :eek:
 
Dom Harvey

Dom Harvey

Contributor
What would a a more complete taxonomy look like?

- W, U, B, R, G generic: Tarmogoyf, Doom Blade, Blade Splicer. Cards that cost 1W, R, 2U, etc. that don't have additional colour requirements and can go in most decks of that colour. Most of the 'good-stuff' filler goes here. These are the most broadly applicable cards; if these are also the best cards without extra effort, the optimal strategy quickly becomes taking as many of these as you can support

- W, U, ... aggro: Kytheons and Goblin Guides, cards that can join up with aggressive cards in other colours but are largely confined to that theatre

- WW, UU, ... time-sensitive: I say 'time-sensitive' because these face the unfortunate tension of needing to be cast early but being harder to cast because of their mana costs. These aren't aggro-exclusive - because of that tension, many of these cards are buffed such that they can slot into midrange decks of their colour - but they tend to be. Precinct Captain is great in any deck if you cast it on Turn 2; a top-end card like Sphinx of Lost Truths gives you more time to find UU and doesn't need to be cast in a specific time-frame. These fit into two colour decks if you have some good fixing

- WWW, ... high-commitment: Geralf's Messenger, Boggart Ram-Gang, Boros Reckoner, but also Student of Warfare/Figure of Destiny. These ask you to be based in the colour and not run colourless lands or off-colour lands if you can avoid it. A BW deck that wants to play Messenger will have to be black-heavy with a light splash: certainly no white 1-drops. Ram-Gang/Reckoner fit in multiple bins here - high-commitment red and high commitment (green/white respectively)

- WU, GB, ... cheap: Fleecemane Lion, Putrid Leech. This is the original focus of this thread: the tradeoff that you get with 1W vs WW cards doesn't apply linearly here, as these tend to be more powerful/interesting and easier to cast in the right context. A lot of the gold 4-drops - Falkenrath Aristocrat, Bloodbraid Elf - are classic midrange threats but also great curve-toppers in aggro

- WU, GB, ... utility: Lightning Helix, Maelstrom Pulse. Familiar effects jazzed up and more versatile.

- WU, GB, ... expensive: Prophetic Bolt, Sorin Grim Nemesis, Mirari's Wake, Bituminous Blast. EDH-style, Battlecruiser Magic that was more common a few years ago and deserves a second chance IMO

- Artifacts: obviously more versatile on average

And then you have a cross-cutting archetype taxonomy
 
Toan

Toan

I feel like I should try making some more custom cards with cascade.

Stuff more in line with Shardless Agent though (Ie; Vanilla creatures/"vanilla" non-creatures that have cascade on them).

The current cascade offering is modest at best, and I suspect it's due to power level concerns.
 
W

whobetta

like the simplistic breakdown from Jason's small land sample size. My question is can anyone help create a table where depending on your COLOR SOURCE breakdown and total what % your next land would be to cast a WW or WG card?

for example

- 10 Plains
- 4 forest
- 1 Mountain
- 1 W/G/R tri color
- 1 W/R search land

say splashing a 3rd color or just have a higher % of one color lands... could playing the Anafenza vs Fleecemane be a different argument?
 
Alfonzo Bonzo

Alfonzo Bonzo

-> thread

I love this post, and it's a shame we devolved into coloured casting cost statistics at the time (can we not again). I've been giving this some thought for a couple of reasons. I've been wresting with the supporting interesting aggro thing and I think there's a lot here. I also subscribe to the school of thought that I don't really want multiple boring 2 power one drops to support aggro (a different argument not for here).

Dom makes reference to it in his post but that you can help to reduce numbers by playing decdent artifact creatures, equipment and vehicles to support aggro. This would in theory lead to decks being similar due to the prevailance of these, but the argment for aggro gold two drops helps to make them feel more distinctive.

What are some of the aggro two (or three drops) which you like (and would like more if you had less restrictions on your gold section)?
 
You must log in or register to reply here.
Top