钢铁雄心4（Hearts of Iron IV）空战研究BUG机制

钢铁雄心4（Hearts of Iron IV）空战研究BUG机制

我选面条，开战之后一会就被盟军把飞机打光了，虽然空战略有优势，但是架不住对面飞机多啊。

所以就研究了一下空战机制，希望能找到有效的对抗手段，这一研究，吓了一跳！

按我们常规的理解，轻战因为机动能力对重战是有优势的，历史上的重战是为了远程护航和对抗重型轰炸机而产生的。但，在游戏中是这样么？

This section exclusively develops on the topic of air combat, aircraft versus aircraft. Using information available in the defines.lua file, it's possible to speculate that air combat simulation approximately follows these steps:

Detection: If enemy airplanes are not detected in the region, interceptors (defending fighters on interception mission) will not take off.

Target acquisition: This phase will determine what's the maximum possible damage that air wings can do and also which type of aircraft will receive the damage depending if it's an air superiority or intercept mission.

Combat simulation: This phase will determine how much damage is done or received in each round of combat.

Rest: The air wing rests for 3 hours after combat, then it can be sent in mission again.

Detection[edit]

The following variables, or defines, are known in order to resolve the detection phase:

Base detection chance from aircraft: 5%. Requires 2,000 aircraft on interception mission to be achieved, interpolated otherwise. This value can be increased through doctrines that provide Interception detection or Fighter detection.

Detection chance from radars: 25%. Radars have a limited range of 20 to 220 pixels from the center of the State they are built in.The range is interpolated based on the building level. It's not clear at this point if the bonus is interpolated from the percentage of the air combat area covered, resulting in a 25% bonus only when 100% of the area is covered, or if multiple radars can stack their detection bonus.

Detection chance during the night: -20%. This value means that interceptors can't detect aircraft at night without a number of air doctrines and/or good radar coverage.

Detection chance from occupied provinces: 10%. Occupying provinces in enemy territory increases the detection chance. It's not clear if this bonus is 10% per province, unlikely, or is interpolated out of the total number of provinces occupied in various States. It's also unclear if this bonus applies to detection in all air areas, including all home areas, or only in the area where provinces are occupied.

Detection chance without search mission: -50%. This define suggests that airplanes with sufficiently high detection chance bonuses would be able to detect enemy aircraft without being assigned to "Interception" mission.

Once the detection chance is established, then it gets modified by efficiency modifiers which serve to determine how many aircraft are detected.

Base detection efficiency: 10%. Once something is detected, the base detection efficiency says that 10% of all enemy aircraft are "found" or located.

Radar detection efficiency: 70%. Radar coverage will allow an extra 70% of all enemy aircraft to be "found" or located.

Random factor: 10%. The defines add a 10% random factor to detection efficiency. It's not clear if the factor can have a negative value or if it's binary (0% or 10%).

Target acquisition[edit]

Once aircraft are detected and the number of detected aircraft is determined, it's necessary to find out who will receive the damage resulting from the combat simulation. In the case of an air superiority fight, the answer is straightforward: the enemy fighters.

Base number of attack passes: 10%. This value multiplied by the attack value of an air wing will give the base number of passes each aircraft can do in single combat. An "attack pass" gives a chance to do damage against an aircraft. Note that, just like army divisions, an air wing with different models of aircraft will average out their value for the entire air wing so that any one fighter will be indistinguishable from any other fighter in the air wing. But the combat simulation is done at the single aircraft level.

In the case of interceptors, they possibly have multiple kinds of targets: escort fighters or the bombers themselves. The interceptors want to intercept or shoot down enemy bombers as much as possible and a few variables will increase or decrease their chances to roll attack damage on the bombers instead of escort fighters.

Base chance to attack bombers directly: 25%. Interceptors have a base 25% chance of "bypassing" the escort fighters and to attempt to inflict damage directly to bombers. (in the defines, this is called the "Combat escort pass chance")

Multiplier to chance to attack bombers directly: 25%. When there's a combat statistics difference (for speed or agility) between the interceptors and the escort fighters, the chance to "bypass" the escort fighters is modified. 25% of the statistical difference is used to add or substract to the base chance to attack bombers directly.

Maximum chance to attack bombers directly: 95%. Regardless of how better the interceptors are, they will never have more than 95% chance to "bypass" the escort fighters.

Now, the number of aircraft in the simulation will almost never be the same. The numerical advantage on one side provides a chance to result in more 2vs1 single combat simulations, instead of the default 1vs1.

Gang chance from combat aircraft advantage: 40%. It appears this value applies on the difference in number of aircraft and the result gives the number of 2vs1 combat situations.

Combat stack limit: 2. This value determines how big a gang situation can occur: 2vs1. This also implies there's no advantage in having more than a 2:1 air superiority for air combat purposes.

Also of note, the defines have a modifier to the number of passes that can be done against bombers or against air superiority missions (in single, 1 versus 1, fight). At the time of writing this, the modifier is set to "1", meaning the number of passes is not affected, but this could change in the future.

Combat resolution[edit]

Chance for pass without event: 60%. For each "attack pass" that an aircraft does, there's a 60% chance it will result in an "uneventful pass", no damage done, but also no damage received.

Factor of the effect of higher speed on attack chance: +25%. This factor applies only in air superiority vs air superiority combat only. During those air combat, the speed avantage of the fastest fighter wing will give that wing a +25% chance to attack first.

For the 40% of passes that are not uneventful, there's still a chance the pass doesn't result in a successful hit for the attacker:

Base chance to hit for each pass: 60%. A hit means that the target aircraft receives 1 point of damage. The cumulative damage will need to reach the air defense value before the target is shot down.

Here's an average (incomplete) result for 1vs1 scenarios: each air wing has an attack value of 16. This means that each plane will do 1.6 attack passes in each round (on average) and 0.96 of these will be uneventful (on average). One of the two planes will attack first and 0.64 of its attack passes will give it a 60% chance to score a success (it's possible that the same define was entered twice in defines.lua, but with 2 differents strings). Out of those 0.64 eventful attack passes, 0.384 (on average) would result in a success and then the other fighter in combat would roll for it's attack value. Rinse and repeat until one of the fighters is shot down or the combat ends.

This base chance to hit for each pass is modified depending on different situations:

Maximum bonus to hit chance due to better stats: +20%. A plane with a "combat statistics advantage" can obtain up to 20% to its chance to hit for each pass it does.

Bonus to hit chance from 2vs1 situations: +18%. When a single combat simulation is done in a 2vs1 scenario, the "ganging" airplanes receive an 18% bonus to their chance to hit for each pass they do.

Maximum damage reduction from higher agility: +20%. Higher agility than the attacking aircraft will help the target reduce the damage it takes.

Also important to note, there is a maximum number of air wings allowed in a combat simulation: 6. With each air wing currently able to support up to 1,000 aircraft, this means there's a theoretical limit to the number of aircraft that can enter a combat simulation: 1,000. More importantly, a low number of aircraft in each air wing will not improve the situation in any way past 8 air wings. It's not clear at the moment if the limit of 8 air wings is for each side or the total for both sides.

Rest[edit]

Once an air wing has completed a combat mission, it needs to rest at the air base for 4 hours before it can be sent in combat again. It's as of yet unclear how long a combat simulation lasts or if the rest period applies to air wings that could not take part in the simulation due to the air wing limit.

Calculation of airplane stats difference[edit]

During the combat simulation, the difference in value of Agility and Speed statistics will impact a number of variable such as the chance to bypass escort fighters or the chance to hit an enemy fighter.

Speed difference multiplier: 25%. A speed difference between two aircraft will provide a "combat statistics advantage" equal to 25% of the speed difference.

Agility difference multiplier: 100%. An agility difference between two aircraft will provide a "combat statistics advantage" equal to 100% of the agility difference.

For example, without any variants, a 1940 fighter will have an advantage of 150km/h in speed and 15 points of agility over a 1936 fighter. This will result in a (150*0.25%)+(15*100%) = 52.5 combat statistics advantage for the 1940 fighter.

Maximum agility factor: 3. This is unclear, but it appears that for the purpose of damage reduction, the aircraft with the higher agility will not get any marginal statistical advantage if its agility is 3 times higher than the lower agility. This is effectively a cap on damage reduction.

中文简译

简译

本章特别阐述了战斗机对战斗机的空战问题。基于在defines.lua文件中找到的信息，我们推测空战模拟会遵循以下几个步骤。

侦查：如果敌方飞机没有碑额侦查到，那么拦截机将不会起飞。

目标捕获：这一阶段将决定哪些敌机会优先受到攻击基于不同的任务

战斗模拟：这一阶段每一轮中将决定伤害和受损失多少

重整：一轮战斗结束后，飞机将重整3个小时，然后再次投入战斗

侦查

以下变量或条件，是已知的影响侦查成功性的因素

基础值：5%

雷达：25%

夜间：-20%

从占领区侦查：10%

未分配侦查任务：-50%

一旦目标被侦查到，将继续进行侦查到的数量判定

基础值：10%

雷达：70%

随机因素：10%

目标捕获：

一旦侦查到的数量确定，讲决定哪些飞机是被首要攻击的目标，如果是空优作战的话，那么目标很明确：敌方战斗机

(Attack Pass)攻击机会，

基础值：10%，每次空战，战机编队的攻击值*10%为这个战机编队中每架飞机的攻击机会次数。

执行拦截任务时:25% 拦截机有25%的机会直接攻击轰炸机。

性能优势：25% 当你有性能优势时，有最大25%的机率绕过护航机直接攻击轰炸机

注意：因为所有的空战在数量上都几乎不可能完全对等，所以每次都会出现部分2v1 的加权（ HOI4的空战模拟计算都是1v1的，即无论双方多少飞机，开打的时候基本都是1v1单挑，其余围观，围观多了才有一个‘2v1’加权，但仅仅是加权参数，不会出现多对一，然后分担伤害这样的战斗模拟）

战斗过程：

攻击机会丢失：60%，即每次攻击机会有60%的可能会啥也不发生，就像对头一样，有一方觉得不好打，提前拉起，然后双机就插身而过。

优先攻击：25%对战双方中，平均速度更快的一方有25%机会优先发动攻击。。（剩下75%估计是随机或者其他因素。）

即使抓住攻击机会的40%当中，仍然有可能不命中。

基础命中率：60%。

性能优势命中率奖励最大值：20% （我的理解就是无论你机动性多高，你也就最大能提升20%的命中率了）

2v1加权命中率奖励：18%

机动性带来的伤害规避最大值：20%

Here's an average (incomplete) result for 1vs1 scenarios: each air wing has an attack value of 16. This means that each plane will do 1.6 attack passes in each round (on average) and 0.96 of these will be uneventful (on average). One of the two planes will attack first and 0.64 of its attack passes will give it a 60% chance to score a success (it's possible that the same define was entered twice in defines.lua, but with 2 differents strings). Out of those 0.64 eventful attack passes, 0.384 (on average) would result in a success and then the other fighter in combat would roll for it's attack value. Rinse and repeat until one of the fighters is shot down or the combat ends.

在以上的战斗描述中，我们看到 and then the other fighter in combat would roll for it's attack value. Rinse and repeat until one of the fighters is shot down or the combat ends. 这一句，一方攻击完了，然后另一方开始攻击，一直轮流直到有一方被击落为止。

所以，机动高的一方并非只是攻击不挨打，也不是双方依靠机动值随机获得一次攻击机会，然后再重新扔筛子随机下一次攻击该谁来攻击。而是

!!!!!!机动高的一方优先攻击，然后挨打!!!

总结，高机动优势一共有以下3个：

1、在攻守回合中优先攻击

2、最大20%的命中率加成（基础60%）

3、最大20%的伤害减免。

看上去很不错哈。。但是，让我们比较一下轻战和重战的数据

36轻战： 防御 10 攻击 18

36重战： 防御 18 攻击 35

基本是两倍以上，我们把上面的数值进行 （最大20%的命中率加成（基础60%）最大20%的伤害减免）这个加成计算。就是：

36轻战： 防御 12.5 攻击 24

36重战： 防御 18 攻击 35

看到了么？这两货完全不是一个档次的！！！！

只要重战能扛住轻战的第一轮攻击不被干掉，那么剩下啦就是完虐轻战了。

实测：意大利37年开展南斯拉夫 ，南 96架一战战斗机

意大利 轻战、重战、战术、战略 同时 5工厂开造的数量

战术

关键词：钢铁雄心4 钢铁雄心4空战研究