This vast difference between 3 and 4 continues in weapons, with the disgustingly effective bolter being S4, while the lasgun, still considered an effective weapon arguably equivalent to modern military equipment, is S3.
These seem like vast differences in a 1 point scale. However, I think GW doesn't get enough credit for actually translating this fluff into a D6 based game. Our tendency to think of all game stats as a consistent, sliding scale is incorrect. Instead, it's an exponential scale of two. That is, a Space Marine, at S4, is at least twice as strong as a guardsman at S3. And consider that in the context of the fluff, guardsmen are generally speaking, effective and well trained and conditioned soldiers. S3 is not the average human strength, it is the average strength of a conditioned warrior who spends much of his life in a space ship going from one warzone to another, probably lifting weights and training.
This scale of doubling plays out in armour too. Power armour is some of the most advanced protective wargear out there. Meanwhile, Terminator armour is capable of withstanding a bewildering amount of punishment. And in game terms, it is at least twice as durable as Power Armour.
The issue with this relatable exponential scale, is the massive difference between 3 and 4 becomes largely irrelevant when faced with the apocalyptic strengths and toughnesses of weapons and Xenos races in the game. Being twice as strong and tough as a highly trained warrior doesn't mean much when getting hit by a literal Mac Truck of muscle and bone.
|Beep Beep, Motha Trucka!|
So let's just translate all this into some relatable metrics. Using a quick lookup of punching force by boxers, we see that professional boxers average around 776 pounds of force for a punch, and double that of a novice. Taking that, let's generously accept this as your average guardsmen, helped by combat stimms and various other genhancing products used to keep them at S3, equaling 775 pounds of force for a single blow.
Using our guess of the scale being exponential, we would expect an average space marine to deliver a blow with 1,550 pounds of force, while a Necron Lord or a Space Marine with an axe, could concentrate that force into a 3,100 pound impact. For comparison, that's about the extrapolated bite strength of a T-Rex. So a T-Rex would be S5 in this scenario.
If we ran all the way up the ladder to a Carnifex with adrenal glands or crushing claws, or a Wraithknight (S10), these monsters could exert a force of 99,200 pounds. For comparison, that would be about the force of getting hit by a car going 50 MPH, or being hit by a 10 pound sledgehammer swung at 545 MPH (keeping in mind it's been nearly a decade since I took a physiscs course).
Lastly, let's look at initiative, which we can compare to reaction time. Using an average reaction time of 266 miliseconds (or about a quarter of a second -0.25 seconds) for human average, that would mean an I4 Space Marine is capable of reacting in 0.125 seconds, while your average Eldar reacts in 0.0625 seconds. Going in the other direction, an Ork takes a half a second to react at I2, and anyone using a Power Fist or axe takes about a full second to react.
Where this exponential scale breaks down a little is when equipment or rules allow for something to increase or decrease by 1 (or more), effectively doubling it's strength or toughness on a linear scale.
Anyway, I find this a fascinating aspect of the game that I often overlook while playing. Through gameplay, a point difference in a statistic is usually pretty minor, but thinking about it in the fluff, it's a huge difference. I often forget the scale of the game and how the 2-3" tall miniatures I'm moving around, if real, would be about as tall as my house, and definitely would have to duck to fit in the game room.