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

Notification

Icon
Error

84 Pages«<1516171819>»
Sepefeets Silverfish update (OG, Kara, MSG)
anczej
#328 Posted : Monday, June 27, 2016 5:24:43 AM(UTC)
Rank: Member

Groups: Registered
Posts: 27

Thanks: 2 times
Was thanked: 5 time(s) in 4 post(s)
Lets make some tests during this night. Lets set up same deck with same combos muligans and try for each other version of sepe silverfish. Probably we will have some info about problems caused by different updates.

Who could help?
Dre
#329 Posted : Monday, June 27, 2016 5:47:24 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)
If you want to test, here are my latest files: https://dl.dropboxuserco...Aggro%20Shaman%20v5.zip (deck is in the mulligan)
parakeet
#330 Posted : Monday, June 27, 2016 5:57:44 AM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
I will try your aggro shaman with 5.8.2 and 121.2SE. My midrange is holding at 55% winrate after 63 games. From Rank 11 to Rank 8
anczej
#331 Posted : Monday, June 27, 2016 6:14:52 AM(UTC)
Rank: Member

Groups: Registered
Posts: 27

Thanks: 2 times
Was thanked: 5 time(s) in 4 post(s)
My midrange is holding at 48% winrate after 70 games, scottsbreak settings. With 5.9 HR and last Sepe SF.
sepefeets
#332 Posted : Monday, June 27, 2016 6:35:51 AM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 899

Thanks: 1028 times
Was thanked: 584 time(s) in 238 post(s)
More than anything I need some automated testing to run through logs with different SF versions to catch regressions but that will take time that I don't have.
darkayo
#333 Posted : Monday, June 27, 2016 6:53:48 AM(UTC)
Rank: Member

Groups: Registered
Posts: 20
Location: FRANCE

Thanks: 1 times
Was thanked: 4 time(s) in 4 post(s)
Testing 121.0SE with 5.9 middlerange Chaman.
Dre
#334 Posted : Monday, June 27, 2016 8:31:24 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)
@sepe

Could you tweak "face" just a bit? This log is a perfect example: http://pastebin.com/RZAqzRCg

You would expect this behaviour for "rush" but on "face" it shoudnt bother killing off cards that are no treath at all like an Elek and a toad. Doing this slows the game down allot. and you want to burst face when the setting is on "face" Ofc it should kill taunts (obviously) and minions that can get out of hand (knife juggler / buffing minions etc) but it should ignore most other minions and go face. Maybe just a tick or 2 more in the direction of face is enough.
sepefeets
#318 Posted : Monday, June 27, 2016 8:32:13 AM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 899

Thanks: 1028 times
Was thanked: 584 time(s) in 238 post(s)
Dre;39071 wrote:
In example what i experience after the 121.1 & 2 upgrade.

I use this combo:

Code:


// [Combo] [1] Tunnel Trogg (LOE_018) > [0] The Coin (GAME_005) + [3] Feral Spirit (EX1_248)
LOE_018,0;nxttrn;GAME_005,0;EX1_248,0;mana:1;bonusfirst:30;bonussecond:60;hero:shaman


Ive checks tens and tens of logs and this allways worked perfect. After the update it suddenly does this:



it ignored my combo and that never happened before in the previous versions. Trust me I checked this so many times. It allways did the combo of Tunnel trog > coin + feral.

So uh, do you have any combos/did you change any combos with coin that had a penalty on coin? When I sim that board with only that combo then 121.2se, 118.7se, and botmaker's last build make the exact same move. As soon as I raise the penalty for coin outside the combo even by just 1 then it makes the combo move.

I also sim'd those not attacking with doomhammer logs from 121.0se and the one where it was played turn 5 and didn't attack is fixed by 121.2se. The other one where it was played turn 8 and attacked once and then didn't use it the next few turns is strange because after that first attack and continuing for the next few turns it thought it couldn't attack. I'm not sure wth happened there, possibly HR sending bad info for multiple turns or buggy SF code that I haven't touched?
Dre
#319 Posted : Monday, June 27, 2016 8:42:48 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;39092 wrote:
So uh, do you have any combos/did you change any combos with coin that had a penalty on coin? When I sim that board with only that combo then 121.2se, 118.7se, and botmaker's last build make the exact same move. As soon as I raise the penalty for coin outside the combo even by just 1 then it makes the combo move.

I also sim'd those not attacking with doomhammer logs from 121.0se and the one where it was played turn 5 and didn't attack is fixed by 121.2se. The other one where it was played turn 8 and attacked once and then didn't use it the next few turns is strange because after that first attack and continuing for the next few turns it thought it couldn't attack. I'm not sure wth happened there, possibly HR sending bad info for multiple turns or buggy SF code that I haven't touched?


Nope no penalty's on coin. Here are my files: https://dl.dropboxuserco...Aggro%20Shaman%20v5.zip I didnt chance anything since 23-6 in there. Also Yesterday i changed back to 121.0 and iam smiling looking at the logs again. Not only i had 55% wr again. But the plays it made amazed me sometimes even. Small things like the post above about the face-setting and other smalls things maybe. But whole different then the 20% and 30% wr i had with 121.1 and 2. I know how to make the silver files but iam not so experienced in code that i know what can cause this :(

Could someone try my files with HR 5.9 (latest) and silver 121.2SE (latest) - to make sure iam not crazy....

I would be easy if its me, i can uninstall and reinsall everything and start over to test. But iam pretty sure its not me. Bummer iam missing out on the updates tho because things like
(attack with tunnel before playing feral) should be fixed in 121.1 right? I think with thoses fixes the deck could reach so much more even.
sepefeets
#335 Posted : Monday, June 27, 2016 9:24:00 AM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 899

Thanks: 1028 times
Was thanked: 584 time(s) in 238 post(s)
Well the sim for that board never changed so maybe it never had the chance to do that play or you didn't notice it before but it's easily fixed with the penalty of 1 on coin outside combos.

Yes .2 plays the overload and then attacks with that board.
Dre
#336 Posted : Monday, June 27, 2016 9:49:12 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;39096 wrote:
Well the sim for that board never changed so maybe it never had the chance to do that play or you didn't notice it before but it's easily fixed with the penalty of 1 on coin outside combos.

Yes .2 plays the overload and then attacks with that board.


Iam confused. you are talking about this combo:
Code:

// [Combo] [1] Tunnel Trogg (LOE_018) > [0] The Coin (GAME_005) + [3] Feral Spirit (EX1_248)
LOE_018,0;nxttrn;GAME_005,0;EX1_248,0;mana:1;bonusfirst:30;bonussecond:60;hero:shaman


because this works flawlessly in 121.0. I have so many logs about this play on 121.0

here is one from last night where i rolled back to 121.0 with HR5.9


Like allot allot of logs, from 121.0 - Because this is the best opening for my sham deck. And then it suddenly misplayed it in 121.1/2. I didnt change the combo. So the play from above is without the penalty on the coin
sepefeets
#337 Posted : Monday, June 27, 2016 12:41:44 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 899

Thanks: 1028 times
Was thanked: 584 time(s) in 238 post(s)
You can put it in test.txt and see for yourself that the sim for that board does not change if you use different SF versions. That 2nd example sims the same between versions too. It looks like the combo is "ignored" on that first board because of the totem golem in hand. Note that with the way 2 turn combos work it is only calculating the first part of the combo this turn, and then on the next turn it will calculate the second part IF the first part is on the board, the second part can even come multiple turns later.
Dre
#338 Posted : Monday, June 27, 2016 12:48:48 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)
sepefeets;39104 wrote:
You can put it in test.txt and see for yourself that the sim for that board does not change if you use different SF versions. That 2nd example sims the same between versions too. It looks like the combo is "ignored" on that first board because of the totem golem in hand. Note that with the way 2 turn combos work it is only calculating the first part of the combo this turn, and then on the next turn it will calculate the second part IF the first part is on the board, the second part can even come multiple turns later.


Ah like that, ok :) well thanks for adjusting the combo then in the rank 5 tread. Altough ive reached rank 5 on 121.0 (pretty easy accualy) hopefully test will show it will work with the newer version aswell
parakeet
#339 Posted : Monday, June 27, 2016 1:19:15 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
Confirmed 36% WR with Dre's Aggro Shaman ver5.8.2.0 and v121.2SE.

Update: Over 8 hours, starting at Rank 8, I reset stats (without recording games played) to run another test
darkayo
#340 Posted : Monday, June 27, 2016 1:50:54 PM(UTC)
Rank: Member

Groups: Registered
Posts: 20
Location: FRANCE

Thanks: 1 times
Was thanked: 4 time(s) in 4 post(s)
parakeet;39106 wrote:
Confirmed 36% WR with Dre's Aggro Shaman ver5.8.2.0 and v121.2SE.

On how many games ?
parakeet
#342 Posted : Monday, June 27, 2016 11:18:40 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
121.0SE

Missed fatal: http://paste.ee/p/RlScA
Misplay/Poor play: http://paste.ee/p/bl99i
Wolves before Trogg (result of combo.txt?): http://paste.ee/p/yGz0w

121.2SE

Allowed Opponent to keep fatal on board: http://paste.ee/p/TjbeP
Not sure why AI attacked Etheral Conjurer with Flame juggler: http://paste.ee/p/e5f5Z
Cast Rockbiter but didnt attack with it: http://paste.ee/p/SAGDk



parakeet
#347 Posted : Monday, June 27, 2016 11:22:27 PM(UTC)

Rank: Advanced Member

Groups: Registered
Posts: 128

Thanks: 18 times
Was thanked: 83 time(s) in 69 post(s)
@sepefeet Whenever I try changing a log to "test.txt" and then run silver.exe I get the: "Input string was not in correct format" error.

Am I missing a step? What exactly do I need to do to simulate a board?
Dre
#341 Posted : Tuesday, June 28, 2016 2:32:32 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;39106 wrote:
Confirmed 36% WR with Dre's Aggro Shaman ver5.8.2.0 and v121.2SE.

Update: Over 8 hours, starting at Rank 8, I reset stats (without recording games played) to run another test


Glad iam not crazy, I had another day of 57% winrate on version 121.0
Dre
#349 Posted : Tuesday, June 28, 2016 3:34:26 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)
I've gone over some of my old logs from the day i was using 121.1 (25-6) and when my wr droped from 60 to 20%. Here are some highlights (i didnt sim this into 121.0 because iam at work) This is my shaman face deck with settings on face. As i mentioned before, on face settings the bot kills so many minions that are no treath at all. Iam not going to list those plays i think this happens in 121.0 aswell. It should only kill taunts or buffing minions on face settings. or a VERY efficient kill i.e. killing a 8/2 with a 2/1

skipped 1st turn when has a finley in hand (yes i have a cardvalue of 9 of finley so it doesnt play it instead of squire on turn on but that shoudnt cause it to not play anything at all on turn one) (the -value on squire didnt work) Also coin totem golem into next turn finley was good (I have no penaltys on those). But it did nothing at all.


Has 4 mana to play faceless but plays the very weak combo of heropower + Flametongue totem, only to kill of the enemy flametongue with squire.


Will add more...
darkayo
#350 Posted : Tuesday, June 28, 2016 3:44:39 AM(UTC)
Rank: Member

Groups: Registered
Posts: 20
Location: FRANCE

Thanks: 1 times
Was thanked: 4 time(s) in 4 post(s)
This night : rank 11 to rank 9.
45 Total victory, 50.00% winrate, middlerange Chaman.
ver.5.8.5.0 with 121.0SE
84 Pages«<1516171819>»
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 0.636 seconds.