AI Improvements

[furballdn]

Testing with flying SoR eternities. AI does some really stupid stuff.

He has 3 readied eternities on the field. He constantly rewinds my RoLs. That's fine. What is not fine is that AI seems to ignore rustlers. Even with 75 time and 5 readied eternities the rustler simply will not be targetted.

I have two minor phoenix on the field. AI flies an eternity and then casts SoR on it. It then ends the turn.

[Tirear]

Testing with flying SoR eternities. AI does some really stupid stuff.

He has 3 readied eternities on the field. He constantly rewinds my RoLs. That's fine. What is not fine is that AI seems to ignore rustlers. Even with 75 time and 5 readied eternities the rustler simply will not be targetted.

Testing in trainer the AI would always rewind rustlers/leaf dragons before RoLs (as expected, since photosynthesis is considered a powerful skill), even after a SoR. I can't figure out how you got these results unless you aren't following proper scientific procedures.


On an unrelated note, the AI is no more likely to target an adrenaline'd scorpion than a regular one. With vanilla creatures and many other skills the AI picks out the adrenaline'd creature perfectly, but it seems that the targeting boost for having adrenaline doesn't stack with the targeting boost for having venom (or other skills that are considering as powerful, like photosynthesis, but it is especially bad when it comes to venom/deadly venom/nuerotoxin because they synergize with adrenaline).

[furballdn]

Testing with flying SoR eternities. AI does some really stupid stuff.

He has 3 readied eternities on the field. He constantly rewinds my RoLs. That's fine. What is not fine is that AI seems to ignore rustlers. Even with 75 time and 5 readied eternities the rustler simply will not be targetted.

Testing in trainer the AI would always rewind rustlers/leaf dragons before RoLs (as expected, since photosynthesis is considered a powerful skill), even after a SoR. I can't figure out how you got these results unless you aren't following proper scientific procedures.

Really? I've tried at least twice, and I haven't seen the AI rewind my rustlers/leaf dragons at all when testing against my own arena AI. It freely rewinds RoLs (Even RoLs without mitosis over RoLs with mitosis some times).

[Tirear]

Really? I've tried at least twice, and I haven't seen the AI rewind my rustlers/leaf dragons at all when testing against my own arena AI. It freely rewinds RoLs (Even RoLs without mitosis over RoLs with mitosis some times).

It sounds like you are playing hope in these tests, which makes light emitting creatures much higher priority and using it but not mentioning it in your reports is what I had in mind when I said "not following proper scientific procedures".

Hope making RoL too high priority of a target is already reported here and here in this thread, but those cases are so extreme I think it must have been toned down since then, making it fair game for a report on its new (relatively speaking) state.

[furballdn]

Really? I've tried at least twice, and I haven't seen the AI rewind my rustlers/leaf dragons at all when testing against my own arena AI. It freely rewinds RoLs (Even RoLs without mitosis over RoLs with mitosis some times).

It sounds like you are playing hope in these tests, which makes light emitting creatures much higher priority and using it but not mentioning it in your reports is what I had in mind when I said "not following proper scientific procedures".

Hope making RoL too high priority of a target is already reported here and here in this thread, but those cases are so extreme I think it must have been toned down since then, making it fair game for a report on its new (relatively speaking) state.

Ah. Hope. Good point. I remember I had hope up the first game, but am unsure about the second. I think you might have just solved the mystery of that.

AI is still a bit dumb when I had the two phoenixes out though.

[choongmyoung]

Anyway, 1 Rewind per turn is not that good idea.

[jumpingbeans]

Playing Instosis against Paradox. He doesn't get many RoLs early, and I have the combo less a shard, so I play it, Chimaera'd it, and dealt him 182 damage. Paradox then decides to TU my Chimaera - not a great move (I had no shield), but manageable. He then TUs his Chimaera again. That was useless.

[choongmyoung]

Playing Instosis against Paradox. He doesn't get many RoLs early, and I have the combo less a shard, so I play it, Chimaera'd it, and dealt him 182 damage. Paradox then decides to TU my Chimaera - not a great move (I had no shield), but manageable. He then TUs his Chimaera again. That was useless.

Was that just before you dying? AI's use all possible cards if he ensures that he can win this turn.

[jumpingbeans]

No, it wasn't. He made a 25|25 Chimaera or something by TUing mine, and then TUed the Chimaera again. I won the next turn.

[furballdn]

AI fails I saw today:

Plays Pharaoh, plays SoR on Pharaoh, makes one scarab and then ends turn.
I have a ready pharaoh and 2 3|2 scarabs. AI rewinds both scarabs but doesn't touch my ready pharoah.

[dragonsdemesne]

Adrenaline on a 0/1 shard of focus when there's an 11/5 momentumed shrieker and I have an otyugh to devour said shard next turn?  Probably not a good idea...

[jumpingbeans]

AI plays a Pharaoh. It then plays SoR on it. Good, good. Then, it doesn't produce any Scarabs at all, but instead plays a Mitosis on the Pharaoh. It then SoRs the Pharaoh again, and Mitos two more Pharaohs.