Skip to content
MARPLA / Help center

Compare games

Compare games

Compare how several projects perform over the same period.

How to use it

  1. Add games to Favorites and select them for comparison.
  2. Choose a metric and a period. Select line names to show or hide them.
  3. Select a row in the table below the comparison to inspect one game.

Advice and practical context

Meaning of comparison

Define the question first: present scale, growth rate, or stability. Use matching daily mean CCU for scale, equal consecutive windows for direction, and ranges plus peak preservation for stability. One chart should not silently combine these different decisions.

Fair example

A young game can beat a mature one on growth while remaining smaller in absolute CCU; both findings can be true. The mistake is declaring one universal winner. State the criterion and peer set under which one line appears stronger.

A comparison improves when peers are chosen by age, Visits scale, and genre before viewing the result. If exact peers are scarce, relax one condition explicitly. Keep that relaxation visible: a mixed group adds context but warrants a more cautious conclusion.

Technical detailsCalculations, permissions and behavior

Using each view

Genres & studios

Open a genre or creator to see its games. The selected studio portfolio stays at the top while you scroll.

Compare

Add up to ten games with Compare in Favorites. The selection persists on this device after reload; removing every game leaves the view empty instead of inserting demo projects. The view opens directly with one full-width table and keeps its rows, headers and metric section aligned to the same width.

Rank is first, studio remains a subtitle under the game name and selected rows have no green fill. A horizontal switch below offers CCU, CCU / age, CCU / visits, Universe age, one-hour and 24-hour CCU trends, visits, rating, session, favorites and Signal. CCU uses the standard Playing now cell while taking the average of real MARPLA snapshots from the trailing 24 hours for a steadier comparison; the visits ratio is a percentage and the age ratio is per Universe-age day.

Positive and negative trends use distinct colors while their scales compare absolute change magnitude. The lower table headers sort its rows and a second click reverses the order. Signal uses the same standard numeric cell as every other list.

The selected metric shows its exact value and a linear scale relative to the maximum without an artificial minimum length. Select any row in the lower table: once the chart appears, the page places its top edge directly below the shared MARPLA header, even if loading changed the block height. Its header shows the project name, and a repeated set of the same metric buttons lets you switch the chart without returning to the table.

The chart uses saved MARPLA observations; metrics that were not stored previously receive no fabricated points and begin accumulating exact history after this update. The compare button in the upper table removes a game.

Observed sample

Sum of CCU for games in a MARPLA snapshot. The sample expands and is not all of Roblox. Use the external platform index for platform-wide activity.

Compare within groups

Sign in to see the same private groups used in Favorites above your comparison list. All compared games shows the full saved selection; Ungrouped shows selected games without groups. Choose a group to compare only its games already added to comparison.

Each count is the number of saved comparison games in that group; a game may belong to several groups. Search, sorting, metric tables and scales stay within the selected group. The chart hides when its game is excluded by the current filter.

Switching groups never removes games from comparison. Edit memberships through a game's star or in Favorites; updates are picked up automatically. If a group is deleted, all compared games are shown again.

Selected game chart lines

In the chart below the comparison, Line and Smooth modes connect all available points of the selected metric, including gaps between sparse snapshots. Selecting the five-minute step does not split the line when a game was observed less frequently. Markers and tooltips still show only real saved observations; a connecting line does not imply measurements inside the gap.

Bars mode has no connecting lines.

Useful articles: examples and decisions

The guide covers the steps in the service. These articles explain the metrics, examples, techniques and how to evaluate results.

Action essentials and individual concepts (3)