I played a set of 10 games against my buddy Jonny, 2 maindeck, 8 sideboard, and went 8 wins 2 loses overall. Obviously this is a variance upswing for me, but it's nice to know that I can win this and I'm not just a dead fish.
g1: Win on the play. I Charmed both his Bobs and was able to kill a Liliana by attacking, he ran out of steam at that point.
g2: Win on the draw. Jon stalled on 2 lands, I was able to develop a nice Snapcaster board but couldn't combo kill him, eventually Cryptic Command buried him. Ooze gave him a slight chance.
g3: Win, play. I had a slow hand that just traded resources until a surprise Batterskull got him.
g4: Lost, draw. He curved into Goyf, Liliana, backed by discard was a little to much to deal with.
g5: Win, play. He stalled on 2, triple Pestermite denying a 2 drop stole the game.
g6: Win, draw. He mulls to 5, I'm able to leisurely close out the game.
g7: Loss, play. I got double Tect Edges off my Kiki/Batterskulls :( 24 land or Crucible of Worlds woulda been nice.
g8: Win, draw. He stalled out, I killed a Liliana with Vendilion Clique. Seems that every time I can attack down a Liliana I win.
g9: Win, play. Traded cards in the early game, Cliqued him and saw Oooze, Bob, Thoughtseize, Fulm Mage let him keep it and beat him down after he Thoughtseized and Bob. Dark Confidant is back breaking if it lives on turn 2, but I've noticed Pestermite/Vendilion Clique beatdown can trump Bob if you're later in the game.
g10: Win, draw. Grinded out a long game with Grim Lavamancer and Relic of Progenitus.
After this I would go 0 wins 5 loses against GB Rock on Magic Online, confirming my suspicion that the set of 10 was me getting lucky. I still learned that the first 3 turns are absolutely critical in this match up, if you're able to contain GB Rock during them the mid and late game look very promising but you'll get absolutely crushed their 2 drop and 3 drop are able to stay in play since any removal or discard spell will usually tempo you out at least a full turn.
Grim Lavamancer and Relic of Progenitus pulled their weight in a lot of my games, they where never dead vs GB Rock and either neutered a green creature or in Grim's case traded with some miscellaneous removal spell. If I'm not lazy I'm going to try out another set of 10, possibly after a rebuild of the deck since not everything feels like it's synergistic
Modern: Ofelia Twin
4 Scalding Tarn
3 Misty Rainforest
3 Steam Vents
1 Stomping Ground
3 Sulfur Falls
1 Desolate Lighthouse (I'd play moar if I could)
2 Tectonic Edge (Extremely ambitious but so much value)
4 Sleight of Hand (Working well)
2 Peek (Snap + Peek eot after bluffing the combo attempt)
2 Spell Pierce
2 Cryptic Command (Solves a lot of problems and makes your Snaps absurd, could cut 1)
2 Izzet Charm (I'm biased but it's been pulling it's weight)
3 Lightning Bolt
2 Flame Slash
16 Combo Stuff n Things
3 Snapcaster Mage
2 Vendilion Clique
2 Deceiver Exarch
4 Splinter Twin
1 Kiki-Jiki, Mirror Breaker
1 Dismember (For Linvala, Spellskite, and random removal, could be a Combust)
1 Ancient Grudge (For Value)
3 Blood Moon (For UWr, the 3rd is due to Wear/Tear and needing to occasionally cheese out Jund)
2 Batterskull (So hard to kill, occasionally rofl stomps opponents)
2 Engineered Explosives (The Goyf Slayer)
3 Grim Lavamancer
3 Relic of Progenitus
Changes since last post:
Relic of Progenitus + Grim Lavamancer vs GB Rock. Relic's are to contain Tarmogoyf, Deathrite, and potentially Scavenging Ooze. The Grim Lavamancer's I've found very effective vs Dark Confidant, Goyf, and Deathrite. Try to prioritizing removing creatures from both graveyards since a "late game" Scavenging Ooze can get outta control otherwise, and also hoarding card types in your bin that Deathrite Shaman can't eat is a potential soft exploit. Since you control the effects of both cards I haven't found them dis synergistic with Snapcaster Mage over the course of 11 sideboard games (not the biggest sample size mind you). These numbers are currently very high, they'll probably be brought down later.
Sleight of Hand > Serum Visions: This is basically a tempo based decision, when I'm casting my cantrips I need to draw gas immediately to present as many options as possible so I can get some type of tempo or board advantage. I'm older more combo focused versions of Twin like my old Grixis Twin list that I came in 19th with Serum Visions was far stronger since the plan was to combo kill and looking at as many cards as possible, even at a lethargic rate, was superior to getting a good card now. Where as this variant is split between controlling the pace of the game and combo killing, and in other to control the game you need to see relevant cards as quickly as possible.
Minus the Combo Package: This was mostly to make room for the Grim/Relic package, things are still in flux. Though one things I've noticed is the match ups where I boarded in more combo where relatively good match ups, so it might be win more, the exception being Burn.
-1 Desolate Lighthouse, -1 Cryptic Command
+2 Fetchlands, or +1 Fetchland +1 Mountain
I've noticed Grim Lavamancer's don't come online as consistently or quickly as I'd like on occasion, roughly 1/5 games. Also the 3 colorless land count has occasionally made forcing the combo slightly harder then it should be. I don't want to cut the Tectonic Edges since they help hedge against flooding, provide mana choke tactics, and are critical vs Treetop Village, Celestial Collonade, and Raving Levine. Lastly hitting 3 mana is critical in Twin, so every land is super helpful.
+1 Izzet Charm
-1 ?? I have no idea
If my plan vs GB rock is to contain them the first three turns Izzet Charm has always preformed in this regard, Goyf living for a few turns is acceptable but if Dark Confidant flips on card on turn 3 it's over me about 90% of the time.
Pulse of the Grid/Thirst For Knowledge: It's probably not correct in this build at the moment, but if I'm able to commit more cards to the board then my opponent, say like Grim Lavamancer, Spellskite, and Shrine of Piercing Visions (oh how I miss you) you'll be able to consistently keep your hand lower then your opponent while maintaining resource parity so that Pulse will rebuy itself. Also it works splendidly with bluffing the combo attempt.
Guerrilla Tactics: Kills Dark Confidant and the majority of turn 1-2 plays and forces Liliana to not +1, and a bad shock isn't the end of the world to Snapback.
Well that's it for this post, lots of thoughts and stuff left to tweak before the Grand Prix. If you've got any questions comment or email me, till next time!