YGOPRO Forum

Welcome Guest! To enable all features please try to register or login.
Options
Go to last post Go to first unread
VoidAzure  
#1 Posted : Thursday, April 11, 2019 9:26:28 PM(UTC)
VoidAzure

Rank: Newbie

Joined: 8/25/2018(UTC)
Posts: 2

Has production really stopped on the AI development? If so that is a disappointment.

Edited by moderator Thursday, April 11, 2019 9:38:35 PM(UTC)  | Reason: Not specified

neftalimich  
#2 Posted : Monday, May 13, 2019 4:06:16 AM(UTC)
neftalimich

Rank: Member

Joined: 2/21/2015(UTC)
Posts: 77

Thanks: 25 times
Was thanked: 13 time(s) in 11 post(s)
Originally Posted by: VoidAzure Go to Quoted Post
Has production really stopped on the AI development? If so that is a disappointment.


I know, it's very sad, I'm still here, I miss developing AI :cry
guestno1  
#3 Posted : Tuesday, May 28, 2019 7:19:23 PM(UTC)
guestno1

Rank: Newbie

Joined: 6/19/2016(UTC)
Posts: 28

Thanks: 7 times
Was thanked: 1 time(s) in 1 post(s)
I hope they are still developing it. I can wait no matter how long but i can't wait which has been stopped
kexarakis  
#4 Posted : Tuesday, June 04, 2019 5:15:09 PM(UTC)
kexarakis

Rank: Member

Joined: 11/27/2013(UTC)
Posts: 16

Thanks: 1 times
ygopro from percy has the single most worst dev team ever i dont care if you ban me but that the truth
cybercatman  
#5 Posted : Tuesday, June 04, 2019 8:58:09 PM(UTC)
cybercatman

Rank: Advanced Member

Joined: 5/15/2013(UTC)
Posts: 1,299

Thanks: 5 times
Was thanked: 256 time(s) in 203 post(s)
Originally Posted by: kexarakis Go to Quoted Post
ygopro from percy has the single most worst dev team ever i dont care if you ban me but that the truth


If it is the truth, explain to me how we managed to keep uploading cards with an 2 year old core for the exception of 3 cards (that the other team still not have managed to make too)?

It is nice to says we are "bad", but you need to give fact on why

Did you even tried to put on paper how the AI should act ? because I tried, and I can tell you that it is an hell
Summoning a link take a ton of parameters:
- where is the optimal place to summon (because there is time summoning using the link from EMzone 1 to EMzone 2 is a good play to get more zone to summon from the extra deck)
- what is the optimal materials to use
- shall I use links already on the field
- is summoning a link a good idea or shall i summon another extra deck monster instead

and it is just the basic stuff without even worrying about some of the more "complicated" stuff like "when summoning Apollousa, Bow of the Goddess/Saryuja/Topologic is optimal"?

Add on top of that the older summoning stuff that need a similar reworking ("is it a good idea of summoning that Synchro monster in the EMzone or should i wait until I get an link monster out before" for exemple)

and I'm not even mentioning all the "position matter" stuff that exist (Mekk-Knight or Tiamaton for ex)


But to says we are the "worst dev team ever" (that I remind you, is not paid and only work for passion for the game itself), I guess you are better than all of us and so can provide us stuff like a complete decision tree?


I really cannot says When the new Percy AI would be ready because an "generic" AI is way more complicated than an AI designed to play a specific deck properly (given a lot of "meta" deck follow the same sequences of play)
kexarakis  
#6 Posted : Tuesday, June 04, 2019 10:15:36 PM(UTC)
kexarakis

Rank: Member

Joined: 11/27/2013(UTC)
Posts: 16

Thanks: 1 times
Originally Posted by: cybercatman Go to Quoted Post
Originally Posted by: kexarakis Go to Quoted Post
ygopro from percy has the single most worst dev team ever i dont care if you ban me but that the truth


If it is the truth, explain to me how we managed to keep uploading cards with an 2 year old core for the exception of 3 cards (that the other team still not have managed to make too)?

It is nice to says we are "bad", but you need to give fact on why

Did you even tried to put on paper how the AI should act ? because I tried, and I can tell you that it is an hell
Summoning a link take a ton of parameters:
- where is the optimal place to summon (because there is time summoning using the link from EMzone 1 to EMzone 2 is a good play to get more zone to summon from the extra deck)
- what is the optimal materials to use
- shall I use links already on the field
- is summoning a link a good idea or shall i summon another extra deck monster instead

and it is just the basic stuff without even worrying about some of the more "complicated" stuff like "when summoning Apollousa, Bow of the Goddess/Saryuja/Topologic is optimal"?

Add on top of that the older summoning stuff that need a similar reworking ("is it a good idea of summoning that Synchro monster in the EMzone or should i wait until I get an link monster out before" for exemple)

and I'm not even mentioning all the "position matter" stuff that exist (Mekk-Knight or Tiamaton for ex)


But to says we are the "worst dev team ever" (that I remind you, is not paid and only work for passion for the game itself), I guess you are better than all of us and so can provide us stuff like a complete decision tree?


I really cannot says When the new Percy AI would be ready because an "generic" AI is way more complicated than an AI designed to play a specific deck properly (given a lot of "meta" deck follow the same sequences of play)



then perhaps you should quit cause give empty promises is one thing

Edited by user Tuesday, June 04, 2019 10:16:21 PM(UTC)  | Reason: Not specified

cybercatman  
#7 Posted : Wednesday, June 05, 2019 7:51:16 AM(UTC)
cybercatman

Rank: Advanced Member

Joined: 5/15/2013(UTC)
Posts: 1,299

Thanks: 5 times
Was thanked: 256 time(s) in 203 post(s)
Originally Posted by: kexarakis Go to Quoted Post


then perhaps you should quit cause give empty promises is one thing


What empty promise?
Did we stopped working on the AI? because then you have more info than me xP

Are we taking our time? yes, because the complexity of the stuff is another level compared to MR3 AI
Basically, it would be like making a Chess AI when you had a base of a Checkers AI
And I don't think you would want an AI that do really stupid moves like using a Firewall Darkfluid to summon a Security Gardna?
kexarakis  
#8 Posted : Wednesday, June 05, 2019 8:17:33 AM(UTC)
kexarakis

Rank: Member

Joined: 11/27/2013(UTC)
Posts: 16

Thanks: 1 times
Originally Posted by: cybercatman Go to Quoted Post
Originally Posted by: kexarakis Go to Quoted Post


then perhaps you should quit cause give empty promises is one thing


What empty promise?
Did we stopped working on the AI? because then you have more info than me xP

Are we taking our time? yes, because the complexity of the stuff is another level compared to MR3 AI
Basically, it would be like making a Chess AI when you had a base of a Checkers AI
And I don't think you would want an AI that do really stupid moves like using a Firewall Darkfluid to summon a Security Gardna?


well for me when you say "its in development stay tune" and waiting for like 1000000000 years while other creators have done sounds like an empty promise to me

and before you say "ok then go to the other creators and leave us alone" of corse i will ok
cybercatman  
#9 Posted : Wednesday, June 05, 2019 9:06:24 AM(UTC)
cybercatman

Rank: Advanced Member

Joined: 5/15/2013(UTC)
Posts: 1,299

Thanks: 5 times
Was thanked: 256 time(s) in 203 post(s)
Originally Posted by: kexarakis Go to Quoted Post


well for me when you say "its in development stay tune" and waiting for like 1000000000 years while other creators have done sounds like an empty promise to me

and before you say "ok then go to the other creators and leave us alone" of corse i will ok


Then you show that you don't know what you are speaking about since you are comparing AI tuned to play a specific deck (and so always followed the same sequence of play) to what we had aka an "Generic" AI that could more or less play any deck

The difference and difficulties of implementation are like heaven and earth

Remember the episode when Pegasus gave his deck with a sheet of paper to a kid to play? it is basically the way the other AI work, it check what card it control and what cards it have in the hand and depending on that do a specific play
but if you need to update the deck, that mean reworking the AI of said deck completly

and of course, giving a custom deck to that king of AI is useless since it would not "understand" how to use it



PS: To give you an idea of how complicated an generic AI is difficult to do, You should Try Duel Link (Konami official and professional game) "Auto-play", and see how it work, because for having playing Six sam on it, I can assure you that there is time I wanted to kill it for not doing simple stuff like special summoning Shien while I had 2 sam out
and I'm speaking of an AI that don't have to deal with MR4 or placing in the right zone, along a reduced field/hand/Cardpool (so less cards to check)
kexarakis  
#10 Posted : Friday, June 07, 2019 12:02:20 PM(UTC)
kexarakis

Rank: Member

Joined: 11/27/2013(UTC)
Posts: 16

Thanks: 1 times
Originally Posted by: cybercatman Go to Quoted Post
Originally Posted by: kexarakis Go to Quoted Post


well for me when you say "its in development stay tune" and waiting for like 1000000000 years while other creators have done sounds like an empty promise to me

and before you say "ok then go to the other creators and leave us alone" of corse i will ok


Then you show that you don't know what you are speaking about since you are comparing AI tuned to play a specific deck (and so always followed the same sequence of play) to what we had aka an "Generic" AI that could more or less play any deck

The difference and difficulties of implementation are like heaven and earth

Remember the episode when Pegasus gave his deck with a sheet of paper to a kid to play? it is basically the way the other AI work, it check what card it control and what cards it have in the hand and depending on that do a specific play
but if you need to update the deck, that mean reworking the AI of said deck completly

and of course, giving a custom deck to that king of AI is useless since it would not "understand" how to use it



PS: To give you an idea of how complicated an generic AI is difficult to do, You should Try Duel Link (Konami official and professional game) "Auto-play", and see how it work, because for having playing Six sam on it, I can assure you that there is time I wanted to kill it for not doing simple stuff like special summoning Shien while I had 2 sam out
and I'm speaking of an AI that don't have to deal with MR4 or placing in the right zone, along a reduced field/hand/Cardpool (so less cards to check)


sorry but all i hear is "ai is sooooo hard to do and blah"
just say u arent going to do it/ do it
dont whine about it

cybercatman  
#11 Posted : Friday, June 07, 2019 4:15:59 PM(UTC)
cybercatman

Rank: Advanced Member

Joined: 5/15/2013(UTC)
Posts: 1,299

Thanks: 5 times
Was thanked: 256 time(s) in 203 post(s)
Originally Posted by: kexarakis Go to Quoted Post


sorry but all i hear is "ai is sooooo hard to do and blah"
just say u arent going to do it/ do it
dont whine about it



and all I see is that you are whining and cannot make a proper argument, prefering complaining instead of seeing how complicated it is to make something good even after I pointed the difficulties we were facing

We are working on it, we don't know when it will be ready because we want something that can do basic work and not something that get stuck after the first turn because it would have spend all its ressources into summoning more and more links for nothing

If you can't understand that, then I'm asking you, what are you doing here?
TDOANE is a perfect place for you
Users browsing this topic
Forum Jump  
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.

Notification

Icon
Error