Land Stacking vs Land Normalization

[Chris H.]

Land Stacking vs Land Normalization

I had an idea the other day.

Land "stacking" gives the computer a long string of card draws that are non-land cards in the mid-game portion. It can be hard for us to keep up with this since we are getting lands mixed in with our own card draws. Land "stacking" can give the computer a real advantage when using an AI tuned quest deck.

When we turn off the land "stacking" flag, the computer becomes a potential mana hosed/flooded victim. This has a tendency to place the weak computer AI at a disadvantage.

I wonder if a land normalization routine might be workable. The non-land cards would be shuffled and then the lands would be placed in the deck at fairly even intervals. The lands would not be clumped together in some areas and altogether missing in other areas of the computer's library.

We could also use Dennis' land "threading" algorithm. This would help with the designing of AI multi color decks. The land "threading" algorithm does not clump all of the land at the start and end of the library. It is designed to alternate the land types and colors in an ordered distribution.

[DennisBergkamp]

I think this is a good idea... I'll give this a shot sometime. Another thing that could be useful is that when some fetch spell is played by the AI, we could "normalize" and alternate lands instead of doing a regular shuffle.

[Mr.Chaos]

I gues it could work...
But you can get manahosed/flooded too so it is all part of the game really.
(nothing worse than having 2 lands in your opening hand, then waiting 8 turns for your 3rd land!)
I mean, to keep things fair, just shut off land stacking for questmode because the decks are hard enough as is. But in the standard playmode, you could implement the Land Stacking or Land Normalization where it sort of functions as a way to level up the AI.
No land fixing = lvl 1
Land Normalization = lvl 2
Land Stacking = lvl 3

Just a thought that hits me as I type this so I am just dropping it here:
Make the "Generate Deck" function actually use multiple copies of cards instead of just 1 of each as it does now.
That could make randomly generated decks stronger.

[Chris H.]

But you can get manahosed/flooded too so it is all part of the game really.

I mean, to keep things fair, just shut off land stacking for questmode because the decks are hard enough as is.

Prior to the release of quest mode I would not Mulligan. I would concede the game and then restart the match. In quest mode, I Mulligan if my opening hand looks bad. I do not want to record a loss.

I have now played two quick 10 match quest games with the new and easier decks in the recent questData file. Have you had a chance to play test the new questData file? I am interested in other people's feedback.

The old hard decks are now only faced at the end of the quest. The weaker decks are now played in the early and mid quest stages.

There are a few files to download and it is requires a wee bit of work to put these together and to beta test these new decks.