Once again, just cutting some commons and uncommons that I don't think do enough and putting in some lands instead!

I've removed 3 commons (9 total cards) from the pool and dropped the scry lands from 2-ofs to 1-ofs. This is to accommodate the inclusion of Fetchlands and Shocklands, which will be proxied into the cube. These, along with Ral, are indeed non-theros cards. However, the cube currently desperately needs the mana-fixing.

As it stands, all of the cards in the cube come strictly from one of the 4 Theros sets. This change marks a divergence from that with Ral, Storm Conduit. The UR Keranos deck is pretty terrible as it stands, split between wanting to Scry with no payoff and wanting instants and sorceries in the graveyard. Ral provides a lot of what's missing, and he acts primarily as a guidepost for players drafting this deck (since Keranos doesn't).

UR should be casting instants and sorceries to power up its chimeras and Spite of Mogis'. Since most instants and sorceries in the cube scry, the deck also seeks to manipulate the top of the deck to find wincons like Thassa's Oracle, Keranos himself, and Prognostic Sphinx. Ral achieves much of this by either letting players Scry for spells or copying spells (which also scry, which generates more card advantage). On top of that, the pinging effect stacks up overtime and presents a clock to the opponent, so they're likely to focus on removing Ral while the UR player digs for Keranos.

As for the flavor, which I really do care about, I figure Ral is a planeswalker who has at least met Ajani and Gideon and has therefore heard about Theros. Having a rivalry with the God of Storms absolutely seems up his alley.