The Face/Off chart comes in two forms:

**Face/Off (Time Series Line)**is exactly as we've shown in this article, visualising metric Face/Offs as individual lines over time**Face/Off (Stacked Time Series Bar)**is identical computationally to the Time Series Line, but presents the data differently, as stacked bars for each discrete period

*π Note: unlike for the Stacked Time Series Bar Chart, the column total values in the Face/Off (Stacked Time Series Bar) will always be a sum of the values of the selected series.*

As a rule of thumb...

use the **Time Series Line** for:

**comparing relative series trends over time (relative***position)*relative values

**benchmarking segments against totals**overlapping series

all value types

use the **Stacked Time Series Bar** for:

**comparing series splits over time (relative***contribution*)absolute values

**splitting totals into segments**non-overlapping series

no averages

Let's take a look at why this is the case, with 2 examples.

## Benchmarking segment averages against the total - Line π wins!

Benchmarking the average health score of companies in my High-touch filter and Low-touch filter, against the average for all companies in my portfolio.

*Time Series Line*

In this example, the line view is optimal, since I'm interested in benchmarking, which means I care more about relative than absolute numbers, and have overlapping series, since both the High- and Low-touch segments are also represented by the portfolio total.

*Stacked Time Series Bar*

This view is not so good. Not only is it difficult to visually compare the segment averages with the total average, since each section size is similar and they are indexed to one another, instead of 0, but representation as a single bar doesn't make sense, as these three series overlap, resulting in double-counted calculations. For example, the portfolio total in week 20 is 4.56, but the week 20 bar total shows 17.8, because its double counted the High- and Low-touch segments.

Further, since these are period **averages, **segments do not relate to or interact directly with each other (in the same way as, for example, a sum), meaning the stacked view is not intuitive.

## Evaluating segment contributions to the total - Bar π wins!

Evaluating the relative contribution of different companies to my portfolio's total ARR.

*Time Series Line*

This view is less than optimal. Together, all these lines represent my total portfolio, but they appear unrelated: I'm less interested in each companies' relative position to one another, and more interested in their relative contribution to my portfolio overall. There's no benchmarking here, and the series are mutually exclusive.

*Stacked Time Series Bar*

This stacked bar view is perfect: since each series is simply a non-overlapping split of the overall series, this view clearly shows me the relative contribution of each company to my portfolio, and how that has evolved over time. Absolute values on each discrete segment, and on each discrete time period bar also enables me to visually evaluate, for example, how the ARR growth of individual companies has fed into total ARR growth, over time.