Help with improving deck

[Eddygp]

So I have used Xenocidius' Simulator to fight a given Earth-rich rainbow deck against another modified deck that is almost identical (adding one card or removing one card, for instance), for 5000 games. If the change is positive, I would include it and otherwise ignore it. I thus adjusted pillars, number of shards etc but the problem is that I have reached a 42-card dead end that is not too good and I can't quite get out of it.

By the way, Xenocidius' simulator has a bug with SoSac, because it always stays "on" if you know what I mean. Once it reaches count 0, you still get healed.

Anyway, here i the deck, hope you find it interesting. Oh and yes, for some reason two Shards of Serendipity or none were BOTH worse decks than one shard. Weird.

Hover over cards for details, click for permalink

Deck import code : [Select]

6qq 6qq 6qq 6qq 6qq 6qq 6qq 6qq 6qq 6qq 6uq 71a 71a 778 778 778 778 778 778 778 77c 77c 77g 77g 77g 77g 77g 77g 77k 77k 77k 77m 77m 77m 77m 7ba 7ba 7ba 7dr 7dr 7km 7km 8pm

[Aves]

The simulator runs an older patch, I believe 1.28 or 1.29.

That's an interesting process to change the deck. When you ran the numbers, how large was the deviation from 50% in each run? Running that deck against itself for 3 runs of 5000, I got 49.44%, 49.6%, and 49.52%

[InsignificantWeeaboo]

Just going to let you know that we have a separate forum section for help on decks. (http://elementscommunity.org/forum/deck-help/)

Seeing as this looks like a stall deck, I don't really see the reason for Arsenics and SoSe. I think you can go without them. Have you considered Supernovas?

Sometimes, it's better to test decks on AIs. The trainer is good for this, since you have access to every card in the game: www.elementsthegame.com/trainer

[Eddygp]

Thank you for your replies and sorry for not having posted it in the appropriate forum. Hopefully a moderator will be able to move the thread.

I kept trying the decks o the trainer too. The deviations that I considered were those that were above around 51% after 15000 games, above 52% for 10000 games or above 53% for 5000 games. I tried removing the Arsenic and the SoSe, but it performs worse against its unaltered version - although this was a few modifications earlier, I will check today in case it is now better to drop them.

EDIT: The original deck wins 87% of the time vs the modified deck dropping both Arsenics, running 5000 games. The original deck wins 54.6% of the time vs the modified deck dropping the SoSe, running 10000 games.

EDIT2: Current improved deck:

Hover over cards for details, click for permalink

Deck import code : [Select]

6qq 6qq 6qq 6qq 6qq 6qq 6qq 6qq 6qq 6qq 6uq 6uq 6uq 71a 71a 778 778 778 778 778 778 778 77c 77c 77g 77g 77g 77g 77g 77g 77k 77k 77k 77k 77m 77m 77m 77m 7ba 7ba 7ba 7ba 7ba 8pm

After one-card improvements, clearly better. Beats the deck in first post 60.4% of the time in 5000 games. It becomes even more stall-ish, beating the other already stalling deck with an average of 22 turns to win.

[dragtom]

You are supposed to be a better player than an AI.
testing with or without arsenic and/or SoSe is a lot better than using the simulator.

Personally i've found SoG to be very underwhelming when used without mark of .