Trying Land Denial vs. Ramp

Approach:
+4x Spreading Seas
+3x Demolish
-4x Burst Lightning
-3x Lightning Bolt

I presumed the Valakut deck went:
-3 Pyroclasm
+3 Acidic Slime, because Acidic slime does a pretty good job of wrecking my deck, and Pyroclasm does nothing.  It is possible that taking out Harrow for Summoning Trap/Fatties would have altered the matchup, because Harrow was pretty beastly, and Summoning Trap + Fatties fall into the trap of the mana denial plan, but whatever.

Participants:
I drafted my girlfriend into playing the Valakut deck, with some assistance from me, as to when things should be done or not.  In general, Valakut is a puzzle to find out how to spend all your mana every turn, between taplands, ramp spells, and so on.  We played four games before she was hearing the clarion call of level 85 in WoW, and I finished out a 10 game set playing against myself.

Reasoning:
The theory behind this approach is that the threats Valakut presents are mostly/all double-green (Primeval Titan, Avenger of Zendikar, Summoning Trap, Acidic Slime, Rampaging Baloths or Terrastrodon if they have them), but the Valakut decks can only afford to fit in about 5 Forests, and typically a Raging Ravine or Khalni Garden to suppliment this.  If you can turn off their first few forests, they may not be able to play any of the "big stuff", and thus, you could have free reign to do whatever you like, and eventually win.

In past games, I've tried to attack their Valakuts, to 10-30% win rate, via Demolish, etc.  We'll see how weak the mana base of Valakut is with this testing.

Results:

1 Win, 9 Losses.

My win was due to a active Ascension, then two active Ascensions, combined with some Spreading Seas, Demolish, Call to Mind, and so on.

Problems:
The losses were a combination of:
A) KHE, Overgrown Battlement come down T2, and provide amazingly steady sources of Green mana.
B) Valakut has multiple ways to set up double, triple, even quadruple-Green if they want it, in a way my deck can't disrupt fast enough.  Cultivate searches out two land and is only G to cast.  KHE is only G to cast, Harrow is only G to cast, and searches out two Forests pretty easy.

Lands like Terramorphic Expanse, and Evolving Wilds allow the Valakut player to "save up" their Forests, while still hitting land drops.  Something as simple as TE -> Go -> Sac end of turn for a Forest -> Play a forest -> Cast Primeval Titan was often an easy win for Valakut.   They have something like 20 ways to get so many forests out at end of turn, or just kill you, that this is not a reasonable plan.

Overgrown Battlement is a nightmare.  They need only a single G to cast it turn 2, that you can't possibly disrupt, and you can't reliably counter it (or risk losing to T2 Primeval Titan off Summoning Trap).  This thing sits in play, takes two burn spells (2RR) to remove, or a bounce COMBINED with killing whatever G source they have at the time.  That's alot of mana to be spending, when they can just drop another TE/EW and cast it in a semi-non-counterable way next turn.

KHE turn 2 is something that virtually guarantees that your land denial plan will not work.  It's pretty high variance for it to not work.  You can bounce it, like Overgrown Battlement, but with the same problems.

Harrow was always good.  Forest -> Cultivate for 2 Forests was game over most of the time.  (Unless you have an active Ascension + Demolish + them having nothing else.

At times, I felt that 4x Acidic Slime was doing a better job denying me options than my 3x Demolish + 4x Spreading Seas was doing denying the Valakut deck of options.  Slime your 1x Island or Mountain, good game?

This plan seems to be a resounding failure.  It's extremely high variance, and will work not at all against non-Valakut ramp decks, which aren't so green-light.  (GU, RUG, MGE).  Time to try something else.

PS:
Attacking the Valakuts tends to not work, because PT fetches them out so fast, and all the ways available to destroy them are sorcery speed, so you can often just be dead the turn they cast PT, between Harrow, Cultivate, KHE, and so on.  PT is the real problem, so we'll be trying to Mindbreak it in the future.