# Fun and Engaging Science: Part Three

There are still a lot of myths surrounding detonations. Let’s delve into some practical demonstrations.

One of the most telling myths surrounding detonations is that they’re entirely random. Now, thanks to Sub_Octavian’s answers, we know how much of an influence RNG actually has on detonation mechanics (which is still substantial). I wanted to experiment with torpedoes and their large blast radius to see just how effective these munitions were at setting off a ship’s magazine. Even with the dev answers we got last time, there are still a lot of questions surrounding just how big these blasts are and how armour interacts with the damage potential of torpedoes. To this end, I put together a little experiment.

Let’s start with Yamato as our test subject. Few ships are as resilient to torpedo hits.

- She has 97,200hp.
- Her two magazines, forward and aft, have 48,600hp each.
- She has the best torpedo damage reduction in the game, provided hits strike her anti-torpedo bulges which should allow multiple strikes in the same area without sinking her with direct damage.
- Yamato (and all tier 10 ships) have only one hull variant, so there’s no differences between their levels of protection for their magazines between stock and upgraded hulls.

To attack Yamato’s magazines, I used a Minotaur’s torpedo armament.

- She can launch eight torpedoes per side.
- These torpedoes can be launched one at a time to ensure accuracy on a single point.
- These torpedoes have a maximum damage of 16,767hp*

_{* Note the listed damage in port may not (and probably does not) reflect the actual amount of damage the weapon does against modules and should be used as a guideline only. Minotaur’s torpedoes have hidden stats like 47,000 alpha damage, 1,100 damage and 1,000 alpha piercing HE. Yeah, I can’t make sense of that … yet.}

From what we’ve learned so far, we know the following:

- Individual hits on Yamato’s magazines will only have a 1% chance of setting it off so long as the magazine has taken less than 14,580 damage.
- The best possible chance we have of detonating a Yamato is 10% per hit, but only if we can fully reduce the hit points of the magazine.
- Torpedoes have a large “blast” radius which can affect magazines even behind multiple layers of protection.
- The amount of protection around a magazine will reduce the amount of damage taken per hit.

From this I was expecting to see only hits directly striking the area around the magazines to set off a detonation. By firing eight torpedoes into the same area, I hoped to deplete the magazine’s hp completely. From an earlier test, I knew that 8 torpedo hits from Minotaur was not enough to sink Yamato outright, provided the hits all landed in the same area and she would flood out if she didn’t detonate. This experiment would be help illustrate a couple of unknowns: how much damage the torpedoes were actually doing and how much would Yamato’s armour protection reduce the damage of the torpedoes?

At a minimum, I expected 8 torpedo hits in the same area to give me an 8% chance of detonating a Yamato (8 strikes * 0.01 detonation chance = 8% overall detonation rate). However, if the maximum damage listed in port was being dealt to the Yamato, we could expect to see as high a detonation rate as 67.6% (1.6% with first hit, 6% with second, 10% for each subsequent hit). This is, of course, assuming that after a magazine is 30% depleted, the progression is linear.

**Test #1 – Mike Control**

Lert was kind enough to assist me with this section of the test. First, we needed to establish that not just any ol’ hit would set off Yamato’s magazines. To this end, we aimed our torpedo hits to strike Yamato between her bridge and funnel, right into her center of mass. Each Yamato took all 8 hits in this same location. We repeated this test 100 times. Not one Yamato detonated.

From this we could conclude that hits amidships from Minotaur would not detonate Yamato.

*Boring, but necessary.*

**Test #2 The Nose Knows**

The next thing to test was how big the blast radius actually was. There are anecdotes a plenty from players claiming they took a torpedo hit on the bow and then detonated because of it. So once more Lert and I took out our Minotaur and derped fish into one location over and over and over again until we had sunk 100 Yamato. Not one of them detonated.

From this we could conclude that hits on the forward section of the bow from Minotaur would not detonate Yamato.

*Bow aiming point.*

**Test #3 – Exhaust Port Two Meters Wide**

The second part involved aiming directly for Yamato’s magazines. There’s a handy little ‘square’ of armour along the waterline beneath her #1 turret that made a good aiming point that’s visible even at a distance. Torpedoes were shoved repeatedly into that little gap (lewd) and the results recorded. Out of 100 Yamato sunk, 13 detonated. Here’s how many hits each one took before exploding:

- 4 hits
- 7 hits
- 7hits
- 1 hit
- 7 hits
- 4 hits
- 7 hits
- 2 hits
- 5 hits
- 1 hit
- 1 hit
- 6 hits
- 6 hits

Out of 800 torpedoes launched, 754 were needed to sink 100 Yamato. This works out to an average of 58 torpedo hits to cause one detonation. From this we could conclude that hits here would detonate Yamato.

*Just a little more to the left…*

**Test #4 – RNG is RNG**

This test was very important. We established that derping fish into a specific spot could cook off the magazine, but what about “random” hits? To simulate a strike more akin to what might be found in Random Battles, Lert and I conducted this test. From a range of approximately 5km, we set the Minotaur’s torpedoes to launch as a salvo instead of individually, ensuring that all 4 torpedoes per launcher would strike along the length of the Yamato. We weren’t concerned with spacing them exactly. The idea was simply to put 8 fish ‘randomly’ into the side of the Yamato and observe how often the magazine cooked off. At best, I estimated that two to three torpedoes would strike along the magazines, both fore and aft. After launching 800 torpedoes at 100 Yamato, 4 detonated. Here’s how many each victim took before exploding, though few if any of these took more than 2 to 4 hits to the same magazine:

- 3 hits
- 3 hits
- 2 hits
- 5 hits

Out of 800 torpedoes launched, 773 were needed to sink 100 Yamato. This works out to an average of 193 torpedo hits to cause one detonation. From this we could conclude that random hits against the Yamato could cause a detonation, but at a significantly lower chance than if you aimed shots specifically at the magazines.

*Spreadin’ the love. That looks like three** hits to the magazines on that salvo. Two forward, one aft.*

**Premature Conclusions**

Let’s be clear of one thing right from the word go: Sinking 100 Yamato is not enough to give a clear indication of the odds. A fluke detonation or two (or the absence thereof) throws the math way off. I would be more confident drawing *any* conclusions with 500 or more sunk* *on a given test. However, let’s do some preliminary work.

First, we have to make a pretty bold assumption: That I was hitting the magazine with every torpedo. There’s no feedback provided to let you know when you successfully struck a magazine or not. The closest we have for AP shells is a “citadel hit” ribbon, but that doesn’t differentiate between hitting the citadel itself or the magazine within the citadel (*if* the magazine is within the citadel). The waters get even muddier when you start using attacks with an explosive radius. I simply do not know if all 800 torpedoes I fired in test #3 hit the Yamato’s forward magazine or not. I’m dealing with a lot of unknowns:

- Just how big is the blast effect of Minotaur’s torpedoes?
- How much damage can they do?
- How exactly does Yamato’s armour mitigate the damage?

A hit a little more to the left or right may have made all of the difference if a torpedo hit struck the magazine. Assuming that all 800 torpedoes were actually hitting the magazines, we can extrapolate approximately how much damage each of Minotaur’s torpedoes was doing. The 13% detonation chance achieved is 5% higher than the 8% minimum we expected, but a far cry from the 67.6% chance had we been depleting the magazine on our 3rd hit. If we assume our torpedoes are doing a mere 3,000 damage per hit, we get the following odds per strike:

- 45,800 / 48,600 – 1%
- 42,800 / 48,600 – 1%
- 39,800 / 48,600 – 1%
- 36,800 / 48,600 – 1%
- 33,800 / 48,600 – 1.1%
- 30,800 / 48,600 – 1.9%
- 27,800 / 48,600 – 2.6%
- 24,800 / 48,600 – 3.4%

This looks fairly accurate based upon what was seen so far. But again, the small sample size can skew this result wildly. To better isolate it, we’d need more repetitions of test #3.

What this test has shown is that torpedoes are dangerous — something we all knew. But hits have to be taken at the magazines to cause a detonation. Hits along Yamato’s belt armour amidships will keep her safe. Hits to the extremities of her bow and stern are also safe. Of course, this only applies to Minotaur’s torpedoes against a Yamato. There are no guarantees that a Shimakaze torpedo won’t have a larger blast radius to endanger your battleship, or that your Großer Kurfurst has the same levels of protection.

**More and More Questions**

As an overall experiment, this one was a rather poor use of my (and Lert’s) time all told. There were too many unknowns to draw any form of reasonable conclusion. We still don’t know how big torpedo explosions are. It would be handy to be able to take a cross-section of a ship and be able to map where a given shell needs to strike to be able to affect certain modules. Similarly, we don’t know how much initial damage these explosions come with. If we had these two elements, we could begin to isolate where would be the best parts to aim at a ship to guarantee the maximum damage to magazines and thus skew detonation results as much in our favour as possible.

The next step I have planned is to try and isolate blast radii of HE shells. This burst radius **is** a stat that has been datamined (Minotaur’s is 1.2 for her torpedoes, for example), but we don’t know how it translates to in-game measurements. Once we know that, we can begin to isolate how much damage is being done per attack and from there, we can

*really*start to figure out which parts of a ship are vulnerable.