# conflict model: second method doesn't work

**URL:** <http://discourse.iapct.org/t/conflict-model-second-method-doesnt-work/13371>\
**Category:** CSG1995\
**Created:** [October 10, 1995, 6:12pm UTC](http://discourse.iapct.org/t/conflict-model-second-method-doesnt-work/13371 "1995-10-10T18:12:43Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![William\_T\_Powers4](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/w/977dab/32.png) [@William\_T\_Powers4](http://discourse.iapct.org/u/William_T_Powers4)\
**Post date:** [October 10, 1995, 6:12pm UTC](http://discourse.iapct.org/t/conflict-model-second-method-doesnt-work/13371/1 "1995-10-10T18:12:43Z")

</div>

[From Bill Powers (951010.1210 MDT)]

Kent McClelland --

A quick note about the conflict discussion. The cost-benefit model I  
described doesn't work. The reason is that if both systems are  
completely symmetrical, both incur the same costs and the same benefits,  
and there is nothing to offset the balance or get rid of the extra  
effort when the conflict is removed. The first method, with the leaky  
integrators, should work, but the second (the one I tested first, just  
now) only shows the drawbacks of verbal reasoning. I'll test the first  
approach when I get back from town this afternoon.

> **···**
>
> -----------------------------------------------------------------------  
> Best,
> 
> Bill P.

---

<div class="post-metadata">

**Author:** ![Kent\_McClelland2](http://discourse.iapct.org/letter_avatar_proxy/v4/letter/k/278dde/32.png) [@Kent\_McClelland2](http://discourse.iapct.org/u/Kent_McClelland2)\
**Post date:** [October 10, 1995, 9:15pm UTC](http://discourse.iapct.org/t/conflict-model-second-method-doesnt-work/13371/2 "1995-10-10T21:15:27Z")

</div>

[From Kent McClelland (951010.1600 CDT)]

Bill Powers (951010.1210 MDT)

> A quick note about the conflict discussion. The cost-benefit model I  
> described doesn't work. The reason is that if both systems are  
> completely symmetrical, both incur the same costs and the same benefits,  
> and there is nothing to offset the balance or get rid of the extra  
> effort when the conflict is removed. The first method, with the leaky  
> integrators, should work, but the second (the one I tested first, just  
> now) only shows the drawbacks of verbal reasoning. . .

Thanks for checking it out. Makes perfect sense, now that you put it that  
way. As I've said before, I find a lot of this stuff  
counterintuitive--it's hard to think straight about it without a model.

Some time ago you talked on the net about a possible syndrome in which a  
control system near the limits of output and unable to resolve errors just  
gives up, turns the gain down, or whatever. Do you recall that idea? Does  
it have any application here?

Kent