AI Deck Prioritization

Mburn7

I know this has been said before, but there really should be some way to tell the AI how to pilot your deck.

Currently it seems to run creature first, then spell, then support, with the highest cost/rarity of each category coming first.
But for a lot of decks this isn't the optimal way to play.

Why not put in a quick checklist when you enter a deck into a node?  Something like

-Default
-Creatures First
-Creatures Last
-Supports First
-Supports Last
-Spells First
-Spells Last

Or something similar (maybe a 1-2-3 ranking you assign to each category for priority order).

Within each category the AI's existing priority rules could still apply.  This way if you make a deck that does not want to summon creatures (my embalm/eternalize decks are a good example), you can let the AI know to cast the supports and spells first instead of saving up for big creatures that you want discarded instead.

This would fit in with Oktagon's (so far successful) attempts to make the AI more skilled and difficult, and make matches much more interesting and challenging (instead of always knowing that expensive card being charged is a creature)

Thoughts?

UweTellkampf

Yes, please! Anything to give the AI a direction would be great. I like the way how you set it up through categories, but it would also be nice to give the AI the priority simply via the order of cards in the deck. The first card you put in would be the one the AI should prioritize. Then the second and so on. 

julianus

Interesting idea, but remember this is a commercial product they work on. Do you think a change like this would result in enough additional revenue to justify the time and expense of coding it?

Mburn7

UweTellkampf said:

Yes, please! Anything to give the AI a direction would be great. I like the way how you set it up through categories, but it would also be nice to give the AI the priority simply via the order of cards in the deck. The first card you put in would be the one the AI should prioritize. Then the second and so on. 

That would be a little more annoying to set up, and possibly difficult to code.  I set up mine the way I did because it would fit nicely with the existing rules, just with a different type order

bk1234

julianus said:

Interesting idea, but remember this is a commercial product they work on. Do you think a change like this would result in enough additional revenue to justify the time and expense of coding it?

What if adding card prioritization to a deckslot was a feature that you have to use crystals for?  I know quite a few people who would do it to troll around a bit. 

asm0deus

I think it might not work as well as we think. Imagine a hand with 2 rishkars expertise and reason. Putting rishkars as more of a priority than reason would mean the ai would cast both rishkars first before reason when i would set it up as rishkars/reason/rishkars. Rishkars might not be the best example since it depends on hand size but i hope you get the idea.

Mburn7

asm0deus said:

I think it might not work as well as we think. Imagine a hand with 2 rishkars expertise and reason. Putting rishkars as more of a priority than reason would mean the ai would cast both rishkars first before reason when i would set it up as rishkars/reason/rishkars. Rishkars might not be the best example since it depends on hand size but i hope you get the idea.

Obviously it won't be perfect, nor would we expect it to be.  But, for example, if the hand has a rishkar, a reason, and a Gaia's Revenge, you would want to play the rishkar first, not Gaia like it would now

julianus

bken1234 said:

julianus said:

Interesting idea, but remember this is a commercial product they work on. Do you think a change like this would result in enough additional revenue to justify the time and expense of coding it?

What if adding card prioritization to a deckslot was a feature that you have to use crystals for?  I know quite a few people who would do it to troll around a bit. 

Might work - hard for me to say, without knowing what parts of the game people spend real money on. I have a hunch that most of the people who'd be keen to troll are the same people who never need to spend money on crystals.

Another possibility could be including this as part of a hypothetical set of game-wide system and feature changes, aimed at making the game more generally appealing to play and spend money on. I guess the open question is whether this kind of change would contribute to that.

FindingHeart8

Mburn7 said:

I know this has been said before, but there really should be some way to tell the AI how to pilot your deck.

Currently it seems to run creature first, then spell, then support, with the highest cost/rarity of each category coming first.
But for a lot of decks this isn't the optimal way to play.

Why not put in a quick checklist when you enter a deck into a node?  Something like

-Default
-Creatures First
-Creatures Last
-Supports First
-Supports Last
-Spells First
-Spells Last

Or something similar (maybe a 1-2-3 ranking you assign to each category for priority order).

Within each category the AI's existing priority rules could still apply.  This way if you make a deck that does not want to summon creatures (my embalm/eternalize decks are a good example), you can let the AI know to cast the supports and spells first instead of saving up for big creatures that you want discarded instead.

This would fit in with Oktagon's (so far successful) attempts to make the AI more skilled and difficult, and make matches much more interesting and challenging (instead of always knowing that expensive card being charged is a creature)

Thoughts?

I'm not sure if I'm on board with this.  I'd hate to give the AI the ability to activate Omniloop turn 1, or Rivers Rebuke every other turn.  This model sounds good on paper, but I could see this very easily abused and definitely not an enjoyable experience.

Kinesia

Simple changes might be lands/mana cards first, mana spells not to be played without other cards in hand.

Then even randomising the rest of the hand would be more interesting than now.

morgue427

and you think they complain about the ai now? lol

Mburn7

morgue427 said:

and you think they complain about the ai now? lol

I know, I find it hilarious.  This AI still isn't as good as it was when I first started playing, and yet everyone thinks its op as heck.

I would love to see it get even harder.  I hate perfect scores (especially since I never seem to get them)

FindingHeart8

Mburn7 said:

morgue427 said:

and you think they complain about the ai now? lol

I know, I find it hilarious.  This AI still isn't as good as it was when I first started playing, and yet everyone thinks its op as heck.

I would love to see it get even harder.  I hate perfect scores (especially since I never seem to get them)

I don't think that would be fair to lower level players.  High entry barriers do not make for a profitable game.

Mburn7

FindingHeart8 said:

Mburn7 said:

morgue427 said:

and you think they complain about the ai now? lol

I know, I find it hilarious.  This AI still isn't as good as it was when I first started playing, and yet everyone thinks its op as heck.

I would love to see it get even harder.  I hate perfect scores (especially since I never seem to get them)

I don't think that would be fair to lower level players.  High entry barriers do not make for a profitable game.

The thing is, thanks to the (admittedly flawed) tier system, low level players face mostly other low level players.  It doesn't matter how smart the AI is if the decks aren't very good.

Also, my proposed change won't really effect most low level decks anyway.  But if its really an issue, make it so you don't unlock the controls until Gold tier (or even platinum).  Then its an extra incentive to level up but not a barrier for noobs

DBJones

An important note: most new players start in story mode, a few of the early battles would be quite awful with bad cards and really good AI. However, I think the biggest issue with story mode is how it gives new players the false impression that leveling up your Planeswalkers is always better.

FindingHeart8

Mburn7 said:

FindingHeart8 said:

Mburn7 said:

morgue427 said:

and you think they complain about the ai now? lol

I know, I find it hilarious.  This AI still isn't as good as it was when I first started playing, and yet everyone thinks its op as heck.

I would love to see it get even harder.  I hate perfect scores (especially since I never seem to get them)

I don't think that would be fair to lower level players.  High entry barriers do not make for a profitable game.

The thing is, thanks to the (admittedly flawed) tier system, low level players face mostly other low level players.  It doesn't matter how smart the AI is if the decks aren't very good.

Also, my proposed change won't really effect most low level decks anyway.  But if its really an issue, make it so you don't unlock the controls until Gold tier (or even platinum).  Then its an extra incentive to level up but not a barrier for noobs

I get what you're saying.  I just want to see what sets rotate out of standard, and what cards get "rebalanced" before I'm comfortable with any harder difficulty than we already have.

If you feel like the game is too easy for you, you could always play a deck full of commons/uncommons or play a deck with planeswalker Hautli

Mburn7

DBJones said:

An important note: most new players start in story mode, a few of the early battles would be quite awful with bad cards and really good AI. However, I think the biggest issue with story mode is how it gives new players the false impression that leveling up your Planeswalkers is always better.

I wouldn't call that a false impression.  Especially with the Origins and Zendikar walkers which were specifically designed to encourage people to level them up (there used to be an issue where there was no point leveling Gideon past 20).  Some of the new walkers don't need to be level 60, but that's more because of terribly designed mana curves than anything else.

Why would you not want to level your walkers?  To force easier opponents in events?  I can't think of any other benefit

DBJones

Exactly, leaving your Planeswalkers low-level means you face other people at low levels. Most importantly, more of the people you face have reasonable card collections, people with level 60 characters generally have better cards. In Silver, my level 1 Liliana, Death's Majesty with one rare, the rest uncommon and common is my most consistent non-cycling strategy. Nicol Bolas and the improved AI have dropped my win rate a bit, but I still usually go 4-0 in Training Grounds. Having a few level 60 Planeswalkers for PvE is important, but low-level is often the way to go for PvP.

morgue427

i have leveled all of mine( except the last few i am still working on that i got recently) sorry i think it is cheating to try to  play lower levels hoping to get easy matches. I am  firm believer in slow and steady play to get there, that said in tg good night i go 4-0 but far from every night, usually i go 4-1, i live with it it. yes it is with a level 60 in platinum, i got there and feel like it is cheating to use a low level just to get a better chance to win faster.
  I dont think that the suggested improvments are a bad idea but i cant picture some wanting to play against them if the are trying to min/max wins and rewards. i would try it but some of those decks i wouldnt want to face ai smart enough to play them perfectly by anymeans, nobody wants to lose everytime though

Mainloop25

To me, using low level walkers is akin to using all-in cycle decks. Sure you can win easier, but are having as much fun doing it? 

FindingHeart8

Mainloop25 said:

To me, using low level walkers is akin to using all-in cycle decks. Sure you can win easier, but are having as much fun doing it? 

I agree with this 95%, except with Gideon1 (who I read somewhere on these threads is more powerful at level 10 than at level 60).  I never leveled him up past that and I do enjoy using him for weenie battles.