Dies to Removal

Saw this on ChannelFireball.
Finding the Optimal Aggro Deck via Computer Simulation
http://www.channelfireball.com/articles ... imulation/

Very interesting. He discusses it from the point of stronger card pool vs. weaker card pool to find the optimal mix of cards. In a card pool with better cards, his simulations suggest that 20 lands are best since the power of each card is relatively high. Whereas in a weaker card pool, less than 20 lands are best because land flood is punished more severely.

Interesting article; thanks for the link.

One thing I found odd is actually right at the beginning. His third best deck "had a 21st City of Brass instead of the 11th Putrid Leech". That seems like a very strange trade-off -- increase the land count while simultaneously lowering your curve. Any thoughts on why that would be? I can't think of any possible advantage you'd have doing that, which makes me question his methods (though they seem pretty rigorous).

I'm glad he went into detail on the mulligans. That was a very smart approach he took and led to some interesting conclusions.

He didn't give any example code or raw data (which I'd love to see), but my guess is that the 21st land resulted in playing more of his spells on curve. The 20 land version likely had a statistically higher chance to get stuck with cards in hand but no mana to cast them than the 21 land version.

That would be my guess as well. He mentions later in the article how the first deck wanted more lands because the power level of the cards allowed it to fight through mana-flood.

I was quite fascinated with this article and its applications to modern. Here is a "perfect" aggro list I came up with using his numbers:

[deck]Creatures:27
4 Diregraf Ghoul
4 Goblin Guide
4 Gravecrawler
4 Rakdos Cackler
3 Goblin Deathraiders
4 Gore-House Chainwalker
4 Spike Jester

Spells:13
4 Bump in the Night
4 Lava Spike
4 Lightning Bolt
1 Shard Volley

Lands:20
4 Arid Mesa
4 Blood Crypt
4 Marsh Flats
2 Mountain
4 Scalding Tarn
2 Swamp
[/deck]

You actually going to try that, Sasky? It'll be fast, but probably flop to opposing aggro decks.

Nevertheless, I am very interested to see how that list performs.

Same. You die fast to aggro, and pretty much every deck that plays creatures (even pod), because your mana base hurts you so much (you really do need perfect fixing for a "perfect" deck to work, though) and because you don't play much removal, and almost half your shit can't block. Jund seems beatable, though, because they also cut through their life total with the mana base and confidants.

I did not say that the list was performance perfect. It's a "perfect" list based off the article's calculation. It is a starting point but requires heavy tweaking.

I was quite fascinated with this article and its applications to modern. Here is a "perfect" aggro list I came up with using his numbers:

[deck]Creatures:27
4 Diregraf Ghoul
4 Goblin Guide
4 Gravecrawler
4 Rakdos Cackler
3 Goblin Deathraiders
4 Gore-House Chainwalker
4 Spike Jester

Spells:13
4 Bump in the Night
4 Lava Spike
4 Lightning Bolt
1 Shard Volley

Lands:20
4 Arid Mesa
4 Blood Crypt
4 Marsh Flats
2 Mountain
4 Scalding Tarn
2 Swamp
[/deck]

I might be missing something (not that experienced , so if it's staring me in the face please forgive me ) , but how do Arid Mesa , Marsh Flats and Scalding Tarn help? As you don't
need plains or islands , wouldn't those just be better replaced with Mountains and swamps respectively? :S

All three fetch Blood Crypts.

All three fetch Blood Crypts.

Ah of course , thanks a lot

I just copied the manabase I had for a boros deck, but that deck ran grim lavamancers. So yeah, you don't need so many fetches. Derp.