[mpugsley]

An Essay on Beery's Combat Stress Model
by Matt

After initially deciding to not play with the mod, I've learned to be ok with it. (Note: by the end of writing this, I've actually come to like it a lot.)

As I see it:
Pros:
1) Less micromanagement is good. Of course, Beery's mod is not the only response. You could tweak fatigue values--both the values of growing fatigue and the values of regeneration.
2) Combat stress is something that would be nice to model and I think Beery's mod does a good job of that, modulo the cons I will list shortly, although I think simple fatigue is not something to completely ditch either. It'd be nice to have both (which is obviously not possible in SH3).

Cons (criticism as a model of combat stress):
Given the aim of modelling combat stress, there are three things I'd do differently. It's not clear to me whether any of my suggestions can be implemented at all, given that we are forced to tweak the game's fatigue feature (which is originally designed to model fatigue) to model combat stress instead.

So many thanks for Beery for the good idea of modelling combat stress and for giving us a good implemented model of combat stress. It may be (as far as my concerns go) the best model that we can get, given that we have to work with the original fatigue feature.

1) Combat stress should be induced largely by combat. Simply running from point A to point B shouldn't contribute (much) to combat stress. Beery's suggestion is to run at high TC as much as possible. But there are plenty of times when I'm doing hydrophone checks, manually searching the horizon, etc. where I don't want to run at high TC. And those times simply shouldn't contribute to combat stress. If they do, it should be an incredibly negligible amount.

Now I grant that the threat of combat can itself induce some stress (i.e. chasing down a merchant at low TC suggests that action is going to occur, which would induce stress, though not as much as actual combat. Seeing/hearing enemy planes or warships, even if you don't engage them, would give everyone the heebie jeebies.

But just moving around, with no contact, shouldn't be all that stressful (although as the patrol gets longer, mere length should put some strain on the crew's nerves).

So I don't like the idea that we're forced to use the fatigue feature to model stress. It makes stress depend solely on where you are and how long you are there (at low TC). It has nothing to do with whether combat is imminent, actual, how many combats the crew has seen on this patrol, etc.

2) The complete lack of regeneration. I agree that very tense situations can make the crew unable to perform their duties. (Think of Johann in "Das Boot", who cracks after a DC bombardment even though he's been on nine previous patrols.) But after the tenseness is over, relaxation should reduce stress. How you feel during a wasserbomb attack is quite different from how you feel two days after you've survived said attack. You should be much closer to 'normal' (albeit still worse for the wear of the attack). Here, think of how the crew in "Das Boot" came together when they hit rock bottom near Gibralter. Everyone (including Johann) worked their butts off (and efficiently before they got super tired) to save the boat.

3) It seems to me a plain fact that stress can actually increase your efficiency, given a reasonable amount of stress and given that it doesn't go on for too long. Stress can help you focus in important situations. We have adrenaline for a reason! It helps you put aside any distractions and accomplish what you need to accomplish in a tense situation. Your achievements with tension can be better than your achievements without tension.

I assume we've all heard the folk-tales of the people who have been able to lift a car or some other ridiculously heavy piece of equipment (well beyond their strength) in order to save a child who was trapped below the equipment. And beyond that, just think of situations you face in your own life where some stress has helped you--an important test, for example, perhaps in school, perhaps in martial arts.

So I would argue that long-term combat stress should have a bell-curve effect on the efficiency of individual crew members. As they leave port, they are a bit 'too relaxed' from their leave. As the patrol 'gets serious', everyone buckles down and focuses on his own job. As the patrol continues to be serious, the stress begins to wear on everyone's nerves. Indivudal crew members' efficiency goes down as this continues.

Thus:

e |
f |
f |---------------*--*
i |------------*--------*
c |---------*------------ *
i |-------*---------------- *
e |-----*------------------- *
n |---*---------------------- *
c |--*------------------------ *
y |----------------------------------
-----long-term combat stress




So it would be nice to have the following implemented:
(a) Stress increased as combat or threatened combat or heavy U-boot damage situations occur.
(b) Regeneration in crew's quarters when outside these situatiions, but limited to a new max, which is continually lowered as situations mentioned in (a) occur or continue. The effects of how the red-bar-max is limited are cumulative during a single patrol.
(c) The effects of long-term combat stress are implemented as per discussed in 3) above (bell-curve model).

It might be objected that I am trying to combine both short term and long term combat stress here where (a) Beery is only trying to model long term combat stress and (b) we certainly don't have the resources to model both kinds of stress, given the SH3 framework.

Fair objection. I would say that regeneration is more a matter of short-term stress that goes away after the dangerous period is past. Regeneration shouldn't apply to long term combat stress, which is (by definition) something that continually grows as time goes on. I was trying to accomodate both by having a single stress meter. This meter would go up with stress, and some of it could be reduced with rest. However much cannot be reduced counts as 'long term'. The amount that can reduced is itself reduced as more and more tense situations occur. I think that this would work, if we could implement it. And if we could get both short term and long term combat stress, wouldn't that be even better?

In other words, long term stress should be linear with time, short term stress should not.

I think that the following are ways to most simply modify Beery's mod so as to address my concerns:
a) Choose to model either long term or short term combat stress.
b) If the choice is short term, implement some regeneration so that stress levels are not linear with time. We can't do limits (as far as I know). So have no limit, and allow all the stress to be eliminated. We have, after all, chosen to model short term stress (by hypothesis).
c) If the choice is long term, do not implement any regeneration. This is because we can't dynamically change stress limits during a patrol, and we don't want the crew to ever completely reduce long term stress and long term stress should always be cumulative.

But since we have chosen to only model long term combat stress, I would suggest that the values of the current mod are a bit too punishing. They are reasonable as values for short term stress, but the current mod is not doing short term stress (which is consistent with the decision to not have any regeneration). So my point is that if we're doing long term only, then let's do long term only. In other words, it should take longer to get stressed (as a matter of long term stress only).

So to keep with the current theme of Beery's model as a model of long term stress, I am in perfect agreement with no regeneration, but hesitantly suggest that perhaps the stress should accumulate at a slightly slower pace. I only hesitantly suggest this because I have played very little, and so haven't worked with an experienced crew. It may be the case that Beery's numbers are right on the money as far as an experienced crew goes.

My main complaint has just been that my sonar man (who I can't easily replace with a 'ton' of guys or a major officer) gets quite quickly stressed even without combat. Granted, this was a first patrol, but it was also without combat. I was just trying to get out of the Kiel area and past Norway. That seemed both unrealistic to me and also extremely punishing, since I need my sonar guy to operate efficiently and I can't 'overload' the compartment with stressed guys to get the efficiency up.

So if we're doing just long term stress and if we assume that Beery's numbers are the best, then I'm in complete agreement with the mod (modulo my discussion of the effects of stress in 3), whether this be long term stress, short term stress, or a combination of both, which probably can't be implemented).

This has been very pleasant procrastination on the papers I have to write. It's writing, but fun. I feel productive, but am really not getting anything done that needs to get done!