Skyfall predetermined...

Aeroplane

Playing Koth in TG. He has one red match , but goes for a match with plains which skyfalls  a 5 match off screen which leads to another 5 match. I'm playing Bolas so matching plains wouldn't do anything to me. There was no reason whatsoever unless the skyfall was predetermined.
 Too note the M19 packs were garbage for mythic rate. Was going naip before , but after this my time ,I'll will  split my time with Puzzle and Dragons which runs way smoother than this one and more generous with gifts. Great game but way too much fishy stuff , especially, after the pink gem thread i just read,poor rates, freezes and lack of communication with the player base with what's going on.

Volrak

Sorry about your bad luck.

Regarding strange matches, the AI has been crippled so that when the AI's best match according to its normal algorithm gives "too much" mana, the AI is forced to take a match which it evaluates as weak instead.  I am guessing that's what you saw.  And, of course, any match anywhere has a chance of causing cascades from off-screen.  (Ironically, if Greg's forced-weak matches *never* cascaded, then that would actually be evidence for predetermined skyfalls.)

Aeroplane

Actually, I enjoy the new challenge with the balancing of the new set and out with Omni/huf , but it looked strange seeing such odd behavior. i should have put this under "let off some steam thread" as my draw of cards were horrendous yesterday then I had some weird matches. Oh well, thanks for the pat on the back.

[Deleted User]

Noooooo! don't give in! its all rigged, rigged I say!

Talahamut

I've noticed this too... If Greg makes an off-color 3-match when there is an on-color match available on the board, there's a cascade coming. 

I've never heard of the "too much mana" condition, but I'm not sure how that could apply to a simple on-color 3-match when the AI will almost always take any available 4- or 5-match.

Talahamut

OK, I take it back. I just saw Greg take an off-color match-3 when there was an on-color match-5 available, and there was no cascade afterwards.  Sometimes...he's just horribly dumb.

Volrak

Talahamut said:

I've never heard of the "too much mana" condition, but I'm not sure how that could apply to a simple on-color 3-match when the AI will almost always take any available 4- or 5-match.

Usually when I've noticed a dumb match, the best match which Greg avoided was more than just a simple on-colour match, often 2 matches in the same move.  But I don't know the exact rules.

HarryMason

Greg is definitely dumber than previous incarnations. In many ways . Seems to be less spammy with his first ability when using non bolas pws . Still has magical Greg luck with bolas. Cascades were painful ,but I don't like this Greg . Feels like beating up a blind guy. 

hawkyh1

Talahamut said:

OK, I take it back. I just saw Greg take an off-color match-3 when there was an on-color match-5 available, and there was no cascade afterwards.  Sometimes...he's just horribly dumb.

what lead to greg having a match 5 available?
rng or you setting it up?

HH

jtwood

But I still see Greg passing on match-5s. How do you reconcile that with the idea the AI is cheating?

DBJones

I think that the most recent Greg nerf was more adding randomness to his choices than adding conditions to his choices. That means occasionally his choices will end up working better than the optimal choice he would have made before, but usually they end up worse. It also makes him harder to manipulate, which was simple before, though tricky to actually set up most of the time.

DumasAG

Was there a change to Greg recently, though? In TDW, I've noticed several opposing PWs using their second and third abilities, which Greg never did after he was neutered. 

Mburn7

DumasAG said:

Was there a change to Greg recently, though? In TDW, I've noticed several opposing PWs using their second and third abilities, which Greg never did after he was neutered. 

Yeah, he definitely got a slight buff now.  He's very inconsistent, too.  I've seen Bolas leave my creatures out for 2-3 turns before finally using his first on them.  Can't find a pattern anymore.

Stormbringer0

The point is that we know that Greg is as dumb as a stone and that’s why it feels unfair, when he gets cascades after making a bad move.

DumasAG

Stormbringer0 said:

The point is that we know that Greg is as dumb as a stone and that’s why it feels unfair, when he gets cascades after making a bad move.

Apparently people also thought it was unfair when he got cascades as smart Greg. Poor Greg...

ZW2007-

DumasAG said:

Was there a change to Greg recently, though? In TDW, I've noticed several opposing PWs using their second and third abilities, which Greg never did after he was neutered. 

I am of the belief that if Greg cannot use and benefit from both his normal first ability and the node specific first ability, then he will not use the ability. I was facing a Vraska on the node that destroys an Enchantment, an Artifact, and a creature respectively. I had a single enchantment support in play and she never used her first. I cast a second one and then she used her first to destroy both. I have seen similar things happen like Samut not cast her first because I didn't have a creature in play for the "Your first creature deals damage equal to its power to your opponent's first creature" ability.

Talahamut

hawkyh1 said:

Talahamut said:

OK, I take it back. I just saw Greg take an off-color match-3 when there was an on-color match-5 available, and there was no cascade afterwards.  Sometimes...he's just horribly dumb.

what lead to greg having a match 5 available?
rng or you setting it up?

HH

I'm not 100%, but I'm pretty sure I didn't notice it and left it there myself. (Surprisingly my name is not Greg :P)

hawkyh1

if greg needs code to reduce the amount of match 5's
taken, then what does that say about the amount of
possible match 5's greg gets?

HH

Volrak

hawkyh1 said:

if greg needs code to reduce the amount of match 5's
taken, then what does that say about the amount of
possible match 5's greg gets?

It says that that amount is more than the player base can handle.  And given that Greg gets the same number of match-5 opportunities as us (or less, through careful play), that's a little disappointing for those who like the challenge aspect.

hawkyh1

Volrak said:

hawkyh1 said:

if greg needs code to reduce the amount of match 5's
taken, then what does that say about the amount of
possible match 5's greg gets?

It says that that amount is more than the player base can handle.  And given that Greg gets the same number of match-5 opportunities as us (or less, through careful play), that's a little disappointing for those who like the challenge aspect.

I like your data collecting but I don't agree with this.
imo greg definitely gets better rng(gem matching,
gem conversion, gem placement etc). I can't prove
it or explain it. it's a bit like when you get dealt 3
cards jack, queen and king. and greg gets dealt the
same 3 cards. but in secret your matches are always
jack versus queen, queen versus king and king
versus jack. results in favour of greg despite the
rng showing no technical preference. it's not about
the challenge. I believe greg should take each and
every match 5, because match 5's should be a
hugely rare occurrence.

HH

Volrak

Everyone will form their own opinions, and we don't have to agree.  But one enlightening finding from the cascade data collection was that play style seemed to have a heavy influence on how many matches per move you could expect to get compared to Greg.  So it may well be that for some players with certain styles, Greg's programmed style really does consistently come out on top.  I think this phenomenon and perception bias are two explanations for what is sometimes observed, both of which are more likely than a bias towards the AI which is so far unmeasurable through observation, undetectable through code analysis, and would represent added game complexity (a cost for the developers) without a clear benefit.