OK, if you want to keep doing this, lets keep doing this. I apologize to everyone who doesnt really care about the fine details of warhammer statistics that we seem to clutter up this particular thread.
First it is obvious that a basic lesson in probability calculations is in order. All probability calculations assume an infinite number of samples, and therefore the average wounds caused (in this case) is what is calculated. The fact that piranha blade can potentially cause 10 wounds in a single turn is therefore not interesting. that issue is often refered to as "burst potential" and it plays a very large role in for example PvP in world of warcraft. overly simplified it can be said that there it is better to have a small chance to cause a large amount of damage at once, than to have a large chance of causing a small ammount of damage. this is because the first case is much harder to react to for a player who is healing, and thus can result in a dead player. this is not applicable in warhammer, since there is so very little healing, and there is no real "reaction time" since it is turn based, so we can drop that argument at once.
Now on to the example with a T4, 1+ AS and 4+ ward character.
Mr GW hits on 4+, wounds on 2+ and the save is 5+/4+. that means mr GW has a 14,81% chance of actually causing a wound with each attack. (note that i took poison in to account, though it is a very small contribution here. without poison it is 13,89%)
Mr P(as in piranha blade) hits on 4+, wounds on 3+ and the save is 3+/4+. that means mr P has a 5,56% chance of breaking through and causing TWO wounds. According to all normal probability analysis this is analog to having a 11,11% chance of causing ONE wound.
As you can clearly see, since they both have 5 attacks you can see that since 14,81>11,11 mr GW will come out on top in this comparison.
Now, there is actually an effect that is not included in this type of analysis. that is the potential "overkill" of mr P. If you are fighting an enemy with an even number of wounds, this does not show up (assuming mr P is the only one attacking), but if the enemies number of wounds is odd, then it actually becomes significant. I did take this in to account when it came to the obvious case when fighting models with only one wound. In this case mr P does not get any benefit from his sword, since even if his sword causes 2 wounds, the enemy only has one wound to lose. If you want to be nit-picky though, the same phenomenon exists when fighting models with 3, 5, 7 wounds and so on. The effect grows smaller when the number of wounds increase, but against a normal lord, who has 3 wounds, mr P can either cause 2 or 4 wounds, meaning that the 4th wound is "wasted". you would still get overkill for it in a challenge, but it should not be included in an analysis of how good mrP and mr GW are at killing stuff.
This means, that in a fight against a guy with 3 wounds, 1 wound in every 4 that mr P causes would be "overkill" and hence should be removed from the comparison. mr GW will never cause any accidental overkill in that way, but understanding the difference between causing excess wounds with normal attacks (and they can both do that) and causing it with the doubling of the wounds made by mr P IS a quite subtle point, and many people will have trouble understanding it in the case with 3 wounds, even though they think it is very obvious in the case with only 1 wound. If you are not truly interested in these fine points, just take my word for it that this effect does exist and that it can be significant.
All in all that means that the probability to wound for mr P should be reduced by 25% to 8,33% to take the potential overkill into account. one wound in every four that he causes (against a 3 wounded character) will always be "wasted" since he can only cause 0, 2, 4, 6, 8 or 10 and never 3.
As you can see here, the problem with overkill does exist even for mr GW, since he could potentially cause 4 or even 5 wounds in a single turn, but the loss he gets from this is much much smaller, since it is to be compared with mr P scoring 8 or 10 wounds in a single round. It can happen, but the reason that i didnt count with teh possibility of 6+ wounds on mr P and 4+ wounds on mr GW is that it is a whole magnitude less likely to happen, and can thus be ignored. as an engineer it is my job to simplify where possible
In closing i would like to press the issue that i do not pretend that my strategical thinking is beyond reproach, but that if you question my maths you'd better know what your are talking about. the last post contained counter arguments that are clearly not valid mathematically. strategically though mr P has one edge; that he can potentially kill an enemy character (or monster or whatever) before said character can ever strike. mr GW does not have that luxury. whoever, as the calculations show, mr GW is more likely (in the cases i calculated) to EVENTUALLY beat said character/monster. My tactical choice here is to go with the "safe bet" on mr GW rather than the "wildcard" of mr P. If any one choses diferently that is fine by me, as long as they do not claim that it is a choice build on mathematical analysis.
)
