Please take your time and read the blog rules

May 14, 2013

Artillery in Real Life

By EnsignExpendable

With the upcoming artillery nerf rebalancing coming, I thought it would be interesting to explore what artillery could do realistically. I have previously looked at how well HE works in real life, and it seems to work pretty well with even small calibers. Large calibers, like 152 mm, work even better.

Ferdinand after a direct hit from a B-4 203 mm howitzer.
The thing is, these tests only cover what happens on direct impacts, which, as any artillery player will tell you, are few and far between. Even the Object 261, heralded for its accuracy among the SPGs, has an accuracy of 0.42, which is pretty miserable for a tank, especially when shooting all the way across the map.

That range is where the problem with real life artillery comes in.

Long range accuracy for the 76 mm model 1942 gun at full propellant load, and 122 mm model 1938 howitzer, at first propellant load.
This figure from a Soviet artillery textbook demonstrates how inaccurate a gun can be at long range. The 76 mm gun has similar ballistics to the T-34's F-34 gun, which does not miss a StuG-sized target at 950 meters. That's pretty accurate. However, at a range of 3 km, the horizontal dispersion is 8 meters, or 2.66 StuG widths. At 5 km, that distance increases to 20 m, 33.6 m at 8 km, and 30.4 meters at 10 km. At those high ranges, it is also easier to overshoot, especially if you are using a low velocity gun with a high arc.

Since in-game accuracy is much poorer for all guns than real-life accuracy, let's do some conversions. The model 1938 howitzer (M-30) exists in-game on the SU-85. Its accuracy at 100 meters is 0.55 (somehow, the model 1930 howitzer on the SU-26 is slightly more accurate), meaning that shells at 100 meters would fall within a circle that has a radius of 55 cm. At 3000 meters, this circle would be 16.5 meters in radius, or 33 meters in diameter. Compared to the real life circle, 12 meters in diameter, 122 mm artillery is roughly three times less accurate than in real life. . Compared to how badly the D-25T got shafted (5.75 times), the equivalent caliber artillery gun gets off easy.

All right, so indirect fire in real life is inaccurate too. But everyone knows that arty relies on splash! Let's take a look at how well splash works in real life.

Craters from high explosive shells, with the fuse set to explode in the ground.
As you can see, the 152 mm shell digs out quite a pit with its explosive power. A 420 mm shell (not shown due to scale issues) displaces 250 cubic meters upon impact. Thankfully, there are no artillery pieces with such a massive caliber in this game. Even if a tank is not destroyed by the explosion, it will have a hard time getting out of the pit.

This figure only illustrates how much destruction is caused by the explosive force of the shell. However, artillery shells damage tanks not only with expanding gases, but also with shrapnel. The spread from shrapnel is not in a circle. It flies out more to the sides than front or back. The measurement of a shell's shrapnel effect is a rectangle within which 50% of the targets are struck by shrapnel. The following figure illustrates these areas for some Soviet guns.

Areas of damage for a 152 mm shell, 122 mm shell, 107 mm shell, and 76 mm shell. The direction of shot is from left to right. Shells are set to explode in the air.
Of course, these rounds are designed to destroy infantry and soft vehicles, and we are dealing with tanks here. However, shrapnel can still destroy tracks, optics, and wheels. In game, the damage from an artillery shell should be some mix between damage from the explosion and damage from shrapnel.

The in-splash on the SU-26 122 mm model 1930 howitzer is 2.49 meters. The splash on a stock S-51 152 mm gun is 3.66 meters. Comparing this to the crater sizes, the shrapnel is considered as barely contributing to the destructive effect at all.

And now for the biggest difference: the "satellite view". Obviously, artillery cannot aim like that! However, there are ways to aim at things you cannot see, although you need people correcting your fire.

Before the start of a battle, the battlefield is carefully studied by the battery commander. Forward observer points are established, distances to various landmarks (a house, a lone tree, a fork in the road) are recorded. When a unit radios in for artillery support, the artillery battery aims in at a visible target at the same distance, and then changes the angle to aim at the invisible target.

Aiming at a target you cannot see.
This is where some math comes in. The forward observer and artillery battery will see the angle of the explosion differently. Using the difference in angles, and some quick trigonometry, it is possible to dial in. Since it would be decades before cheap calculators that could handle trigonometry, you would retrieve your data from a table of pre-calculated values.

Calibrating your shots on a target you cannot see. The battery commander must work with an observer to accurately gauge the distance to the visible target, then adjust the angle to hit the invisible target.
This is quite a bit more complicated than "point-and-click", of course. It takes some time, and some shells, to be able to fire accurately at a target you cannot see. The aiming does not need to be as precise as in-game. Instead of only one gun, you are firing with an entire battery.

In-game, your interface is an abstraction of the data that a well-trained crew can provide you. Assuming that the people spotting for you are trained artillery observers, and your map is in good order, an artillery crew should be able to dial in reasonably quickly, especially on maps as small as in-game. Hitting a moving target, however, would still be very difficult. A huge bloom when rotating your tracks is very much realistic.


- according to EU "developer" Quasar, 8.7 might bring a tier 8 medium premium tank
- the CW event is taking shape: it is confirmed that lowtier tanks will have a place within it
- SerB states that while vehicles are sorted in the team list by MM weight, in case of equal weight, they are sorted alphabetically, but some mods mess this sorting up
- it's unclear whether the random/company player stats split will happen after all
- AMX-13/105 won't appear in WoT
- the roles of Chinese tank crews (where sometimes a commander acts as a radioman and sometimes it's the loader) won't be unified
- in 0.8.6, champion companies will have arty up to tier 8, the medium tier company will have arties up to tier 6 and junior companies up to tier 4
- tier point limits for companies won't change in 0.8.6
- compensations for players who specialized in one type of arty (for companies etc.), which gets changed are not planned
- press accounts are not transferrable (eg. they are not allowed to be played by other people than the journalists)
- the "ghost" houses on A3 in Live Oaks are apparently bugs
- it's possible maps in various times of the day (dusk, dawn, night, day) are possible, won't be implemented for now
- WoT Blitz doesn't have arty, because its maps are too small
- WoT now apparently has an anti-botting mechanism implemented, nothing will be disclosed about it
- modern tanks will never be implemented, SerB stated that for modern tanks (from the 1960's till now), a whole different game would have to be made
- apparently there was some cybersports account named Cyber_Sport_Federation, punished with 7 days for fixing battles (SS: I am not into cybersports, but I can find out more if anyone is interested)
- tanks in platoons don't have bigger MM weight than without them
- players with high winrates are NOT being put by the MM mechanism to the bottom of the team more often
- Q: "What will happen to VK2801 in 0.8.6 in connection with the HEAT nerf? It's not fun to drive a scout without a gun..." A: "Don't drive a scout without a gun then." (SS: VK2801 in 0.8.6 gets a 105mm HEAT nerf, but also a depression nerf from -10 to -8)

- the Leopard 1 ammorack positions are historical
- archievements for travelling certain distance with a tank are apparently not planned
- it's not certain that captured tanks will appear only after historical battles
- apparently the increased E-25 L/70 penetration is just arbitrary, not historical

0.8.6 VK2801 and A-Panther changes

Source: gamemodels3d

You might be "happy" to know that the VK2801 gets nerfed. Specifically: its depression gets nerfed by 2 degrees from -10 to -8 and the 105mm HEAT shells get the same nerf penetration nerf as the Panzer IV - from 150 to 104. Furthermore, the 105mm shell price is changed also: the 105mm derp shells (for all vehicles) get twice as expensive (from 60 to 120), with the HEAT shell price moving from 7g to 10g.

Aufklärungspanzer Panther gets buffed tho:

It's MM weight gets reduced (from 32,4 to 27), it gets 50 HP more (from 950 to 1000), but the price will increase by 40k. Tracks will get a slight HP buff and the Konish will get an aimtime buff (from 2s to 1,7s). Furthermore, the Konish silver ammo price will be halved.

Oh yea, and Leopard 1 (apart from the 20 pen nerf for HEAT shells) gets better MM weight too: from 120 to 100 (the MM weight for E-50M stays at 120 tho).

0.8.6 GER supertest arty stats

Order of pictures: Stock, Elite, Elite gun

GW Mk.VIe (tier 2)

Sturmpanzer I Bison (tier 3)

Wespe (tier 3)

Pz Sfl.IVb (tier 4)

Sturmpanzer II (tier 4)

Grille (tier 5)

Hummel (tier 6)

GW Panther (tier 7)

GW Tiger P (tier 8)

GW Tiger (tier 9)

GW E (tier 10)

0.8.6 US supertest arty stats

Order of pictures: stock, elite, elite gun

T57 (tier 2)

M7 Priest (tier 3)

M37 (tier 4)

M41 (tier 5)

M44 (tier 6)

M12 (tier 7)

M40/43 (tier 8)

M53/55 (tier 9)

T92 (tier 10)