Although I have the rules, I have admittedly not played a game of Warhammer Fantasy Battle 8th Edition yet. However, while following online discussions of the game and perusing the new system, I noticed an interesting trend. Units had, across the board, become much more lethal in this edition, with myriad special rules to allow more and more attack rolls.
To be sure, Warhammer has always been known as a "bucket of dice" game, although I have never seen this as a particular fault. The average attack in 3rd Edition, for instance, must roll to hit (an average of 2 in 6) and then to wound (3 in 6) before the opponent can make a saving throw for armour. It might be seen as much quicker to simply allow the attacker a 1 in 6 odds of mortally striking his opponent (which has the same odds as the two previous rolls), yet the rather simple and elegant mechanic of requiring multiple layers of dice rolling allows the game designer to subtly tweak the odds. For instance, a 3 in 6 chance to hit and to wound (or 25% chance to mortally strike) cannot possibly be represented on a single six-sided die roll.
But the recent disturbing trend seemed to force these elegant mechanics to their extremes. By significantly boosting the number of attack dice being rolled, the difference between subtle modifiers to one layer of rolls or another became blurred. The 8th Edition in particular has a number of rules to this effect, whether allowing two ranks of models to strike in melee, or a third with 30+ models in the unit (a "horde"), or an additional further rank with spears or double attacks from additional hand weapons. It soon became quite possible to levy 40 attack dice in one round of melee. Similarly with missiles, archers could now fire in two ranks or half of every subsequent rank for volley fire. An archer regiment arranged in four ranks and ten files could roll 30 attack dice in this manner.
With a sneaking suspicion, I then looked back at previous editions of Warhammer and found that this is not at all a new development. In fact, units have been creeping in lethality since the very first version of the game. I decided to graph the number of unsaved wounds a unit might inflict if arrayed in three ranks and six files (a number that simply made the math much easier). I gave the units a 5+ save and either two-handed weapons, additional hand weapons or spears and crunched the numbers against an identical foe (ignoring initiative, charging bonus and so on). This is what I got:
As you can probably see, lethality has been increasing regularly across the different editions. Some pairs of editions worked very similarly, and more or less represented minor incremental updates on the previous rules (as with 2nd and 3rd, 4th and 5th and 6th and 7th). What's the take away of all this? A cynic might assume Games Workshop has been tinkering with the system over time to make larger and larger armies more necessary. Another option is that perhaps the game designers have been trying to reduce the time it takes to play an average game of Warhammer, making it more accessible for those with a busy schedule. I suspect both of these answers tell part of the story.
Importantly, this trend very much changes the style of game from older editions to newer. In my mind, Oldhammer is very much a detailed, gritty warband skirmish game. Newhammer, perhaps a mass-battle game in 28mm scale. Something that might support this are the changes starting in 4th Edition to radically increase the lethality of combat resolution, a rule change that greatly compounded the general trend in boosted attacks. Here, the leadership of the losing side in a round of combat is almost reduced to nothing, making an average leadership role nearly impossible. Losing combat by only three points, for instance, means a human regiment will flee 83% of the time, whereas there was no such modifier in 3rd Edition and the same human squad would have a better chance of sticking to the combat than fleeing (nearly four times better odds to remain). Furthermore, units that did flee were entirely wiped out, whereas in 3rd Edition they would merely suffer three or four further casualties from the rout. The result was that 3rd Edition warbands, when they did flee, could easily expect to return to the battle later. Later editions made these units much more expendable, resulting in much faster games but perhaps less character and narrative to the warriors fighting the battle.