Based on ACTUAL data sets from more than 80 websites.
It seems to be related more to content than links along with other micro factors. In general:
- Sites that had active link acquisition and low index bloat and had a tight index of performance pages saw an INCREASE in rankings and traffic.
- Sites that had no active link acquisition other than general background flux in random links / scrapers but had a large content pool where more than 50% of the index underperformed saw significant drops of 30% plus
- Sites that had active link acquisition with good domain metrics and a good portion of relevant links but large index bloat also saw drops of 20-30%+
- Sites that had no active link acquisition but had a good tight index of performing pages (less than 20% underperforming index) saw an INCREASE in rankings, as much as 25%
Factors where there was no real correlation:
- Any change in weight on site speed (think with google)
- Site age or establishment over time
- Brand trust/citation
Factors where there was SOME correlation:
- Poor quality links
- Negative Link Patterns i.e. Blogspot abuse
- Massive skew between links at a root domain level and lower level pages
- Anchor text profiles