Are missile fire casualties too random?

[dazzam]

I have been using a few missile heavy armies recently and been paying a bit of attention to the casualties relative to hits. There seem to be a lot of strange results. I've had a few 2 hits producing 2 casualties and a few 0 hits producing 3 casualties. One would have thought that the number of hits should be the main determinant of casualties but it seems to me it is a little too random. I understand from the rules that there are overlapping % casualties for numbers of hits (although I don understand why) but what I don't know is how that range is determined ie for 0 hits the range is 0.01% to 1% for 2 hits its 0.5% to 5% which explains why the above results are possible. However how is that range determined ie is it 0.5% or 5%? Is it normally distributed or not? My guess is that the % casualties determined is not a normal distribution..but that's my guess. In general I find missile fire to be a bit of a lottery.

Anyway someone asked me the other day about my pet gripes with the rules and shooting casualties random nature would be up there. I also saw some comments about troop valued and there were some who thought bows are not worth the points. I tend to agree with that but due to the fact that the results are too random. Does anyone else agree or care?

As my last poll was so much fun I thought I would do another on this subject.

[Blathergut]

This connects to the age old concern with casualties in both shooting and combat and how it varies so extremely....1% to 18%. Much has been said but there's never been any comment on it all from the developers as far as this ancient brain can remember. Or comment on exactly how the game arrives at the results.

[dazzam]

Yes I've seen plenty of commentary on the random results before. I can live with the impact and melee combat as over time I think the results tend to be fair. Very occasionally u get some LF slingers beat up some Elite legion and it's great but extremely rare. But archery fire..too me is just too unreliable. I can see as a game why you might want to tone down their effectiveness but I would expect the results over time to have some sort of normal distribution rather than what seems more evenly distributed over the range. These games should always be a combo of skill and luck but I feel the missile combat systems strays too far toward the later.

[stockwellpete]

My answer is "yes". I would like to see the range of results moderated in the same way that I would like to see melee outcomes altered - these would be steps towards making the game more realistic, in my view.

[deeter]

Ditto! and for all combat results. I know several people who were initially excited about the game that quit playing because they were turned off by the wild swings in casualties.

Deeter

[mceochaidh]

I think the wild variability of all combat results is a serious problem; I propsed a change using a bell curve system to reduce the frequency of all extreme combat results, but not eliminate the possiblity of having these results. I sponsored a poll on the subject in the fall and the responses were about 70% in favor of moderating the extreme results, with a reasonably large number of responses. The bell curve approach should require very little programming, just changing tables.

Among other things, I think the present system may encourage the use of more horde type armies, as lower priced BGs can cause outsized casualties on the more expensive troop types, making them easier to disorder and fragment. This is compounded by shooting casualties received before impact.

As far as I am aware, the design team has not commented on this issue. It would be great to hear from them on their reasons for creating the large variation in results and the frequency with which the extreme results occur.

[dazzam]

Thanks for your comments mac..I just had a look at your poll and agree with the view that a normally distributed rang of casualties would be better. My own view is that the combat results from hand combat are not as bad as missile shots. In combat with the POA system dictating the percentage chance of hits to me gives a fairish set of probabilities of the number of hits and with the superior number of hits being a big determinant in the number of casualties. For example a 2 hits v 1 hit will give 5-14% v 0.25 to 3%. At least there is no overlap of casualties. I feel that over time you get the balance of combat results is acceptable but maybe it could better with normally distributed casualties. But that's just my opinion and I see that there are those who feel stronger about this.

However to me the area they have not got the right mix is in shooting casualties. As an example in a recent game I had a 4 BG of MF Longbow fire into some protected MF for a total of 8 casualties which the enemy survived the cohesion test. A few hexs along some LH with bows fire into a MF protected inflicting 2 casualties and they get disrupted. While I appreciate that a sample space of 2 is not statistically significant it's anecdotal evidence of relative events that happen far too often. I realize I'm touching on a different issue here with regard to the results of cohesion tests from missile fire but I think that part of the problem. However my first step would be to rectify the results from casualties. The overlaps of casualties based on numbers of hits are too large in my experience 0.01 to 1% for zero hits and 0.5 to 5% for 2 hits just seems wrong in the absence of a normal (bell curve) distribution. Actually statistically speaking you may want to play around with the kurtosis to give it slightly fatter tails but at least make it more normally distributed than it is.

The frequency that a unit can be disrupted from 2 casualties out of 300 while another unit can lose 30 out of 300 and be ok also seems wrong. Massed arrow fire just does not have the impact you would expect.