A Guide to Thievery (Non Halfers and Rogues)

[Bo To]

That's why I said faery is more resource controller: with fb/lighting strike/tornadoes you do more dmg I agree but you dont need mystic perso for them. The best about faeries is they have also tog. While elves should do pf/ms/chastity which consumes a lot mana/runes, faeries can do everything else better or they can feed dragon fast if needed. Should you spend a whole perso just for +1 mana... I dont think so :). Now this will be my last post about that because its off topic...

About the topic: generally if u are going to use your thieves for something more than intel you will need atleast 10% TDs. Are you going to have the room for them is another question. Also when war starts you will lose/gain land and your tpa/TDs will drop so it will be harder to do other ops than intel and to be honest: i really hate when someone of the attackers in KD spent his stealth for something else than intel.

[noobium]

First off the basic mechanics of thievery, max gain ranges, and the typical success rate of ops, is useful information to know. Riots is a surprisingly easy op to land, but requires a lot of thieves sent to get good results. Sometimes an Avian with 3-4 TPA can manage it, and it might make sense to have attackers chance riots if your dedicated riot machines are out of range.

For an attacker TPA is often use-it-or-lose-it, so it can make sense to take risks with stealth and thieves, even with an estimated 50-60% success rate. Really depends on the thief strength of core, and the likelihood of attackers unloading their stealth in retaliation. Low-yield robbery and nightstrike do not risk too many thieves, but enough stealth needs to be held back for intel gathering. Against an undead core I'm far more likely to use offensive thievery with core attackers, since the need for defensive TPA vs. other attackers isn't as necessary.

Attackers can run a fairly wide range of TPA and still be good, though orcs and undeads typically keep 1.5-2 at the most, and sometimes less. Against less coordinated kingdoms an attacker can often get away with 1tpa. Avian, Dwarf, Human range anywhere between 2-4, Human can go higher. Those races at the lower end of nw chart might consider adding tpa, if they believe they are less likely as initial attack targets. Reading your opponents' TPA and crime sci levels beforehand is helpful for making adjustments, whether defensive or offensive.

Training extra thieves during war is a pretty useful tactic. It seems expensive to train lots of tpa, but thieves are in a lot of ways like a dragon project, if they are built on provinces that are reasonably secure against losing thieves. The problem is knowing when and where to do this, and balancing thief training with other economic needs for a kingdom. In the long term, a kingdom that can set up an mTPA advantage has some advantages in landing ops.

Any thief role is very luck-dependent, and selecting which ops to use follows a strategy similar to poker - do the thing that has +EV every single time. Unfortunately there is no clear-cut knowledge of success rates, but players can make rough guesses, and can control how many thieves they lose in the worst-case scenario. A big newbie mistake with using thieves on attackers is to assume that they're going to land all of their ops all the time, and they'll always get awesome damage, and that no one will adjust their TPA or watchtowers during war if they see a core with strong thievery. So consider, if you train thieves on an attacker, you are looking at what those thieves will do later in war, not just thinking "OMG imma nightstrike everything!"... and tbh in a lot of cases it's not necessary, given that nightmares can drop target tpa effectively.

A long time ago I did some rough math and figured that after around 2 raw TPA, a province should look at improving their crime science before they worry about jamming 3-6 TPA on a province, unless they are playing a dedicated thief role and need the TPA for obvious reasons. A bucketload of crime science isn't always available on short notice, while an extra TPA or 2 can be planned with greater ease. A 200 bpa : 100 bpa crime sci advantage is effectively +16.5% TPA, and while this doesn't overpower a similar tpa outright, it makes it far easier to build up an advantage or win a battle of thief attrition.

Thief attrition is something to keep in mind as well, and a major reason why Tactician is so good is that they lose no thieves to intel. This is important both for t/ms and attackers, and really important for Halflings since their really strong point is their ability to win thief attrition battles. Gold dragons are useful for winning thief attrition battles, both for reducing the impact of enemy watchtowers and reducing the utility of thief dens; although the majority of Gold dragon benefit is passive (hence why a lot of kingdoms don't like using them), it has some uses.

Finally;

Rogue and Mystic are only viable on provinces that lock themselves into thievery or magic. Avian, Orc, Undead, and Dwarf should never lock themselves into a mage or thief role, since those races will always want to be playing with a solid offense. Humans can manage Mystic fairly well due to their science granting them the second-best overall wpa, but are on shaky defense; the human/mystic is looking at being able to stack acres and peasants in war, which they can guard a lot easier than other humans, but they are not going to be able to keep very high wpa, so they are looking at using nightmare rather than MS. Humans do not manage Rogue as well for a variety of reasons, since it locks them into running at least 4raw TPA.

As far as picking between Faery, Elf, and Halfling; for most purposes Faery is by design the best pure t/m no question, because of Faery's versatility and eventual science lead over their competition. Elf is capable enough as a pure t/m, but a Halfling is really built for the attacker-thief niche and for winning thief attrition battles. A nub mistake of Halflings the world over is thinking "imma gonna AW and Prop and steal runes with no fails whatsoever and it's gonna rock!!!!!!", only to burn through their thieves with a **** ton of fails and/or get themselves chained or massacred into the ground, or to sit as useless husks while their support gets wrecked. Elf wpa is a lot more consistent, and in a lot of situations a spellcaster simply has to suck up a huge fail rate to land MS on a difficult target - in that battle, Elves are preferable to anything else. Elves are pretty effective as counters to large and unprepared banks, less effective but still decent as counters to other t/m, but their strongest niche at lower tiers is being able to hit easier than faes, and being able to handle land bloat and still cast offensively. Faery's biggest selling point is their defense and econ as much as their damage output, though their damage output against a vulnerable target is better anyway.