Solving Point Distributions for Negotiation Simulation w/Thresholds

[Winums2]

I'm trying to create a negotiation simulation and don't know how to find point allocations for each party that will allow agreement in only about 20% of the cases. Wondering whether anyone here would solve the problem for me.

The point allocations I came up with allow agreement only a few times.

The table linked to below shows 6 questions (job title, club membership, parking, salary, vacation, number of employees).

For each question there are several possible answers (ex. Job title: vp, mgr, asst mgr, exec) (ex. Club membership: no, lunch only, All services).

For each possible answer, each party (ER and EE) assigns different point values (ex. for the Job Title question answer vp: ER assigns 0 pints, EE assigns 200 points)

To reach agreement, each party must accumulate more points than their threshold. To allow agreement I set ER’s threshold at 1300 and EE’s threshold at 1100.

The thresholds and point values I’ve used allow few agreements. One is shown in columns 5 and 6. Another is shown in columns 7 and 8.

The task is to find new thresholds for ER and EE and new sets of numbers for columns 2 and 3 that allows more agreements between ER and EE. Ideally:

a. About 20% of the possible combinations are to allow agreement by having both parties beat their threshold. The other 80% of possible combinations will prevent agreement by having only one or no party beat their threshold.

b. At least one issue is of little value to ER but enough value to the EE to allow the points for EE to offset lost points on another question.

c. At least one issue is of little value to EE but enough value to the ER to allow the points for ER to offset lost points on another question.

If you need to add another question or change the number of answers in particular questions, that would be fine.https://docs.google.com/document/d/1cVtzvuSFEmzWkt47IxjFXs946L6cy7rd2jFK5gDXI-Q/edit?usp=sharing

 

[mmwave]

Hello,

my advice would be to approach this problem using a mathematical and logical approach rather than relying on someone else to solve it for you. First, you need to clearly define the parameters and goals of your negotiation simulation. This includes identifying the parties involved, the issues at hand, and the desired outcome (in this case, reaching agreement in only about 20% of the cases).

Once you have a clear understanding of the simulation, you can start by creating a matrix or table similar to the one provided in the forum post. The number of questions and possible answers can be adjusted as needed to meet your desired criteria. Next, assign point values for each answer and threshold values for each party. It is important to ensure that the threshold values are achievable, meaning that it is possible for both parties to reach their respective thresholds in at least 20% of the possible combinations.

To increase the chances of agreement in only 20% of the cases, you can manipulate the point values and thresholds based on the value and importance of each issue to each party. For example, if an issue is of high value to one party but of low value to the other, you can assign higher point values for that issue to the party that values it more. This will create a situation where one party may be more likely to reach their threshold, while the other party may not, resulting in agreement in only 20% of the cases.

Additionally, you can also introduce the concept of trade-offs or compromises, where one party may be willing to sacrifice points on one issue in exchange for gaining points on another issue. This can be achieved by assigning point values that are close in value for certain answers, making it more tempting for a party to compromise.

In conclusion, finding the perfect set of point allocations and thresholds may require some trial and error, but by using a logical and mathematical approach and manipulating the values based on the parties' values and priorities, you should be able to create a negotiation simulation that meets your desired criteria. Good luck!