"Headless" behavior

by **custos** » 08 Jun 2020, 07:55

I have been trying to balance my decks so I attempted to run each against each in sim/headless mode. I ended up generating calls to cmd along the lines of:

".\\forge.exe sim -d \"{deckNames[i]}.dck\" -d \"{deckNames[j]}.dck\" -n 10"

Then checked the forge log file for results and found a curious behavior, asides from not playing 10 matches, always the first deck won, 100% of times. If I switched the order the same then the other deck won all the time, if I ran a simulation from GUI then it was not one-sided. Btw I did a full 30 vs 30 run, and it was this way for each. What am I doing wrong? Is it just a bug? Since I come from a C# background the project structure is foreign to me, where should I look to try and fix this?

by **custos** » 11 Jun 2020, 20:30

Figured it out after I managed to get the debugger working. For the -d parameter the code expects all decks to follow, and not multiple -d inputs. Because if you use multiple -d for example: ".\forge.exe sim -d deck1.dck -d deck2.dck -n 10" it will only take into account the second one and it will simulate a game against nobody, an win all the time. Also if you have a space in the deck name use "" around it.

<tl;dr> if you use the simulator use ".\forge.exe sim -d deck1.dck deck2.dck -n 10" instead of ".\forge.exe sim -d deck1.dck -d deck2.dck -n 10"