Originally Posted by
Ethan
[Found the long post - just expands on the simple point above.]
Ok, lets do this again then. You say a 238.65 defense is good. You get 827.63 total military with 30% TGs, congrats.
I use 19.1% TGs, 10.9% Forts. I have 57.25 elites, 42.75 dspecs. I get 238.66 defense, so I'm just as safe, and I put out 610.79 offense. Giving me a total of 849.45. Oops, looks like a mix won again.
Maybe you can do better. Lets try and get a higher total with your approach again.
Lets say, seeing my great new numbers, you go to 59 elites, 41 dspec. Gives you 664.81 offense, and 205 defense, for a total of 869.81. Great, you retook the lead by 20 points.
Now, can I do better? Lets see... I go to 22% TGs, 8% forts. I use 62.32 elites, 37.68 dspecs. 676.37 offense, 205.04 defense, total 881.41. Up another 11 and change points!
What, I beat you again? How is this happening? Well... you'll just get more offense and less defense yet again - but wait. You've finally reached the point where it is a bad strategy to lower your defense more, cause you'll open yourself to doubles, or triples, or whatever. Congrats, you finally found the target defense - the point where more offense isn't worth the lost defense.
And a funny thing happened at said target defense... the TG/Forts split was better. But wait, you cry, the optimization maybe made the defense go too low! Nope - *all* of the defense drop happens when you try and beat the total military by getting more elites and fewer dspecs. Ever time *I* optimize, the defense either goes up or stays the same, so *my* optimization is always safe. While your must, eventually, fail due to insufficiently low defense. You were already in TT range of yourself before I stopped the chain as it was.
If your argument is that your *current* defense is higher than it needs to be... then you are being dumb even before you try and optimize. Your target def is X, and you are running 1.2*X. Lower your defense to X and you'll be better by definition. NOW, after getting to your target defense, optimize TG/Forts. Whatever the optimal result... you *can't* do better, again by definition.