What circumstances make Heighten Turning worthwhile?

Question

The feat Heighten Turning has the following as part of its benefit:

When you turn or rebuke undead, you may choose a number no higher than your cleric level. Add that number to your turning check, while subtracting it from your turning damage roll. (Libris Mortis 27)

So we're all on the same page, here's a summary of the special attack turn undead:

• When a cleric makes a turn undead attempt, the cleric first determines the maximum Hit Dice of the largest single undead that can be affected by making a turning check (1d20 + Charisma modifier), the result found on a chart that yields a Hit Dice outcome of between the cleric's level −4 and the cleric's level +4.
• Then the cleric determines how many total Hit Dice of undead are actually affected by making a turning damage roll (2d6 + cleric level + Charisma modifier). There is no upper limit to this result.

The feat Heighten Turning lets the cleric increase the turning check by +1 but decrease the turning damage roll by −1. The cleric can do this to a value up to the cleric's level. Reversing the polarity would've made the feat's use obvious in the face of a horde of low-powered undead creatures, but the way it's written instead makes the feat just weird: The cleric tries to turn a lone powerful undead creature while simultaneously limiting the size of the creature and his ability to affect it.

What are the optimal circumstances in which to use the feat Heighten Turning? That is, what numbers have to be present on both the cleric's side and the enemies' side for the feat Heighten Turning to be a good idea? Is the feat one that looked good on paper but is useless in actual play? Is there ever a certain point at which, or a certain build wherein, it's the ideal feat?

Answer 1

What are the optimal circumstances in which to use the feat Heighten Turning?

The optimal circumstances would be obviously a situation, where you reap only the benefits. But is this possible? It is.

The turning check as D20 has a 15% probability to end up with a value of three or less, the damage roll is guaranteed to be 3 though, assuming you are always at least level 1.

Level 2 Cleric versus 4 enemies with 1 hit die

A level 2 cleric is guaranteed to turn 4 hit dice, but has a chance of 30% to fail completely, by rolling 1-6 on the D20. This chance to fail is constant.

Chance to fail completely: 30%
Chance to turn all enemies: 70%
(100% - 30% chance to fail completely)

The group has a 30% chance to end up with 4 enemies. The group either ends up with no enemies or all of them.

Let's now use the feature to add the +1/-1 modifier to the rolls:

Chance to fail completely: 25%
Chance to turn all enemies: 73%

Okay, we improved the maximum hit dice odds by the 5%, we only need to not roll 1-5 now. But how can we get a better result by adding a negative modifier? Well, you don't, you still decrease your odds, but in sum and comparison to before, it doesn't matter.

There is a chance of 1/36 to roll 1 + 1 as damage roll, which is 3% rounded up. Only this way you end up with less than 4 total hit dice. So, there is a 97% chance to roll any other combination that saves your day. There is also a 75% chance to not fail the maximum hit dice check. The probability of not failing both checks (97% * 75%) is 73%.

Whenever the enemies total hit dice are near your guaranteed damage roll, you can reap most benefits.

Otherwise, the feature gives you options. 4 enemies together are stronger than the 4 individuals. It's usually more beneficial for the group to have a guaranteed reduction of enemies instead of sometimes many and sometimes few enemies.

Level 2 Cleric versus 1 enemy with 6 hit dice

This is a simple one. Your probability to make the maximum hit dice check is zero. Which is a really bad probability. But you only need to roll a 4 or more on 2d6 for the damage roll, which is quite a good chance - 91%! But it's a waste. What's the point of a nearly guaranteed damage roll if there is a guaranteed turning check failure?

A +2/-2 modifier will give you at least a chance of 5% to make the maximum hit dice check. And the chance to at least roll 6 as damage roll is 72%, which isn't that bad. Your overall chance (5% * 72%) is still crappy 3%. But 3% is more than 0%.

Answer 2

Speaking simply, on the high end or the low end.

When you desperately want to turn something with more HD than you have cleric levels, (aka turning a single high-CR undead), you might not need that much 'turning damage' aka 'amount of hd turned' but want a higher chance to actually affect it at all. The only real way to calculate this is through metagaming, or a GM who treats knowledge checks as 'you know the monster's stats'.

On the low end, when you just don't want to roll like, a 2, and not be able to turn the low-HD ghouls despite being a higher cleric level than they have HD, you can just take away a few 'turning damage' points to push that from 'needing to roll a 4 or better' to 'needing to roll a 1'. Aka, making the initial turning check a shoe-in.

Is the feat any good?

Nope. Get the ones which give +effective turning levels. Get the one that activates the relic item that does the same. Counting as a higher level cleric makes the turning check relatively pointless, with the amount of etls you can get for the cost of a feat. And if you're caring about turning, be a Sun Cleric and just straight up burn undead into actual cinders with Greater Turning.

Is there a certain build where it's good

Sorta. The maxed out charisma turning cleric build. Where you already have all the stuff to make your turning as murderous as possible, in an undead-heavy campaign. The high charisma means you can spend points of 'turning damage' without making it useless, so the 'high end' undead can be affected by your 1d20+charisma+heighten turning bonus'd check. Since 2d6 is unlikely to roll less than four, you're effectively going to be making the turning check 1d20+charismax2, with when you're really going for it, 1d20+charismax2+3 (average roll on 2d6 being 7), and with a charisma of 5 and average of 10.5, pushing the turning check into 'very unlikely to get below level+3' territory.