Updated Dre configs on github.

@er009 - Someone noticed you had the wrong card id for kindly grandmother in your hybrid hunter configs and gave a PR with the fix. A lot of the boss match cards share names with the playable cards but all playable cards start with "KAR_" in their id's.

Deckrules dont effect silverfish AI, read my guide how to set up silverfish. So you can leave that on default always. In the files for the shaman settings.txt is included, like always :)

Yep :) Tested the warrior myself short though, wasnt good...

midrange shaman, turn 2 vs lock. coin into lightning storm on argent squire and dark pedler. this is silverfish AI idea to do such play?

How do I run them on basic settings? Tell the bot to not run the extern AI and then choose one of the 5 default settings (ie deckrulesrush.xlsx)??

Edit: here's what I did and will continue to run through 50 wins unless it's wrong (playing the beast druid deck) ...

I moved my entire Silverfish folder to my desktop to preserve everything after testing, then downloaded the newest versions Sepefeets put up a few hours ago. All I did was say to use externai and didn't modify anything, but I'm a little worried as the mulligan and combo files in there are for standard midrange hunter. Does it not matter because none of those cards are in the deck?

http://i.imgur.com/R7aXe9y.jpg

No no dont do this - just take out the mulligan and combo files and leave the setting.txt in there what comes with silverfish. Dont turn of silverfish. Also dont run a deck with wrong mulligan because it will discard all.

I ran the Dragon Warrior deck for 30 games starting at Rank 18. I made it to Rank 15; my stats are as follows:
16-14 (53.3%)
Druid (2-0) Hunter (4-3) Mage (1-4) Paladin (2-2) Priest (2-0) Rogue (1-0) Shaman (0-1) Warlock (1-1) Warrior (3-3)

from mid next week i can contribute a bit to the testing part. these days i dont have much time :)

New season looks good. So far and from other forums and inputs from friends. aggro shaman and mid level are some of the best botting decks (even gone till rank 3). Not sure, but there should also be a few more decks of course :)

All the best,

Dildogaggins

Thnx for the 2 tests! So basilcy they have both potential, guess it's writing time hehe.

@Carins ah yeah sorry, i tought it would do nothing with the Fandral card, will replace it by something else. - its not mandatory

@BlackOak fancy doing the warrior so i do the druid?

Yes, also since I do this for quite a while i kinda know what the bot finds dificult and what not. So first i look at the deck like that. For idea's I open up like 4 good rated Beast druid decks from different websites and check differences. With the bot in mind I now made the following changes: -Fandal + An extra Tiger (it only had one) and -Druid of the Flame +Mounted Raptor. Yes i see the potential for Fandal but it will be to hard for the bot to combo right all times, also allot of beast-druid decks don't run him. I personaly see more value in the Raptor then the Druid of the Flame. Also other decks made this swap. Also the raptor is better for the bot.

Now i make a mulligan for all classes, think about nice curves and combinations to keep. I start without any combo file and then let it run a night. The next day I have allot of logs and that is where the work begins. From here I look what classes i was weak against and make changes in the mulligan for that class. And then track and trace all misplays I find in the logs. At first I look mostly at opening turns because i can force them easly in the combo files and openings are the most important. AI misplays i post in the silverfish treads and deck misplays i fix in the combo files. Indeed like you said. from 10 misplays to 9 etc. Fixing a misplay in the combo file can result in 2 new misplays being made. So there is were some experience comes into play xD

And yes, you can test my Warlock - It has complete files and all. Posted in the same post. It hasnt been tested at all last 2 months. But it might be good again.