Critical hits were scored in approximately 50% of the shots. Shots fired into the side the hull (tracks) location failed to do any hit point (HP) damage and were excluded from all further comparisons. All other shots resulted in HP damage, ranging from 5% to 58%. Only one of these 43 shots (2% of the shots) penetrated the KV armor. This shot was placed in the rear hull (scenario 4) and caused the 58% damage result. This is surprising since the armor penetration of the 105 mm shell is only 43 and the KS's armor in the location hit is approximately 70 (perhaps a lucky dice roll or weak spot was hit). The damaged caused by this hit (320 HP) was close to the +/- 25% range of the 105 mm shell damage (410 HP). Excluding the 58% datapoint (and Scenario 1), damaged ranged from 5% to 24%, averaging 16% per shot.
With the exception of scenario 1 (side tracks), damage did not greatly differ by scenario, as shown in Figure 2.
Figure 2 shows the average damage by scenario, with error bars representing standard deviation (SD). No damage was observed in 5 shots to the tracks (scenario 1), indicating that the explosion did not reach the hull. Although critical hits on tracks were noted, this is not an effective location to target. Results from scenario 1 were excluded from statistical analysis.
The letters above each column indicate statistical significance such that columns with the same letter are not meaningfully (statisitically) different. For example, scenario 6 ( A,B label) is not different than scenario 9 (A label) because both columns have an "A" label. In contrast, Scenario 3 is different from all columns whose labels do not have a "C" (scenarios 1.5, 2, 5, 6, and 9).
Among the other results shown in Figure 2, it is clear that shots to the side of the turret away from the hull (scenario 3: average damage 10% ) did less damage than shots to the other locations, including the same turret side at a location closer to the hull (scenario 2). The scenario 2 shot was placed much more closely to the hull than the shot in scenario 3, suggesting that damage rays from the explosion in scenario 2 may have penetrated the top of the KV hull instead of only the turret (as in scenario 3). Because less explosion rays strike the hull in scenario 3, the damage is slightly less (i.e., 10% damage versus 17% damage). However, damage in scenario 2 (side of turret near hull) was not different than in scenario 1.5 (side of hull), suggesting that the top surfaces of the hull presumably struck by explosion rays in scenario 2 are not more easily damaged than the side of the hull.
As with the result observed for the side of the tank, the top surfaces of the rear hull (engine deck) were not more easily damaged by a shot placed on the lower portion of the turret (scenario 5) as opposed to shots placed directly into the rear hull (scenario 4). Damage for the direct hit rear hull shots averaged 23% versus the rear turret near the hull damage of 16%. Although, the 23% average does include the 58% damage shot that penetrated. To enable an apples-to-apples comparison of non-penetrating HE damage, this 58% result was removed from the dataset and the statistical analysis was repeated (as shown in Figure 3). In Figure 3 it is clear that shots on the rear turret near the hull do similar damage to the shots directly on the hull (15% versus 16%). Thus, the top surface of the hull on the rear of the tank (engine deck) appears to be no more easily damaged than other parts of the hull.
From the above results for the sides and rear of the tank, it may be possible to conclude that there is no advantage to targeting the hull directly (scenarios 1.5 and 4) versus trying to splash damage onto the top of the hull by hitting the lower turret (scenarios 2 and 5). Strangely enough, Figure 3 suggests the
top of the hull on the front of the KV appears to be weaker (or a critical hatch or seam area is present, allowing more damage). Damage for the lower turret on the front (scenario 9, average damage 22%) was significantly greater than damage in the lower glacis plate (front hull, scenario 8, damage 14%). This represents nearly 50% more damage in scenario 9 than the other scenarios. One would expect the opposite since the front turret armor is 90 mm and the hull armor is 75 mm.
Consideing all the above results from the front, side, and rear of the tank, aiming for the lower portion of the turret is the most optimal aiming strategy, as shots that land in this area will do the same damage (sides, rear) or more damage (front) than shots landing directly on the hull. I make no claims as to your ability to do this accurately (yes, HE is inaccurate at most battle ranges), but put the center of the recticle in this lower turret area and hope for the best.
All statistical analyses in Figures 2 and 3 indicate that one can cause damage by landing HE shells underneath the tank (scenarios 6 and 7), however, damage is no greater than shots fired directly into the front or rear hulls. Thus, the underbelly hull is not more easily damaged by HE than other hull areas: there is no advantage to landing HE shells underneath the tank.