Just ran a comparative benchmark on the new Majestic Trust Flow (TF) and Citation Flow (CF) update against the previous dataset. The delta is significant, and not in a uniform way. It appears the recalibration has introduced some latency spikes in score stability for certain domain profiles.
Key observations from my test set (500 sampled domains):
* **Volatility in Mid-Tier Scores:** Domains previously scoring TF 25-35 exhibited the highest variance, with some shifting ±8 points post-update. High-authority domains (TF > 70) remained relatively stable (<±2 point shift).
* **Anchor Text Weighting Change:** There's a noticeable dip in CF for domains with a high percentage of exact-match commercial anchors. The correlation coefficient between exact-match anchor ratio and CF delta was `-0.67` in my sample.
* **Data Freshness Lag:** The update seems to have re-processed historical data, but fresh backlinks (crawled <7 days ago) are not being integrated into scores with the same velocity as before. This creates a "staleness window" that wasn't as pronounced previously.
```
// Sample from my monitoring script showing score drift over 48h post-update
Domain | Old TF | New TF | Δ TF | Old CF | New CF | Δ CF
--------------|--------|--------|------|--------|--------|------
example-blog.com | 32 | 27 | -5 | 38 | 31 | -7
example-portal.net| 41 | 43 | +2 | 45 | 40 | -5
```
This isn't the first time a metrics provider has shifted the goalposts, but the opaque nature of the change makes longitudinal studies tricky. Has anyone else done a before/after analysis? I'm particularly interested in the impact on large-scale site audits where these scores are used for tiered prioritization. The "gotcha" here is making decisions based on a moving target without understanding the new baseline.
-- perfwise
p99 or bust