New model worse than previous?

[Emx77]

Is there any reason why Shining Star III has so much worse soft and hard attack ratings compared to Shining Star II? New model is just upgrade and have same weapon load (10 x 500 kg Precision Bombs).

[Akrakorn]

Because its firepower reached its threshold and "converted" the damage for an extra attack per round, it's the stat below the attack stat. Artillery does the same thing.

[zgrssd]

As Akrahorn said, on some units tehre is a maximum damage/attack.
If modified firepower goes beyond that, you get extra attacks.

Artillery is the most common example, but apparently it applies to Aircraft as well?

[Emx77]

Guys, thank you for explanation. It make sense. However, that opens new questions. For example: Which of these two fighter bombers is having more chance to destroy 300mm Polymer plating monitor tank? In other words is it better to attack such tank with single 2325 hard attack or two 1056 hard attacks?

[DeltaV112]

ORIGINAL: Emx77

Guys, thank you for explanation. It make sense. However, that opens new questions. For example: Which of these two fighter bombers is having more chance to destroy 300mm Polymer plating monitor tank? In other words is it better to attack such tank with single 2325 hard attack or two 1056 hard attacks?

It depends on the hitpoints of whatever you're attacking. The single attack has slightly higher value, so it is slightly better if the defender's HP after bonuses is greater than it, but quickly becomes worse than two attacks if the defender's HP is less than it. Given typical HP values this means the single attack is probably better against anything with at least 150 or maybe 100 polymer.

[Zanotirn]

As far as I understand there are two factors at play here, caliber/armor thickness and firepower/hitpoints calculations.

Splitting of the attacks should not affect the caliber calculations, so this one is irrelevant

During the attack resolution the game rolls a number from 0 to attack power for attacker and from 0 to hitpoints for defender and compares the two to determine if a hit was made.
The monitor in question can have different number of hitpoints depending on linear tech, design roll and various bonuses, but let's say it has 4K hard hitpoints.

If my calculations are correct at 1056 hard attack you have on average 13.2% chance to score a hit. Two attacks give you double that (yes, there are more complicated chances to hit both attacks or neither, but in the long run it averages out). A single 2325 hard attack should give you 29.1% chances of scoring a hit, which sounds better.
Also I doubt that's all there is to it since there's also the hit effect resolution (e.g. whether it's a kill, rout, pin or no effect). It's possible that having a higher attack roll increases chances of that the hit becomes a kill, but I have no idea how that calculation works.

Actually the numbers in screenshot look strange since the new model has less than a half of the attack of the old one. I.e. even if you just multiply the firepower by number of attacks the new model looks worse. (if it had just half of the attack of the old one, attacking the poor monitor would give you on average the same number of hits).

So either it's a bug or the game takes into account some factor favoring many small attacks that I'm missing here. (many small attacks are better in preventing breakthroughs but it shouldn't be relevant for an aircraft)

[zgrssd]

During the attack resolution the game rolls a number from 0 to attack power for attacker and from 0 to hitpoints for defender and compares the two to determine if a hit was made.
The monitor in question can have different number of hitpoints depending on linear tech, design roll and various bonuses, but let's say it has 4K hard hitpoints.

If my calculations are correct at 1056 hard attack you have on average 13.2% chance to score a hit. Two attacks give you double that (yes, there are more complicated chances to hit both attacks or neither, but in the long run it averages out). A single 2325 hard attack should give you 29.1% chances of scoring a hit, which sounds better.
Also I doubt that's all there is to it since there's also the hit effect resolution (e.g. whether it's a kill, rout, pin or no effect). It's possible that having a higher attack roll increases chances of that the hit becomes a kill, but I have no idea how that calculation works.

The math as I explain them:
1056 Attack vs 4000 HP

There are 4000 possible results. Of those, [4000 - 1056]/4000 % is a guaranteed defender win, as it rolled higher then the attacker can roll.
The remainder is split 50/50 for either side.

[4000 - 1056]/4000 = 2944 / 4000 = 73.6% automatic defender win.

That leaves 100-73.6 = 26.4% that are split 50/50 between defender and attacker win.

So I end up with 13.2% attacker win chance
86.8% Defender Win chance

[GuardsmanGary]

ORIGINAL: Zanotirn

Actually the numbers in screenshot look strange since the new model has less than a half of the attack of the old one. I.e. even if you just multiply the firepower by number of attacks the new model looks worse. (if it had just half of the attack of the old one, attacking the poor monitor would give you on average the same number of hits).

So either it's a bug or the game takes into account some factor favoring many small attacks that I'm missing here. (many small attacks are better in preventing breakthroughs but it shouldn't be relevant for an aircraft)

Looking at my own designs I've managed to track down an increasing penalty for additional attacks. This penalty is added after the multiple attack divider is applied. There is a -25% penalty to the attack/defense values for having 2 attacks, a -33% penalty for having 3 attacks, a ~-42% penalty for 4 attacks, and a ~-58% penalty for 6 attacks. Rounding starts becoming an issue when you get to the larger numbers involved with 4 or 6 attacks giving these awkward percentages.

If we look at Emx77's Shining Star III we can go through all the steps to arrive at the final soft attack/defense values taking this hidden penalty into account.
It starts at 1810 firepower, 60 for the machine guns and 1750 for the bomb load.
132 weapon design increases this by 32%: 1810 * 1.32 = 2389
Precision bombs have an 80% soft attack penalty: 2389 * 0.2 = 478
Though it's not shown I can very safely guess there is a cluster bombs bonus of 50% being applied: 478 * 1.5 = 717
Two attacks divides this value in half: 717/2 = 359
And finally a hidden -25% penalty for having two attacks: 359 * 0.75 = 269

Maybe when Vic was designing the aircraft system he decided to slip in an additional penalty for multiple attacks that was not thought of when ground vehicles were developed? This penalty only appears on aircraft designs.

Similarly there is an issue in the design log for aircraft that makes it inconsistent with design logs for ground vehicles. When looking at the design log for conventional artillery it makes it very clear the additional attacks it gets are from firepower: it lists "base attacks" in one entry and then has a second entry for "attacks increase due to high Fire power". For rocket artillery and missile launchers it's simply a factor of size of the rockets/missiles, there is only "base attacks" and no second entry for "attacks increase due to high firepower". A rocket launcher with medium rockets will only ever have 6 attacks even if its firepower is greater than that of a rocket artillery equipped with large rockets and subsequently 8 attacks. The aircraft design log uses the same language as rocket artillery and missile launchers but they get increased attacks based on a firepower as well; a superheavy bomber with a 160k bomb load getting it 7000 firepower has a "base attacks" entry of 4 attacks, while another superheavy bomber with the same 160k bomb load has a "base attacks" entry of 6 attacks once you introduce missiles onto the aircraft, which allow it to benefit from payload optimization, increasing firepower from 7000 to 9900. For both aircraft and rocket artillery/missile launchers these number of attacks in the design log are noted as "base attacks" despite one clearly getting additional attacks based on firepower.

[zgrssd]

If we look at Emx77's Shining Star III we can go through all the steps to arrive at the final soft attack/defense values taking this hidden penalty into account.
It starts at 1810 firepower, 60 for the machine guns and 1750 for the bomb load.
132 weapon design increases this by 32%: 1810 * 1.32 = 2389
Precision bombs have an 80% soft attack penalty: 2389 * 0.2 = 478
Though it's not shown I can very safely guess there is a cluster bombs bonus of 50% being applied: 478 * 1.5 = 717
Two attacks divides this value in half: 717/2 = 359
And finally a hidden -25% penalty for having two attacks: 359 * 0.75 = 269

Nice sumary, never looked that deeply into it myself.

However I would bet that the splitting into multiple attacks happens after the design roll - so on 2389.
Otherwise we would get into a wierd situations like the unit having 1 Soft but 2 Hard Attacks.

So it would be two attacks at 2389 / 2 * 75% = 895.875.
Wich are then modified for soft or hard, offense or defense combat.

[Tchey]

Hi,

So do you all agree, the Shining 3 is actually worst than the Shining 2 ?

[Emx77]

ORIGINAL: Tchey

Hi,

So do you all agree, the Shining 3 is actually worst than the Shining 2 ?

Based on comments above, and if we put aside fuel consumption, hit points and range, it seems that SS III is worse ground attack aircraft than SS II. Which is weird having in mind same weapon loadout on both planes. [:(]

[zgrssd]

If you think the result is worse, consider reporting it as a bug.
Did wonders to get the Fuel and Worker efficiency of Logistics assets fixed. And I am hoping for a few Airplane fixes.

[Sieppo]

I don't claim to understand one bit on how the combat works below the hood but this mechanistic that instead of giving higher attack value to better weapons you give them more attacks is so very weird when to my understanding (that little) defense is based on comparing hit points to attack value. So if your attack is less than hit points = not good. So two times below those hit points with less attack value should be much worse than killing with one overwhelming blow?

[Sieppo]

And to add to that, making bigger guns just makes them create a bigger boom killing harder to kill stuff more easily, NOT making more smaller booms that those harder to kill objects just shrug off. Hard to wrap my head around this why Vic chose to do it.

[Soar_Slitherine]

We don't know how the combat kill chance mechanic works. If kill chance is flat per attack and having a specific value of firepower split into 2 attacks means the game rolls for kill chance twice while the overall chance of scoring a hit remains the same, then 2 attacks is a lot better than one. On the other hand, if overkill on the attack score increases kill chance, then having multiple attacks might not be so much of an advantage, unless the targets have much lower HP compared to the attack value so hit chance is already very high.

[Vic]

ORIGINAL: Soar_Slitherine

We don't know how the combat kill chance mechanic works. If kill chance is flat per attack and having a specific value of firepower split into 2 attacks means the game rolls for kill chance twice while the overall chance of scoring a hit remains the same, then 2 attacks is a lot better than one. On the other hand, if overkill on the attack score increases kill chance, then having multiple attacks might not be so much of an advantage, unless the targets have much lower HP compared to the attack value so hit chance is already very high.

overkill roll does not increase chance for a kill, so 2 attacks does indeed give an advantage.

attack scores might be slightly lower because the horsepower to weight ratio change.

[Soar_Slitherine]

ORIGINAL: Vic
attack scores might be slightly lower because the horsepower to weight ratio change.

Are you referring to attack versus aircraft? Some quick testing suggests that aircraft horsepower-to-weight affecting soft or hard attack is not a thing.