Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error
84 Pages«<2122232425>»
Sepefeets Silverfish update (OG, Kara, MSG)
Options
Previous Topic Next Topic
sepefeets
#452 Posted : Monday, July 4, 2016 8:34:18 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 899

Thanks: 1028 times
Was thanked: 584 time(s) in 238 post(s)
It's in the OP and I think I repeated it a couple times already... A spoiler every once in a while I can deal with but it's been snowballing lately. The more one person does it, the more others think it's the way to do things.

I guess I can spam thanks from now on.
Back to top
WWW
jsmwood729
#454 Posted : Tuesday, July 5, 2016 1:15:19 AM(UTC)
Rank: Member

Groups: Registered
Posts: 12

Thanks: 23 times
Was thanked: 8 time(s) in 6 post(s)
Having an issue with Alexstrasza's Champion not attacking after it gets the charge buff. I created a combo that nudges the bot to coin out Alexstrasza's Champion, but it will not attack after it gets the buff. In the log, Alexstrasza's Champion is coined out on turn 1, but the bot immediately ends the turn without attacking.

http://pastebin.com/phvYiCEY

http://pastebin.com/9TiPDWZV
Back to top
WWW
 |  Edit by user
1 user thanked jsmwood729 for this useful post.
sepefeets on 7/5/2016(UTC)
Dre
#453 Posted : Tuesday, July 5, 2016 2:40:06 AM(UTC)

Rank: Advanced Member

Groups: HSModerator, Registered
Posts: 888
Man
Location: The Netherlands

Thanks: 216 times
Was thanked: 480 time(s) in 315 post(s)
sepefeets;39318 wrote:
It's in the OP and I think I repeated it a couple times already... A spoiler every once in a while I can deal with but it's been snowballing lately. The more one person does it, the more others think it's the way to do things.

I guess I can spam thanks from now on.


Ah must have missed it, my sincere appologies :)
Back to top
WWW
1 user thanked Dre for this useful post.
sepefeets on 7/5/2016(UTC)
sepefeets
#455 Posted : Tuesday, July 5, 2016 5:05:50 AM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 899

Thanks: 1028 times
Was thanked: 584 time(s) in 238 post(s)
jsmwood729;39325 wrote:
Having an issue with Alexstrasza's Champion not attacking after it gets the charge buff. I created a combo that nudges the bot to coin out Alexstrasza's Champion, but it will not attack after it gets the buff. In the log, Alexstrasza's Champion is coined out on turn 1, but the bot immediately ends the turn without attacking.

http://pastebin.com/phvYiCEY

http://pastebin.com/9TiPDWZV

It looks like it shows up on the board but that SF thinks it's not ready to attack. SF probably doesn't understand the battlecry yet.

It will be a couple more days before I'm done working on the mulligan code and can get to misplays. It already found several typos in existing mulligans...
Back to top
WWW
Elekid
#458 Posted : Tuesday, July 5, 2016 5:49:08 AM(UTC)
Elekid


Rank: Advanced Member

Groups: Registered
Posts: 101
Location: Nijmegen, The Netherlands

Thanks: 29 times
Was thanked: 65 time(s) in 58 post(s)
Not sure if misplay or suboptimal play:

plays swipe to go face and take the only enemy minion (1 health) down in the process.

Instead could have killed minion with hero power, play Twilight Elder and save swipe for a upcoming turn where there would probably be more minions to deal damage to.

not the last swipe which won me the game (go sepefeets!), but the first one.
Back to top
WWW
2 users thanked Elekid for this useful post.
parakeet on 7/5/2016(UTC), sepefeets on 7/5/2016(UTC)
parakeet
#442 Posted : Tuesday, July 5, 2016 5:54:10 AM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
Dre;39307 wrote:
Hmm iam not joking if i say that in 50 games i found atleast 5 games (10%) maybe even more, where it failed to kill the doomsayer. I believe when u say it's all sync, but it happens on doomsayer allot then. Another one here:


Do you have any logs of it wasting minion attacks into a surviving doomsayer - or is it just durability?
Back to top
WWW
1 user thanked parakeet for this useful post.
sepefeets on 7/5/2016(UTC)
parakeet
#459 Posted : Tuesday, July 5, 2016 6:03:34 AM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
Elekid;39332 wrote:
Not sure if misplay or suboptimal play:

plays swipe to go face and take the only enemy minion (1 health) down in the process.

Instead could have killed minion with hero power, play Twilight Elder and save swipe for a upcoming turn where there would probably be more minions to deal damage to.

not the last swipe which won me the game (go sepefeets!), but the first one.


Datasync issue. It did not update the Druid's attack after using heropower. Its initial moveset did not involve casting the swipe.

Not updated: druid 12 30 1 False False 76 True 0 False 0 0
Updated: druid 12 30 1 False False 76 True 0 False 1 0
Back to top
WWW
1 user thanked parakeet for this useful post.
sepefeets on 7/5/2016(UTC)
parakeet
#456 Posted : Tuesday, July 5, 2016 6:09:30 AM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
sepefeets;39331 wrote:
It looks like it shows up on the board but that SF thinks it's not ready to attack. SF probably doesn't understand the battlecry yet.

It will be a couple more days before I'm done working on the mulligan code and can get to misplays. It already found several typos in existing mulligans...


It correctly calculated attacking in both logs. Are you sure this isn't datasync?
Back to top
WWW
1 user thanked parakeet for this useful post.
sepefeets on 7/5/2016(UTC)
Elekid
#460 Posted : Tuesday, July 5, 2016 10:05:26 AM(UTC)
Elekid


Rank: Advanced Member

Groups: Registered
Posts: 101
Location: Nijmegen, The Netherlands

Thanks: 29 times
Was thanked: 65 time(s) in 58 post(s)
parakeet;39336 wrote:
Datasync issue. It did not update the Druid's attack after using heropower. Its initial moveset did not involve casting the swipe.

Not updated: druid 12 30 1 False False 76 True 0 False 0 0
Updated: druid 12 30 1 False False 76 True 0 False 1 0


I guess it's the same thing with this play?:

Plays Mana Wyrm first, play spell after (in this case Fireball).
Back to top
WWW
2 users thanked Elekid for this useful post.
parakeet on 7/5/2016(UTC), sepefeets on 7/5/2016(UTC)
Dre
#443 Posted : Tuesday, July 5, 2016 10:49:50 AM(UTC)

Rank: Advanced Member

Groups: HSModerator, Registered
Posts: 888
Man
Location: The Netherlands

Thanks: 216 times
Was thanked: 480 time(s) in 315 post(s)
parakeet;39333 wrote:
Do you have any logs of it wasting minion attacks into a surviving doomsayer - or is it just durability?


No i can't recall it wastes minion attacks

Good news - I hardly found ANY misplays with my hunter yesterday, 1-2 which iam sure i can fix myself with the combo file.
Bad news - the meta changed to be allot more agressive this season. and my hunter is midrange and not hybrid so bit to slow for the rly fast decks. Got a 48% wr yday. Gonna change a card or 2 to tackle the agro. Might be just unlucky yday.

Need some help:

Can somone explain to me why this:
Wont simulate the play of call of the wild.

Its an emty board, enemy has 8 health left. with the use of call of the wild i could hit directly for 5 dmg (huffer) and have 3 strong minions on the board. The sim will allways give to play the 2 2-drops and hero power. Pretty weak. I lost this match in the end aswell...

I tried some things myself but even aan -100 value on call didnt help. I tried it when with bonuses the play it on turn 8. Eventualy it simed to play it on a 50 !! bonus so i had to use this:"

Code:
// [Nudge] Force [8] Call of the Wild (OG_211)
OG_211,0;mana:8;bonus:50;hero:hunter


To make it play call, in that play in the spoiler.


Another question

Why does the bot allways coin 2 drop on turn 1 like in this hand:


Its allot beter to stay on curve, 1 drop on 1, 2-drop on 2, 3-drop on 3 (all in hand) and use the coin for better cards later. I tried serval things in combo. Raising the penalty to play coin, lowering the value of the 1-drop. But the boards keeps simming to coin the toad.
Back to top
WWW
 |  Edit by user
2 users thanked Dre for this useful post.
parakeet on 7/5/2016(UTC), sepefeets on 7/5/2016(UTC)
Elekid
#463 Posted : Tuesday, July 5, 2016 12:20:58 PM(UTC)
Elekid


Rank: Advanced Member

Groups: Registered
Posts: 101
Location: Nijmegen, The Netherlands

Thanks: 29 times
Was thanked: 65 time(s) in 58 post(s)
  1. attackWithMinion: Fencing Coach target: Bloodhoof Brave
  2. playcard: Polymorph target: Bloodhoof Brave choice: 0 place1


wrong order
Back to top
WWW
2 users thanked Elekid for this useful post.
parakeet on 7/5/2016(UTC), sepefeets on 7/5/2016(UTC)
parakeet
#447 Posted : Tuesday, July 5, 2016 3:23:39 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
Dre;39343 wrote:


Another question

Why does the bot allways coin 2 drop on turn 1 like in this hand:


Its allot beter to stay on curve, 1 drop on 1, 2-drop on 2, 3-drop on 3 (all in hand) and use the coin for better cards later. I tried serval things in combo. Raising the penalty to play coin, lowering the value of the 1-drop. But the boards keeps simming to coin the toad.


There is a large bonus built in for playing the first turn very aggressively. Specifically, anything with attack>3 and HP>2 gets a 20 point bonus on turn 1. I suggested Sepefeets lower the bonus on 3/2 minions to 5 - We will see what he thinks. It might be better to remove it altogether. I am playing with the reduced version right now, and it seems to play intelligently first turn.
Back to top
WWW
2 users thanked parakeet for this useful post.
Dre on 7/5/2016(UTC), sepefeets on 7/5/2016(UTC)
parakeet
#444 Posted : Tuesday, July 5, 2016 3:39:33 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
Dre;39343 wrote:
Need some help:

Can somone explain to me why this:
Wont simulate the play of call of the wild.

Its an emty board, enemy has 8 health left. with the use of call of the wild i could hit directly for 5 dmg (huffer) and have 3 strong minions on the board. The sim will allways give to play the 2 2-drops and hero power. Pretty weak. I lost this match in the end aswell...

I tried some things myself but even aan -100 value on call didnt help. I tried it when with bonuses the play it on turn 8. Eventualy it simed to play it on a 50 !! bonus so i had to use this:"

Code:
// [Nudge] Force [8] Call of the Wild (OG_211)
OG_211,0;mana:8;bonus:50;hero:hunter


To make it play call, in that play in the spoiler.


It calculated there was greater than 65% chance of the mage having flamestrike. You can up your playaround prob and it will be more aggressive. 70% was enough for it to play Call of the Wild. Honestly, the playaround code appears rather weak and not updated for the current Meta. Maybe Sepefeets can make it awesome.
Back to top
WWW
2 users thanked parakeet for this useful post.
Dre on 7/5/2016(UTC), sepefeets on 7/5/2016(UTC)
parakeet
#462 Posted : Tuesday, July 5, 2016 3:43:09 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
Elekid;39342 wrote:
Plays Mana Wyrm first, play spell after (in this case Fireball).


Sepefeets or I will look into this
Back to top
WWW
2 users thanked parakeet for this useful post.
sepefeets on 7/5/2016(UTC), Elekid on 7/6/2016(UTC)
Dre
#445 Posted : Tuesday, July 5, 2016 3:58:22 PM(UTC)

Rank: Advanced Member

Groups: HSModerator, Registered
Posts: 888
Man
Location: The Netherlands

Thanks: 216 times
Was thanked: 480 time(s) in 315 post(s)
parakeet;39350 wrote:
It calculated there was greater than 65% chance of the mage having flamestrike. You can up your playaround prob and it will be more aggressive. 70% was enough for it to play Call of the Wild. Honestly, the playaround code appears rather weak and not updated for the current Meta. Maybe Sepefeets can make it awesome.


How to set it to 70% in settings i can turn this either True or False i tought...
Back to top
WWW
1 user thanked Dre for this useful post.
sepefeets on 7/5/2016(UTC)
parakeet
#446 Posted : Tuesday, July 5, 2016 4:01:22 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
Dre;39353 wrote:
How to set it to 70% in settings i can turn this either True or False i tought...


Use
playaroundprob = 70;

The description is a little counter-intuitive.
Back to top
WWW
2 users thanked parakeet for this useful post.
Dre on 7/5/2016(UTC), sepefeets on 7/5/2016(UTC)
etronpoilu
#465 Posted : Tuesday, July 5, 2016 4:33:58 PM(UTC)
etronpoilu

Rank: Member

Groups: Registered
Posts: 26
Location: Can

Thanks: 17 times
Was thanked: 13 time(s) in 11 post(s)
New update for hearthranger ( v.5.9.1.0) says : "- Bug refix for silverfish data desync."
Should that fix all data sync problem ? Should it work only with the included silverfish version of HR or with sepefeets version too?
Back to top
WWW
1 user thanked etronpoilu for this useful post.
sepefeets on 7/5/2016(UTC)
parakeet
#466 Posted : Tuesday, July 5, 2016 5:38:29 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
etronpoilu;39355 wrote:
New update for hearthranger ( v.5.9.1.0) says : "- Bug refix for silverfish data desync."
Should that fix all data sync problem ? Should it work only with the included silverfish version of HR or with sepefeets version too?


Oh my, how exciting!! I am curious to see what's changed! I would expect it to work with any silverfish or custom AI, but I cannot say for certain.

EDIT: It didn't even overwrite silverfish when I updated - that would imply any version should work. Can anyone confirm a drop in datasync errors? Is anyone seeing increased roping?
Back to top
WWW
 |  Edit by user
1 user thanked parakeet for this useful post.
sepefeets on 7/5/2016(UTC)
sadpwner
#467 Posted : Tuesday, July 5, 2016 8:40:13 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 262
Location: somewhere

Thanks: 3 times
Was thanked: 64 time(s) in 54 post(s)
Data sync seems to be still present.

Some proof

http://pastebin.com/WrSAd7pM

Turn 8, still can't wait for the slime after the sludge belcher. It also tried attacking with minions that already attacked many times before.
__
http://pastebin.com/wwkhLhAj

Turn 4, pointlessly uses two power overwhelmings when one would suffice.
Back to top
WWW
 |  Edit by user
1 user thanked sadpwner for this useful post.
sepefeets on 7/5/2016(UTC)
sepefeets
#457 Posted : Tuesday, July 5, 2016 10:58:34 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 899

Thanks: 1028 times
Was thanked: 584 time(s) in 238 post(s)
parakeet;39337 wrote:
It correctly calculated attacking in both logs. Are you sure this isn't datasync?

The sim calculated correctly but that's only one part of it. Once it's on the board it could be data sync or it could be that SF isn't looking in the right place to see that it has charge since it gains the charge as a battlecry rather than always having it.

Elekid;39342 wrote:
I guess it's the same thing with this play?:

Plays Mana Wyrm first, play spell after (in this case Fireball).

Hmmm, it looks like the fireball was drawn normally on turn 7 but that for some reason it didn't map the name for it so it has "placeholdercard" for the name which throws off everything...

The other log I suspect it probably attacked to trigger the enrage before polymorph was considered worth using.

@Dre - I think with the current meta aoe playaround should probably be disabled or severely reduced in settings. Coining 2 drops bonus will get reduced/removed.
Back to top
WWW
2 users thanked sepefeets for this useful post.
Dre on 7/6/2016(UTC), Elekid on 7/6/2016(UTC)
84 Pages«<2122232425>»
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.

Powered by YAF | YAF © 2003-2011, Yet Another Forum.NET
This page was generated in 0.290 seconds.