Players have learned.
Once a city's defenses are whittled down, a bunch of coordinated attacks break the city to nothing in about a minute. Maybe scoutbombs soften up the troops before a large mech wave wipes up the remains. It doesn't really matter how a city is broken, this thread is about what happens next.
Once the first successful attack does any loyalty damage, it means all the defending troops have been wiped out. Most (if not all) of the wall defenses have been wiped by this point, too. After the first successful attack breaks through, it will be quickly followed by cav spam to grind down loyalty.
This period of cav spam tends to take less than 10 minutes. It doesn't matter if your loyalty starts at 100, or starts at 50. The speed of cav spam will grind the loyalty down to the critical "15" threshold very quickly.
Some basic algebra tells us that income is maximized at 50% tax. A lot of people set taxes to 0% to maximize loyalty. It is my contention that maximizing loyalty does not matter, since it will grind down to 15 very shortly after it the defenses get broken. The extra 50 loyalty, might take the attacker and extra 3-5 minutes to grind down with cavalry.
You should set taxes to 0 if you want max population for queueing troops. But if you can cover the cost of your heros within a tax rate of 50 or less, then I say set taxes to put the gold upkeep in the green. Gold is easy to get, so there's no need to gain it from taxes. But, why not make some profit if you can.
I guess what I'm saying is...
Don't lower your tax rate ONLY because you think it will delay the time it may take an opponent to capture your city.
The delay caused by high loyalty is no more than a few minutes. If you have other reasons for keeping the loyalty high, then by all means, set your tax rate to zero.



Bookmarks