Unit design traits, the best and worst?

I agree, vanilla game has only a few traits and not all of them useful.

I choose fast all the time, like you said thing is just too op, finesse is pretty good for sword users and bloodthirsty for everything except late game empires(since they get it for free at their fortress), precision is cool for archers and mages most of the time.

I do not however use charge or defender at all, and i only use dodge traits with wraiths in which case balance is good but the one that gives +1 every level might be better late game.

Underdog works for the beginning of the game when everything has higher lever than your guys, discipline works for endgame units since endgame fortress gives several free levels which makes it better than precision in late game.

the one that removes 1hp and gives 1 attack might be useful for archers.

That's all i can think about right now.

Well, there's  the Zealot trait (+3 initiative, -1 HP per level) available to Wraiths if you're willing to train troops out of an Infirmary Fortress (grants Endurance I trait for +1 HP per level to units trained here). As for why I say 'out of an Infirmary Fortress' - Zealot + Wraith Blood means you have 0 HP per level. Adding Endurance I from the Infirmary to that brings you back to +1 HP per level. Zealot + Fast gives you +5 initiative, which is a little better than Fast + Impulsive (+4 initiative), but means that you've sacrificed whichever trait you would have put into the slot taken up by Zealot. Overall, I'd say that Zealot is bad to mediocre, depending on where you're training units.

Wraiths also have the unit trait Hallowed Rite, which generates 3 mana each time a unit with that trait kills something. This is another marginal trait but can be useful to fuel early spellcasting.

The Race of Men gets the 'Potential' trait for +25% experience, which stacks both with the 10% bonus from the blood trait and the various army experience bonuses, which can let you push your trained units up to high levels quickly to stack the level bonuses, or help counteract the penalty to the army for stacking champions (though this won't do enough to make that particularly worthwhile even so). It's a reasonable to good choice for Henchmen, but probably mediocre at best for regular troops (unless you're really stacking the level bonus traits).

Mancers get the Experienced trait, which is just like the Veteran trait except it also increases unit upkeep costs. At best, this is a mediocre trait, as one extra level does less for you than most of the other traits do, and the additional upkeep costs don't make it much more appealing.

Underdog and Bully are similar to Finesse except that they work on relative level. Having Underdog and Bully on the same unit guarantees that it gets a +3 attack bonus against anything that isn't of equal level. I don't know if these stack with Finesse, but if they do then Bully + Finesse + Fast will get you +6 attack against most computer faction troops, and Underdog + Finesse + Fast will get you +6 attack against many high-end monsters.

Unit Design Traits:

1. Acrobat, +1 dodge per level, 6 production
2. Balance, +10 dodge, -1 initiative, 6 prodcution
3. Bloodthirsty, +25% attack versus damaged enemies, 20 production
4. Brute, +3 attack versus lower level enemies, 10 production
5. Charge, +1 attack and +2 movement on the first turn of combat, 20 production (can be obtained for free from Watchtower Fortresses)
6. Constitution, +3 health per figure in unit, 10 production
7. Defender, +10 defense when defending, 4 production
8. Discipline, +1 accuracy per level, +1 spell resistance per level, 8 production
9. Enmity, +25% attack versus opposite alignment (Men/Mancers/Amarians/Tarth/Ironeers are Kingdom, Quendar/Trog/Wraith/Dead/Urxen/Krax/Slave are Empire), 12 production
10. Fast, +2 initiative, 15 production
11. Finesse, +3 attack versus lower initiative enemies, 6 production
12. Fury, +1 attack, -1 health per figure, 6 production
13. Ironskin, +3 defense, 14 production
14. Lithe, +10 dodge, -1 health per figure, 6 production
15. Muscle, +1 attack, -1 initiative, 12 production
16. Precision, +10 accuracy, 8 production
17. Rage, +25% attack when under 25% health, 10 production
18. Reap, +1 health each time the unit damages an enemy, 10 production
19. Scout, no movement penalties, +1 sight range, 4 production
20. Spell Resistance, +20 spell resistance, 6 production
21. Stealth, monsters less likely to attack, 8 production
22. Underdog, +3 attack versus higher level enemies, 10 production
23. Veteran, +1 level when trained, 18 production
24. Chain Armor Proficiency, 5 production
25. Plate Armor Proficiency, 10 production
26. Zealot (Wraiths only), +3 initiative, -1 health per level per figure in unit, 8 production
27. Hallowed Rite (Wraiths only), +3 mana per kill made by this unit, 8 production
28. Experienced (Mancers only), +1 level when trained, +1 gildar per turn upkeep (unsure if this is per figure or not), 25 production
29. Road Building (Mancers only), 10 production
30. Potential (Race of Men only), +25% experience gain, 8 production
31. Endurance I (Infirmary Fortress), +1 health per level per figure, free
32. Impulsive (Strike Garrison Fortress), +2 initiative, free
33. Charge (Watchtower Fortress), same as the other Charge

Out of the above, Balance and Lithe do more or less the same thing but Balance has the more severe penalty (-1 initiative from Balance, as compared to Lithe's maximum of -6 health, which basically sets its health half a level behind an equal level equal size unit; Ironeer blood or the Endurance I trait make this a third of a level, while Wraith blood makes this a full level, but regardless this is practically nothing). Fury and Muscle also do more or less the same thing, with Fury being clearly superior as it's once again up to -6 health compared to -1 initiative as well as being half the cost of Muscle. Reap and Rage are kind of useless (Reap's healing is not per figure; if you have a 1 figure unit with Reap, it will heal exactly the same amount as a 6 figure unit with reap will each time it does damage; Wraith Blood has the same issue), and Enmity is a better form of Bloodthirsty if you can guarantee that you'll only be fighting opposite-alignment foes with the unit.

Dodge-based units should probably take Balance + Lithe + Acrobat, though if you want to sacrifice any of these traits to take something else (like Fast), I'd sacrifice Balance first. More or less the only reason to take Precision or Discipline is that you're fighting a good dodge-based unit. Discipline and Spell Resistance can toughen a unit against spells, but a mid-level Mage champion, or even a high level non-Mage champion, will blow through any amount of spell resistance a trained unit can come up with so this is mostly a defense against spellcasting monsters and low-level champions. In my opinion, if you're going to take dodge traits, you should probably go all-out; lots of dodge is much more useful than just a little, and Acrobat is necessary to prevent your dodge-based defenses from being eroded by the accuracy bonus from unit levels (+1 per level, +1 more per level for Mancer Blood, +1 more per level with Discipline; Acrobat will cancel out one of these bonuses by granting +1 dodge per level). Especially with Wraith Blood, it may be tempting to sacrifice one of the +10 dodge traits, but remember that hit chances are computed as (Attacker Accuracy) - (Defender Dodge), and with Wraith Blood + both 10 dodge traits + the 10 dodge robe you can get to 50 dodge (meaning a 20% hit chance for a 70-accuracy unit, about normal for low-level trained troops); dropping down to 40 dodge from this point means that that same 70-accuracy unit hits roughly 50% more often and on average hits for more damage. The difference in the potency of dodge-based defenses gets lower when comparing smaller dodge values, e.g. a 10-dodge unit will be hit 20% more often by a 70-accuracy unit than a 20-dodge unit will, though this is still a significant difference if you're basing your defense around dodging attacks (note also that the higher dodge units will also be hit for on average less damage than the lower dodge units because each figure in a unit makes a separate attack role when determining hit damage). Also note that since units gain accuracy with levels, Acrobat is almost essential for preserving the utility of your dodge-based units against high-level opponents.

Veteran and Experienced are not, in my opinion, worth taking; the production cost is too high and you really don't gain that much from just one extra level, especially if you're training out of a Fortress that already grants a couple extra levels. Henchmen and Sions are an exception to this rule, as they gain traits like champions do and so a free level may be a good deal for them.

Scout is a useful mobility enhancement for your armies, but you'll have to sacrifice something for it. Might be worthwhile if there is a lot of rough terrain around your territory and your champions have compasses or some other way to make up for terrain penalties (e.g. mounts when you're using infantry troops). Stealth isn't worth using except on lone units.

Brute, Underdog, and Finesse all offer about the same thing with different conditions; of them, Finesse is probably the most useful as it's relatively easy for the player to ensure the initiative advantage over computer faction troops and most monsters. None of these are bad, though Underdog will lose its utility if you keep your troops alive and develop them to a high level.

Constitution and Ironskin are both good traits for blood tanks. Constitution grants 1.5 levels worth of health to an average unit, which isn't great but helps with tanking or to counteract health penalties (e.g. Wraith Blood, in which case Constitution grants 3 levels worth of health, or Lithe or Fury). Ironskin makes a unit generally harder to kill and can be especially valuable in the early stages where at best you have leather armor. Defender is good for a tank that is meant to sit still and block (possibly with counterattacks if using a sword unit; this is a case where the initiative penalties of Muscle and Balance might be less of an issue than the health penalties of Lithe and Fury, as the unit is just intended to block the enemy's melee line rather than to be a main damage dealer, and initiative is not that important in this role), but is mostly useless on a unit meant to make attacks on its turns.

Good post joeball.  Thank you.  So many of your posts would make wonderful additions to the Wiki, should you ever feel like contributing.

I will use defender on swordsman units as they can tank and still counterattack.

I go through most of my games with Fast, Finesse Spears with either Constitution/Ironskin/Lithe/Balanced for the last trait depending on mood/spellbooks/essence slots available.  Horses over wargs for mounts for extra map mobility.

Ranged troops will get Precision, Fast and Lithe for dodge.   

I almost always play a No Armor custom faction based either on Mancer or Tarthan blood and the armor slots just seem a complete waste.  The Upgrade button for Armor sucks too as it always tries to remove my Soldier's Gloves and Boots which are also handy.  I end up designing new units with improved shields just to avoid this.

Would enjoy seeing quests, heroes, unique buildings or random treasure drops that add new traits.  Dsraider has done a lot of work in this area with his excellent Blood and Mana mod.

Waiting for an updated D and W to blend with B and M, BMB, Paladin/Sorceror and Monuments mods.  

If you would like to add them to the wiki, you are welcome to do so. I'm not interested in editing webpages, however.

I might just do that, thanks.

Most used:

Fast/Finesse, Constitution, Precision(necessary for ranged units now)

Commonly used:

Bloodthirsty, Ironskin, Defender, Stealth/Scout, Plate Armor, Potential

Sometimes used:

[Muscle/Fury/Brute, Enmity, Underdog]*, [Balance/Lithe/Acrobat, Zealot]**, Charge, Rage, Reap, Spell Resist

Rarely or Never used:

Veteran, Chain Armor, Discipline***, Experienced, Hallowed Rite, Road Building

*I will commonly use combinations of these alongside Bloodthirsty and Charge with Tarth to gain exceedingly sick damage boni in small armies

**I commonly use these traits with Wraiths for high Dodge bonuses

***Sometimes used with Mancers to enhance their Lucky trait, but I just don't care for the +per level idea on trained units

The main caveat to all of this is what kind of combinations I can make with City Enchantments, Fortress Level Ups, or Unit Enchantments. 

Good feedback all thanks, picked up a couple of uses for things there that I hadn't really thought of.

Fast just has to be tweaked I think, if it was less obvious it would bring a lot of other stuff more into play.  Like if it was only +1 Init and say 10 production, straight away it wouldn't be quite a much of a no brainer.

Yeah he's on a bit of a roll, I agree they'd be great in the Wiki.

What's the wiki URL?

http://elementalgames.org/

Personally, I think I'd sooner see it have a drawback like -10 accuracy than have it be cheaper with a lower bonus. Another alternative would be to make it a 'first turn only' bonus like Charge (possibly instead of first turn only, have it work for a few turns and then turn off, or, perhaps better, work for several turns, then turn into an initiative penalty to represent tiring out, though I don't know that the latter is possible with the current XML files).

I tend to agree in principle, I just think it's too easy to stack init bonuses currently.  Another option would be to split it into 2 abilities... something like  this maybe...

Fast (+1 Init), cost 10

Twitchy (+1 Init, -10 Accuracy), cost 5

I have a trait used on some monsters in D&W called "Quick off the mark" which is +2 Init and +1 Move first turn of combat, I guess that could make a good unit design trait which is similar to what you suggest.  What do you think that would that be worth, 10 production maybe?

Yeah you could achieve that by making it +3 init for the first 3 turns, -1 init for the entire battle (and have them stack).  Interesting idea also.

True, but that's an issue of Fast (2) + Impulsive (2) + Aura of Grace (1 to 6) + swords/spears (0 to 4) + non-plate armors (-2 or 0). Fast is just the tip of the iceberg there. If the computer were using the kinds of troops that the player was using, instead of stacking initiative penalties with full plate + clubs and not taking mitigating traits like Fast or training out of Fortresses to pick up Impulsive or using the Aura of Grace enchantment, then Fast would be much less of an issue. Even if I use a unit of identical design to what the computer uses, I can still expect to have a 4 or 5 initiative advantage if I want to (Strike Garrison in a 2 or 3 essence Fortress with Aura of Grace, or one of the rare 4 or 5 essence sites with Aura of Grace).

The issue is more one of the computer spitting out 15 initiative units while I'm spitting out 25+ initiative units. Changing Fast so that I'm 'only' spitting out 24+ initiative units isn't going to make that much of a difference; instead of acting 1.67 times as often, I'll act 1.6 times as often. What would help the computer more is if it would bother training those really low initiative units out of a Fortress with a Strike Garrison (and preferably also Aura of Grace) so that it's looking like 17-20 initiative against my 25+, at which point the unit might actually be competitive since I'd be acting 1.47 times as often (17 initiative to 25) to 1.2 times as often (20 initiative to 25).

I'll also point out that going from 20 initiative to 22 initiative is basically a 10% damage bonus if something like Finesse doesn't come into play. For a large portion of the game, this is about equal to boosting your attack score by 1 point, at least from the math perspective (the exact percentage bonus varies depending on the attack and defense scores in play, but with a base of 21 attack against 40 defense, a +1 attack bonus is about an 8% damage bonus per strike; at a base of 15 attack against 20 defense it's a little over a 10% bonus, while at a base of 9 attack against 10 defense it's a bit over a 17% bonus). It's really only the combination of the computer not counteracting the initiative penalties it picks up on its units with the player either counteracting the initiative penalties or just stacking initiative, in combination with Finesse, that puts Fast over the edge, and reducing Fast to +1 initiative isn't going to solve this (in fact, it may make it worse for the computer faction units which use fast, as for example 24 versus 21 initiative is a better deal for me than 25 versus 22 initiative, though it will very slightly erode my fast unit's advantage over the computer's slow units).

For what it's worth, I feel that altering Fast changes the wrong side of the initiative problem. If we come across unit designs that really need a bit of extra initiative, we can always go find the unit template and add the Fast trait. We can't force the computer to train its units out of an Aura of Grace Fortress with the Strike Garrison improvement, and so to me it's the effects of Aura of Grace and Impulsive that should be changed before Fast is. (I am not saying that Strike Garrison is the best of the Fortress 3 upgrades or that you should always go for it. Which of the three Fortress upgrades you take at level 3 ought to depend on what you're trying to do there; Impulsive and other initiative bonuses are better on slow units like archers or plate-and-maul units, whereas Endurance is great for making tank units and Charge is good for getting your melee line into the fight quickly and potentially getting first strike.)

A different solution would be to increase the base initiative score that everyone gets rather than to change the bonuses we can have; if for example the base becomes 30 initiative, then the computer's worst units might have 20 or so initiative while my best might have 40 so I'd act twice as often, as compared to now where the computer's worst units have 8 or so initiative and my best have around 30 (so I act almost 4 times more often). A more reasonable comparison might perhaps be a -6 penalty on the computer unit to a +5 bonus on my unit, for 14 versus 25 initiative currently (I act about 1.78 times as often as the computer) or 24 versus 35 initiative with a base initiative score of 30 (in which case I'd act about 1.45 times as often as the computer). Note that changing the bonus on my end from +5 to +7 changes the current case to about 1.92 times as often and the case with the base score of 30 initiative to 1.54 times as often. However, doing this would most likely entail a lot of work in changing the initiative scores of the various monsters to maintain the same relative action frequencies so as to avoid upsetting the balance of the expansion phase of the game.

10 probably isn't bad. It's a little weaker than Charge is for ensuring first strike because the movement is a bit lower, so it's alright for it to be a bit cheaper. I would be a bit concerned about stacking Charge with Quick off the Mark, as then you'd have +2 initiative +3 moves +1 attack on the first turn, and that +1 attack could be further improved with Finesse or some other attack-boosting traits (particularly if you pick up Charge from the Watchtower rather than from unit design), so you might want to make it a bit more costly than just 10 production, but I don't think it'd be terribly out of place at that point. Of course, you might also be able to prevent it from stacking with Charge, which would accomplish the same thing.

That was in fact what I was thinking of, though I don't know if it would work (which is why I said something along the lines of 'if the XML allows it').

Yep I think you make good points.  I'd still probably like to see Fast be a bit more interesting somehow, maybe adding in the -5 or -10 accuracy penalty like you suggested might be the way to go.  

Yes I think this is the way to go with initiative.  At some point in future I want to trying modding this in (base init of 30 or even higher) and see how it plays.

That and (as you said) getting the AI to use he init boosting spells... would make a huge difference.  DsRaider has been campaigning for it for a while.  Maybe we can get it looked at in 2.0 assuming that's coming at some point.