About Parameters

Defining a good set of parameters is paramount to successful runcutting in BlockBuster. When defining the rules, you must establish what you are trying to achieve and focus on it.

• Check that the parameter rules accurately reflect the terms of business that are in place at your site.
• Check that the desirable/legal parameter values are the same as the undesirable/illegal rules that are defined in your runcut.
• When recutting runs, make sure that penalties for parameters of illegal runs are higher than those for the legal ones because you want to inhibit the system from creating illegal runs.
• When doing improvements, alter one parameter at a time then cut the runs. This allows you to view the combined effect of the altered parameter with all other defined parameters. Changing more than one parameter at a time obscures the factor that caused the changes in the new runcut. This effectively prevents you from learning how to manipulate the system to achieve the results you require.
• When altering parameters, remember that the desirable run types control the desirable runs in the runcut. In the same vein, the undesirable run types control the undesirable runs in the runcut.
• Do not perform multiple tasks while runcutting (for example, changing /checking parameters and analyzing runcuts). Perform one procedure at a time.
• After you have established a strong set of parameters that achieve the results you require, save them. You can use them as the starting point when making alterations to your runcuts in the future.
• If you change parameter settings when making incremental improvements to a runcut, save them so that you can use the saved scenario as a place of departure when making further improvements.
• Type values only for the parameters that you require. You do not need to type values for all the parameters available in BlockBuster; some parameters control elements that are not applicable to the data or terms of operation at your site. Make sure that all unused parameters are turned off. Check the message that appears at the bottom of the screen on how to turn off or inactivate a selected parameter. Typically, this means setting the minimum value at “0:00” and the maximum value at “50:00”. This practice enables the algorithms within BlockBuster to produce runcuts faster and it also eliminates extraneous factors leaving the relevant ones only. Avoid making large changes when altering parameters. If you must change the value of a parameter by a significant amount to achieve an improved runcut, save the scenario first.
• Afterward, take some time to create a new scenario and reduce the size of the discrepancy between the old parameter value and the one that resulted in an improvement. Perform a runcut and then view the Cost Summary. If the results of this cut yield the same improvement, the new parameter value is preferable. Repeat the exercise until you find the smallest discrepancy that yields the same improvement.
• Changing the value of the parameter by the least amount that still yields the same improvement is the optimal/best value to use because BlockBuster processes the smaller value more efficiently. This value can be considered the threshold.
• Only apply a Cumulative, an Occurrence, or a Worst Case penalty against any one parameter if required until you have acquired enough experience experimenting with BlockBuster using a particular set of data.
• Do not attempt to apply more than one penalty on a single parameter until you have acquired enough experience using BlockBuster with a data set, its work rules, AND have knowledge of the effect to the runcut of each of these penalties separately on any one parameter.
• If you have the experience required to try applying more than one penalty against a parameter or set of parameters, start by applying two penalties, (for example, if you have been using an Occurrence penalty, you may next decide to try using both an Occurrence penalty and a Cumulative penalty to achieve the result you want). Be sure to monitor the values you are using and closely analyze the results you achieve in the runcut using the additional penalty. Adjust the values by small increments in successive runcuts. As mentioned previously, make sure that the additional penalty you apply is the only change you make to the parameter settings for each runcut. If you are confident that you understand the effect of using two penalties on the data set you are using and why it is achieving the result it is giving you, attempt if necessary to apply a third penalty against the same parameter.