There are a few sites growing in popularity in the market of brokering a group date between three guys and three ladies.

This is a concept on how to represent relationship data between six people divided into two separate groups with the least amount of input.

Each individual is presented with a triangle shape, and a singular marker which can be placed in the triangle.

The rules are simple, place your marker nearest to the person you want to go home with.

The vertices of the triangle represent each individual in the other group of the opposite sex. Over the course of the group date, they are asked to place (or move) their marker somewhere on the inside area of the triangle.

The marker's relative distance from each of the icons represents the individual's interest level in each members of the opposite group.

For example, I've chosen the marker placement of two individuals, Mike and Sarah. This is represented as bias in our model. Mike placed his marker nearest to Sarah, and Sarah (with greater bias) also placed her marker nearest to Mike. These two data points would represent a mutual match, and is the most ideal scenario as it's the easiest to determine a mutual level of interest.

With this information we know these two individuals are interested in each other. More so, we know that this is a mutual and exclusive match because none of the other individuals have placed their markers as near to Mike and Sarah as they did with each other.

The following is Mike and Sarah's markers on the same coordinate system, the dashed lines represents a threshold for the bias towards each individual.

### Modeling the entire group

I've represented the area of interest by each group as the vertices of each of their placed markers. This allows us to visualize a different dimension of data for free, where the inner triangle's area represents the groups TOTAL unique selection amongst themselves. The larger the area the more it can be interpreted as the most harmonious group-wise because each individual is exclusively interested in a different individual in the group of the opposite sex.

This data can be modeled with geometry as a time series, where the volume of the shape would again be representing the data in the figure above (Interest by group), however, over time (z-wise in our case). Imagine if the individuals were prompted to place the maker 3 times over the course of the evening. If you were to take the difference of both 3d volumes you could assumably have the total level of mutual interest in both groups. Potentially a measure for a successful experience in my opinion.

The last chart assumes you've been able to poll each individual via a mobile app or something, if that was the case you can potentially chart the velocity of their marker movements as indicators for intent, or as spikes of interest. This could yield something like a vector field, where the most interesting girls and guys would have the vectors pointing towards them. Maybe if the girl in question is a mega-babe, the guys would have moved the marker towards her really fast!?

### Notes and issues

Bias is negative because we are charting positive interest with distance and I didn't want to invert any of the other math.

It's also important to note that a marker placed or left in the center has a smaller coefficient for the bias. This would represent that the individual either has the same level of interest in all the other individuals, and does NOT represent amount of interest, this could be erroneous data unless input is required by the individualâ€¦? Not sure.

When making 1-to-1 comparisons, different comparisons can be made by transposing the coordinate system. This is done by reflecting the triangle x-wise along it's z-axis and bound by the top most vertex.

In the group based examples the location of the individual in their respective triangle is important, and shouldnâ€™t be individually transposed for comparison purposes.