-
Suggesting this in response to 14 meta being popularised over the last few weeks, which some people don’t like for two reasons:
1. The game is unbalanced (I claim the libs may win as much as 75% of…
-
How do relate the rate, that was used for KataGo networks strength benchmark (https://user-images.githubusercontent.com/11942395/107119904-830cd880-6858-11eb-8959-3fc9571246c1.png), and the Elo rating…
-
In the battle mode, KataJigo may choose points with a winning rate of less than 50% that is not in the analysis mode.
Analysis by KataGo or KataJigo:
![图片](https://user-images.githubusercontent.co…
-
Does edge positioning not work with stoploss guard? I had a dry run bot running last night and 15 trades hit stoploss within ~10min despite having stoplossguard configured to lock all pairs after 3 st…
-
I think *4 Charms* sort of conflict with *Safety Charms* with color clues.
#### Solution 1
* *4 Charms* can only be performed by rank clues.
* This solution preserves *Safety Charms*.
#### Sol…
-
* Operating system: Linux Ubuntu 20.04
* Python Version: 3.8.5
* CCXT version: ccxt==1.40.30
* Freqtrade Version: freqtrade develop-adb3fb12
I'm unable to trade in binance with USD…
-
I think this is a good idea and thanks for making this function...:)
I've got a quetion about logics of choosing weakened move.
In some cases winrates of 1st and 2nd best moves would be much dif…
-
[ShiroGo V2](https://github.com/yenw/ShiroGo)
---
* Outline
* Compile MXNet C++
* Compile ShiroGo
* Set Selfplay Task && Run Selfplay
* ...
* many parameters are not adjusted optima…
-
My AI kept randomly not doing anything in some battles, so I did some investigating.
Turns out these errors were being put into the log:
```
INFO - >>> battle-gen8randombattle-1220084985|/timer…
-
A few thoughts:
- At first glance knowing that the deck is winnable might seem like an advantage, but for those who play for winrate, they are already logically assuming that the deck is winnable w…