CAR Mod

[stmartin]

I did merge the CAR mod and it definetly does not reduce turn times by 15% (unless everyone plays huge maps these days and does an ai Autoplay for 400 turns before playing). Even if the average speed gain is much lower than 15% it is still a very nice mod.

Thanks for using our Mod. BTW, do you have some test result that I can take a look at? The speed change in our experiment is on the first 90 turns of AD1000 scenario, so I'd like to know CAR's effectiveness in other experiment settings.

 

[Sephi]

Sorry, I have no statistical significant data. I did two AI autoplays on a Huge Map but the AI acted too different in them for a valid comparision.

 

[xienwolf]

If you maintain the Cache variables in your ::read and ::write functions, but disable the actual use of them and other changes to the code, you can be savegame compatible with/without the mod up. Then you can run tests with a single savegame and random seed disabled. Ought to provide exactly the same game each time (just be sure you load after the initial save so that you are on the savegame seed, and not the actual game seed. They are not the same IIRC. Hence on the FIRST reload you might get a different result, but the result you get then is the one you will ALWAYS get with that save)

 

[Agent327]

I did merge the CAR mod and it definetly does not reduce turn times by 15% (unless everyone plays huge maps these days and does an ai Autoplay for 400 turns before playing). Even if the average speed gain is much lower than 15% it is still a very nice mod.

I usually play on Huge maps (for playtesting as well), so that might still be interesting. I'm assuming you merged with the mod in your sig?