Liminal Alignment

[fabi]

Current trunk and future release 1.9.10 features the new Liminal alignment again.

There is no longer the need to define a time schedule that supports the alignment.
Thus, just set the "alignment=liminal" attribute in [unit_type].

Liminal units receive the lawful_bonus of a time of day as a malus on damage no matter if it is positive or negative.
With the default schedule that means -25% during Morning, Afternoon, First Watch and Second Watch.
The Liminal unit does not get any modification at Dusk or Dawn.

[Mountain_King]

So, to make sure I understand this correctly:
As of 1.9.10, the possibility for liminal units will again exist within the game. However, they never receive any ToD buffs to their attacks, only nerfs. Is that right?

[Espreon]

So, to make sure I understand this correctly:
As of 1.9.10, the possibility for liminal units will again exist within the game. However, they never receive any ToD buffs to their attacks, only nerfs. Is that right?

Yup, no buffs. Just nerfs during day and night with neutrality at twilight.

[fabi]

... However, they never receive any ToD buffs to their attacks, only nerfs. Is that right?

Yes, you understood correctly.

Please understand as well that there is no difference between the two solutions from a gameplay or mathematical point of view.

The problem with the liminal alignment is that it works different than all three other alignments.
Neutral units always have constant damage output, lawful and chaotic ones do have the same amount of time where they get the bonus as they have times where they get the malus, so you can compare their damage values.
At least with the default schedule.
If you take the schedule of the UtBS campaign where a cycle consists of 6 times with lawful bonus, 5 times with chaotic bonus and 4 with no bonus
then the situation is similar to the liminal problem.
You can not compare the base values of a chaotic and a lawful unit anymore (even a neutral/lawful or neutral/chaotic compare is off).
Worst is of course the underground schedule.

So the concept of considering the alignment and the actual time of day schedule is already a skill the Wesnoth player is used to.

If you give the liminal unit 25% better base damage values with the new implementation of liminal you will get the same
values as with the old implementation.
Right?

The old implementation would have boosted a liminal unit by the same factor the unit is harmed with the new one.
That means the designer would have had to lower the damage of liminal units to keep an already balanced system balanced.

For the player the difference in the two implementations is the following:
old implementation: "Liminal units are stronger as their base values suggest because they don't have times of malus."
new implementation: "Liminal units are weaker as their base values suggest because they don't have times of bonus."

Still the result is a unit that does it's most damage at twilight.

On the other hand does the new implementation lift some issues with the older one.
There is no more need to define an extra time schedule.
Every time schedule you can create in Wesnoth will be ready for liminal units.
The effect of liminal is easier to get, it does not depend on extra information as the old one did with the extra "liminal_bonus=" attribute.

So no extra work for the wml coder, just assign the alignment and adjust the damage for balancing reasons.
On the c++ code side the new implementation is very much cleaner and easier.

[Elvish_Hunter]

Just to add one more detail to what fabi already said, with the old implementation the Illuminates ability (and every other illumination-related ability) didn't have any effect on Liminal units. With the new implementation, it should work right out-of-the-box.

[Alarantalara]

Please understand as well that there is no difference between the two solutions from a gameplay or mathematical point of view.
....
If you give the liminal unit 25% better base damage values with the new implementation of liminal you will get the same
values as with the old implementation.
Right?

Wrong. The new 25% damage reduction will result in less damage than the other way around.

Consider a unit that had 16 base damage under the old system.
During twilight, it will do 16*1.25 = 20 damage and 16 at other times.

Now look at a unit that does 20 base damage under the new system.
During day or night it will do 20*0.75 = 15 damage at twilight.

Thus, from a mathematical perspective, the two methods are completely different.

This is not a complaint about the new system, just a clarification. The ability for it to work well with illuminates makes the change superior in my opinion.

[fabi]

Please understand as well that there is no difference between the two solutions from a gameplay or mathematical point of view.
....
If you give the liminal unit 25% better base damage values with the new implementation of liminal you will get the same
values as with the old implementation.
Right?

Wrong. The new 25% damage reduction will result in less damage than the other way around.

Consider a unit that had 16 base damage under the old system.
During twilight, it will do 16*1.25 = 20 damage and 16 at other times.

Now look at a unit that does 20 base damage under the new system.
During day or night it will do 20*0.75 = 15 damage at twilight.

Thus, from a mathematical perspective, the two methods are completely different.

This is not a complaint about the new system, just a clarification. The ability for it to work well with illuminates makes the change superior in my opinion.

Oh yes,
thank you for pointing it out.

We can agree that old_system == new_system +- 1 ?

Having an additional malus for liminal units is in fact a good thing.
The liminal alignment is the best in town, it does not have a malus, no matter which system you use to implement it.

[Drakefriend]

If you want to change a preexisting unit's alignment to liminal, multiply it's old damage with 1.2, as this means the average damage over a ToD (not taking into consideration rounding) remains the same.

Proof

The total boni and mali of ToD should be balancend, so the factors of Day, Twilight and Night should sum up to three (lawful: 1.25+1+0.75=3; neutral: 1+1+1=3; chaotic: .75+1+0.75=3). So the factor x the base damage of a liminal unit is modified with is the one where this leads to a total of three.

0.75x+x+0.75x=2.5x=3 |/2.5
x=1.2
Quod erat demonstrandum.

A liminal unit thus efectivly gains a bonus of twenty percent at twilight and a malus of ten percent at day and night.
On the example given fabi and Alarantalara's correavarage iction, It would be better to use twentyone instead of twenty, as is closer two the average of the original damage, because one point less for four turns a six-turn ToD cycle is more, twice as much, as on point more for only two of six turns.

[fabi]

Thank you for doing the math

[SigurdFireDragon]

Not sure if this possible or even desirable, but as something to consider:

What about giving the limial a 50% bonus at dusk/dawn and keeping the malus of %25 at all other times?
That way one wouldn't need to apply the * 1.2 formula suggesed by Drakefriend to convert a unit to liminal, one could just change the alignment and it would still deal the same average damage over the ToD.

This would make it easier for umc's to convert existing units to liminal, and would make it a bit more obivious to players about paying attention to ToD with liminal units

Though I'm not sure about the effects of how such a swing on damage amount might affect how such units would play

[Alarantalara]

This suggestion would make the code more complicated. The current method is to calculate the time of day bonus normally and then make it always a penalty. Your suggestion would require either a) hard coding a special value for the 0 case, or b) increasing the time of day WML complexity again, with the associated problems with interacting with illumination effects.