**Retrospective: April to July 2009**

Sadly, nothing to do with the fantastic Ozzy Osborne song but a metaphor for dissassembling rule sets and brain racking new game mechanics.

My customizing of rules began by simply extracting the data relevant to the period of WW2 I had nowWRG Rules but had separate charts drawn up which only referred to guns used in the Invasion of France and AFV armour data only listed those AFVs in use at the time. Much more Zen, I thought.

become obsessed with, after all I would never need to look up the armour class of a Tiger again! Hurrah! I still used the main text of the

Another aspect of the rules which was not very satisfactory was the probability ranges that threw up statistical anomalies, for example, Gun A penetrates Armour Class X on a roll of 4 on a D6 at ranges of 0-250m, but what if the range was 25m or 249m? Having the same chance of penetration right across the range bracket was too simple, so I began to draw up new charts based on the WRG probability data which had around ten probability brackets from point blank to extreme ranges. Admittedly,

*within each bracket*the probabilities were still flat & linear but the brackets themselves were now a lot smaller, allowing for more detail in my opinion.

Having realised that the fall-off of hit probabilities is similar for, say, high velocity AT guns and similar across a selection of howitzers I then began designing a

*graphical representation*of the 'to hit' chart which simply showed how the chance of hitting the target 'tended towards zero as the range got closer to the maximum range for that firing weapon. All you needed to know was which range bracket the target fell into for the particular firing weapon because at 'extreme range' the chance to hit is the same for a 2 pounder as it is for a 75mm L24. The only difference is that the extreme range bracket for the 2 pounder is, let's say, 1250-1500m but for the 75mm it may be 1500-1800m.

Attempting to come up with user-friendly graphics in 2D did my brain in for a few months until I realised that the problem lay in that you still had to refer to other charts or graphs when it came to armour penetration or destruction of unarmoured targets, unless you had some kind of curvaceous Henry Moore 3D probability model after factoring in the extra variable of target destruction.

The dice-rolling method for recreating probablilities of taking out a target

*did*work, it just seemed a bit pedantic to me. Roll to hit; roll to destroy. You hit the truck but you didn't destroy it. Roll dice; roll some more dice.

Was there a way of using

*how good a hit you got*to determine target destruction? I thought there was, after looking at my childrens' Cuisinaire Maths Rods.

If there was a 60% chance of hitting & a 30% chance of destroying the target, aside from neutralisation, pinning or immobilisation effects, we are really only interested in the overall 20% chance of total destruction.

So that was the major Game Mechanic I finally invented, now all that was left was to somehow convert the WRG data into my single dice-roll system...