Probability Profiles allow you to incorporate different product mixes into your Label values.
We will start by expanding the example used in the Label Tutorial, where we used lbl_color to differentiate Green and Blue Work Items.
We are going to expand this example by adding some more detail. We now know that 45% of people are female, and 55% are male. Instead of saying that we are equally likely to have male and female Work Items, we are going to amend our simulation model to incorporate this new information.

Name your distribution with the naming convention 'Dist_xxx'. We could call this one dist_gender. This distribution can be edited to reflect the likelihood or your own values occurring. We have 2 values (Male/Female) so we only need 2 columns rather than the 10 displayed here.
Columns can be removed individually by right clicking and choosing 'Delete'. If you have several columns to remove, right click and choose 'Multiple' to add or remove more than one at a time. Columns can also be added to the distribution by right clicking and choosing 'Insert'.

Once you have amended your percentages, you should check the 'Value' associated with each column - the 55% column should have value 1 (representing male Work Items) and the 45% column should have value 2 (representing female Work Items). It is important to check these because we are going to link our Label to this distribution later.
A Probability Profile can be attached to your simulation model using a Label. First, create a Label (we can use lbl_gender from the Label example).

Select your distribution, and the Value box on the dialog will disappear. This is because all the values required have been set within the Probability Profile.
Now you have created and attached a Probability Profile within your simulation model. You can check that the correct proportions of Work Items are being assigned values 1 and 2 by checking the contents of the Male and Female Queues.
Remember, randomness is incorporated within the probability profile, so you won't get an exact 45% - 55% split, it's a percentage chance of a value being assigned.