Centova Technologies Inc.

Products and Services

Centova Cast
Customer Support
Centova Technologies
 

Knowledge Base | General | Troubleshooting

Track weighting issues

Question:

My general rotation playlists are playing the same tracks over and over.

-or-

Tracks from certain playlists are being played more/less frequently than I expected.

Answer:

This is a common complaint resulting from a misunderstanding of how the Centova Cast track weighting system works.

How Track Weighting Works

Track weighting is explained in the General Rotation Playlist Settings section of the Centova Cast user's guide, and a brief example is provided below.

Take the following playlist configuration as an example:

Type                 Track count     Weight
-------------------- --------------- ------
General rotation     1000 tracks     1
General rotation     5 tracks        10

In this example, weighting between the two general rotation playlists is calculated as follows:

1 + 10 = 11

So out of every 11 songs played, on average, 10 of them are going to be chosen from the 5-track playlist (the entire 5-track set will be repeated twice), and only 1 of them will be selected from the 1000-track playlist.

The end result is that you will seem to be hearing the same tracks over and over again on a loop. The solution is to simply weight each playlist based on how often you actually want a track to be selected from that playlist.

Explanation of Design Decision

We're often asked why we designed the track selection mechanism this way. This was a deliberate design decision to make sure that a large playlist doesn't overwhelm your station's programming. To understand why this is necessary, consider another example scenario in which you are operating a Pop Top 40 station.

In a Pop Top 40 station, most of the time you want to hear songs from the current Pop Top 40 charts. Occasionally -- perhaps once per hour -- you may also want to hear an older track as well, so as songs drop off the charts, you move them into an "Old Tracks" playlist.

Assume for a moment that Centova Cast handled weighting based on simple ratios, the way many Centova Cast users expect them to work. We'd like to achieve a 10:1 ratio, wherein Current Top 40 Tracks are played 10 times more often than older tracks, so we've set the weight for the Current Top 40 Tracks playlist to 10, and the Older Tracks to 1:

Type                Purpose                 Track Count         Weight
------------------- ----------------------- ------------------- ------
General rotation    Current Top 40 Tracks   40 newer tracks     10
General rotation    Old Tracks              400 older tracks    1

Unfortunately, this won't work the way you might have expected it to. Because the Old Tracks playlist actually contains 10 times the number of tracks as the Current Top 40 Tracks playlist, you will actually see a 50/50 split of airtime between the two playlists:

40 x 10 = 400
400 x 1 = 400

ratio = 400:400 or 1:1

Now, say you look at the numbers above and realize that by changing the Current Top 40 playlist to a weight of 100, you can achieve a 10:1 ratio:

40 x 100 = 4000
1400 x 1 = 400

ratio = 4000:400 or 10:1

That would work exactly as you expected -- you would hear Current Top 40 tracks 10 times as often as older tracks, at least for awhile. It's important to consider, however, what will happen to your ratios as you change the number of tracks in each playlist. Building on the example above, say after a year you have increased the number of tracks in the Old Tracks playlist to 800. You still have 40 tracks in your Current Top 40 playlist, and the weights are still set to 100 and 1. Suddenly your ratios are no longer the same:

40 x 100 = 4000
800 x 1 = 800

ratio = 4000:800 or 5:1

By simply changing the number of tracks in the Old Tracks playlist, you have skewed your ratios and will now hear old tracks far more often than originally intended.

This is a common example of why Centova Cast does not use simple ratios when calculating playlist weighting. Instead, Centova Cast uses the weighting algorithm described at the top of this article, which guarantees that no matter how many tracks you have in each playlist, your weighting values are always honored and always consistently applied.