Unit healing at day change

This is a follow up to the groundbreaking comments by Vigor555 in the FAQ thread regarding unit healing. Check that post for things not covered here, but my observations pretty much agree with what he said. Here I want to present a perhaps more intuitive, for some, graphical description of how the healing works. The plot below shows the theoretical and observed morale increase as a function of morale before day change. The expected amount of healing is (15 - 0.15 * morale_before_day_change). For example, if morale is at 75 before day change you can expect about 15-0.15*75 = 3.75 morale gain to give 75+3.75 = 78.75. So the reported morale would be around 78 or 79 after day change. For the real data in the plot I'm actually using hit points since it's a more accurate measure of morale. For most, that is about all you will need or want to know, but as Vigor555 pointed out the amount of heal doesn't usually exactly match the expected value due to some randomness and/or rounding, but it's close. Vigor555 suggested that internally the hit points might have more than the single digit after the decimal which could result in differences. However, this would not be enough to account for some of the observed differences, and there might be some randomness thrown in. This is most apparent with the point furthest to the left, near zero. This was a destroyer with 0.1 hit points (calculated morale of 0.33%)before and 4.3 hit points after. The expected morale increase would be 14.95, but the observed is only 14.33. Rounding the final values down some integer doesn't work, because it's clear that many of the observed values are not integers. Neither could I find any way to round the hit points so that the observed values were all predicted correctly. In order for the destroyer to get an expected value of 14.333 it would need to start with about 1.33 hit points, which is much higher than 0.1. There could still be some combination of missing significant digits and rounding at certain points that causes the variance, but for now I'm going to assume there is just a bit of randomness to it. As Vigor555 pointed out, sometimes units at high morale do not show any healing and the same unit will heal on another day with the same hit points. This is good evidence that there is a bit of randomness to it. In anycase, for practical purposes it's pretty simple, near zero morale you can expect around a 15 morale point boost and it will drop linearly to zero as you approach 100% morale.

[/img]

1 Replies

Because 100 % morale units have been observed losing morale while en-route across long sea voyages, as well as, while crossing enemy territory or garrisoning newly taken territory (yes, the latter could be from attempted revolts), I suspect there is some downward or upward pressure on unit morale based on the morale (or "friendliness") of the province they're in. Not too unlike the travel speed modifier of units in owner, friendly, enemy, or naval terrain.

If this "pressure" exists, I would suspect it has a relatively minor influence, in most cases, but might account for your divergent results being attributed to "some randomness and/or rounding" and, maybe, the 1.33 HP DD disparity.

Post a Reply

Please log in to post a reply.

Back to Discussions
Quick Launch