|
Metric
|
Value
|
P-Value
|
Commentary
|
| Root Mean Squared Error (RMSE) |
3.135
1.254
1.173
1.042
0.993
0.897
|
N/A |
Approximately how far the predicted metric is from the actual metric. A RMSE closer to 0 is better.
|
| Mean Absolute Error (MAE) |
2.099
0.932
0.892
0.752
0.689
0.629
|
N/A |
The average distance the predicted metric is from the actual metric. A MAE closer to 0 is better. |
| Pearson Correlation (r) |
0.456
0.916
0.923
0.933
0.933
0.940
|
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
|
Pearson correlation is a value between -1 and 1, where closer to 1 is better. |
| Spearman Correlation |
0.519
0.919
0.925
0.935
0.923
0.931
|
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
|
Spearman correlation is a value between -1 and 1, where closer to 1 is better. |
| Kendall's Tau Correlation |
0.371
0.757
0.764
0.786
0.776
0.787
|
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
|
A Kendall's Tau is a value between -1 and 1, where closer to 1 is better. For your channel, the model thinks one title/thumbnail combination will be better than another, and it is correct around
68.5%
87.8%
88.2%
89.3%
88.8%
89.4%
of the time. For reference, a random predictor would be correct 50% of the time. This is especially useful if you want to rank multiple title/thumbnail combinations using the YouTube Title and Thumbnail Combo Ranker tool. |
These metrics are subject to change as we improve the machine learning algorithms, and collect more
YouTube video
data for building more custom models. In general, the more data the models can learn from, the more
accurate
the predictions will become.
The best way to use this tool is to choose the title, thumbnail, and video duration combination that
leads
to the highest Predicted Views. If you consistently use the tool, you will find that your title and
thumbnails will on average receive more views.
