Arequire

Level 25
Verified
Content Creator
Does Ghostery's behavioral monitoring relies on third party cookies? Because I block all third party cookies in my browser and Privacy Badger is actually useless in this scenario which makes sense. Is it the same for Ghostery's behavioral analysis?
Doesn't appear to rely on cookies:
The Enhanced Anti-Tracking feature is a heuristic approach to real-time, algorithmic tracking protection. So, what does that mean? Like Ghostery’s main blocking mechanism, it monitors all third-party requests; however, instead of simply blocking those requests, it analyzes them for personally identifiable information and then strips out that information before the request is sent. Enhanced Anti-Tracking supplements Ghostery’s block-list approach by catching anything that might have been missed.
 

SeriousHoax

Level 27
Verified
Malware Tester
it monitors all third-party requests
Nah actually here it is. This refers to the third party cookies I'm talking about. So, same as Privacy Badger. If all third party cookies are blocked it won't have any data hence there's nothing to analyze.
Anyway, not everybody block all third party cookies so this feature of the extension would be helpful for most users. But I personally haven't faced any issue by blocking third party cookies so I would suggest everyone to try the same and see if anything breaks for them.
 

Lenny_Fox

Level 11
@SeriousHoax,

I thought you blocked all cookies.

About Ghostery & third-party
Ghostery still blocks the in video ads when blocking third-party cookies. Have you tested it?

About blocking cdn.ster.nl
Easylist does not block cdn.ster.nl. In fact when you search Easylist there is no ster.nl. The block you are noticing is on ADCONFIG filter. the general block on adconfig (which results in blocking adconfig.ster.nl) is actually the wrong one. When you block adconfig, the video's in UITZENDINGEN will not always play.
 
Last edited:

Lenny_Fox

Level 11
Brave has recently overhauled its adblocker making it a lot (69x) aster after Ghostery had published a perfromance benchmark in which Ghostery (surprise oh what a surprise) came out first and Brave adblocker last. Brave adblocker seems to be based on uBlockOrigin engine, using the tokenising of Ghostery (for effective block rule lookup/comparison). Both are open source projects, so there is nothing wrong with reusing their code. The info of Brave adblocker says it is written in Rust (so safer and faster than an extension written in javascript) and seems to monitor domain names only (no cosmic filtering) of Easylist (Ads and Trackers).

So its blocking is comparable with uBlock Origin in very easy mode (only is probably a little bit faster in absolute terms, a lot in relative terms),

_____
Disclaimer: I am no extensions nor adblocking specialist, just did some research for my school project.

Thanks all for the feedback/discussion (y)
 
Last edited:

Lenny_Fox

Level 11
@Lenny_Linux

Ghostery has a dubious past in user data privacy. Do you consider it still good to use?
Cliqz is under strict German and EU law. As long as it is open source and it is possible to opt-out I would say it is okay for power users. Average users probably can't find the opt-out (opt-in enabled by default). But many extension have telemetry and phone home enabled by default. So they won''t win a price for transparency with this enable by default policy, but they are no worse than the majority of other 'free' extensions.
 

HarborFront

Level 52
Verified
Content Creator
Cliqz is under strict German and EU law. As long as it is open source and it is possible to opt-out I would say it is okay for power users. Average users probably can't find the opt-out (opt-in enabled by default). But many extension have telemetry and phone home enabled by default. So they won''t win a price for transparency with this enable by default policy, but they are no worse than the majority of other 'free' extensions.
Open source does not mean it is safe. It just means its source code is open to public for scrutiny. Ghostery extension is owned by Cliqz GmbH, which is still an ad company. And Mozilla has a minority share in it too.

I would prefer uBO or Privacy Badger instead.
 

Azure

Level 25
Verified
Content Creator
Brave has recently overhauled its adblocker making it a lot (69x) aster after Ghostery had published a perfromance benchmark in which Ghostery (surprise oh what a surprise) came out first and Brave adblocker last. Brave adblocker seems to be based on uBlockOrigin engine, using the tokenising of Ghostery (for effective block rule lookup/comparison). Both are open source projects, so there is nothing wrong with reusing their code. The info of Brave adblocker says it is written in Rust (so safer and faster than an extension written in javascript) and seems to monitor domain names only (no cosmic filtering) of Easylist (Ads and Trackers).

So its blocking is comparable with uBlock Origin in very easy mode (only is probably a little bit faster in absolute terms, a lot in relative terms),

_____
Disclaimer: I am no extensions nor adblocking specialist, just did some research for my school project.

Thanks all for the feedback/discussion (y)
@oldschool

Here's the article that mentions the overhaul.
 

SeriousHoax

Level 27
Verified
Malware Tester
@SeriousHoax,

I thought you blocked all cookies.

About Ghostery & third-party
Ghostery still blocks the in video ads when blocking third-party cookies. Have you tested it?

About blocking cdn.ster.nl
Easylist does not block cdn.ster.nl. In fact when you search Easylist there is no ster.nl. The block you are noticing is on ADCONFIG filter. the general block on adconfig (which results in blocking adconfig.ster.nl) is actually the wrong one. When you block adconfig, the video's in UITZENDINGEN will not always play.
Hmm I see. Not all videos are playing on that site with uBlock Origin. It seems the site doesn't load all the videos from same source. The videos that are playing have no ads or blocked by uBlock Origin.
 

Cortex

Level 20
Verified
I used for some time recently Ublock but Google's upcoming plans made me wonder about using AdGuard Desk again as I already had lifetime licenses for 4 devices, so after some initial annoyances mainly to do with VPN's I'm really happy with it - I had configured UBlock with some advanced filters which frankly often broke websites if not for me then others. I now use I think the same setup as Umbra with AdGuard extension integrated with desk -. Unlikely to to change & no issues on any browser including current Brave.
 

vaccineboy

Level 2
On Chrome extension page: When I searched for "adblocker" (the very title of the study), AdGuard is the 3rd result; When I searched for "adblock", Adblock is the 2nd result.
On Firefox addons page, sorted by number of users: For "adblocker" or "adblock", Adblock is 4th, AdGuard & its clone AdBlocker Ultimate are 5th & 7th.
So I'm sorry that I don't understand why Adblock and AdGuard are not included, whereas Disconnect and Privacy Badger, by their own descriptions, are privacy tools to block trackers, not ads per se.
 

Lenny_Fox

Level 11
@Vacineboy

You should ask the researches, the study also provides clues based on technology used, let me recap for you.

1 The two adBlockers with highest numbers of users using AdBlockPlus syntax based content blocking engine using community gathered bottem-up lists as input
2. The two adblockers with highest number of users using their own filtering engine with their own top-down determined blocklist
3. Privacy Badger using a heuristics based filtering engine supported by their own top-down determined black and grey list with option for users to add trackers themselves. The latest Ghostery also applies heuristics to find unlisted trackers so with today;s technology Ghostery would probably be in the same category as Privacy badger.

The study was conducted in 2017,When they would repeat the study today I am sure they would also include Adguard because it uses AdBlockPlus syntax based blocking engine with its own (brute force) blocklist which started as copies of the Easylist community based bottem-up filters.

AdBlock or Adblocker ultimate when they use the same ABP syntax based blocking engine with the same community based bottem-up filters should score identical results (as for AdBlockPlus), so that would not provide any additional insights.

I think the take way is that smaller blocklists are surprisingly effective when you count the number of third-party requests (no matter whether the extension calls itself a content blocker, adblocker or privacy). So when Google's Manifest 3 limits the number of rules to 30.000 that is not the problem, the problem is that the API will be highly limited (probably killing most of the advanced blocking options like CSS manipulation, Scriptlet injection, websocket and xmltHHTPrequest monitoring etc).

That is why Adguard desktop user with lifetime licenses are pulling a long nose to the extension users and some Chrome users are evaluating Brave, Opera and Edge-chromium as alternatives to Google's Chrome.
 

Telos

Level 19
Verified
Content Creator
I use Adguard For Windows with integration mode for the extension (both all stealth enabled) on all my machines except the Linux one (obviously) and its filters are good enough so it doesn't required me to create rules.
Same here. Surprised it wasn't tested.
 

SeriousHoax

Level 27
Verified
Malware Tester
Same here. Surprised it wasn't tested.
The post is about adblocking extensions only so in a way it makes sense that he didn't test Adguard desktop app.
Btw, I installed Adguard yesterday and added 5-6 more filters along with the default one and the UI was telling me that I had too many filters so my the performance would be slower and after testing it was truly slow, I would say extremely slow. I wouldn't like to rely on Adguard filters only and if adding few more makes it slow then it seems Adguard isn't the perfect solution and adblocking browser extensions are still superior. For system wide adblocking I use another app but that's not an adblocker by design and probably not the topic of discussion here.
 
Last edited:

Lenny_Fox

Level 11
(y)Got an A (y)

Remember the project was about learning how to search the internet and debunking fake and forged information. studies and research results. Not to determine what was the best adblocker. Adblocking was a perfect example for learning to research (because it has all the elements of topics which are easy targets for fake news influencing).


I hit the jackpot (according to feedback teacher)
  1. Found a reliable source with large scale research which was relevant for the research question
    100.000 'unwalled' webistes out of Alexa top 200.000 conducted by an organization related to reputable top grade Austrian universities
  2. Found the glitch in the research methodology
    All extensions were tested on default settings except Ghostery.
  3. Found the surprising key finding
    Small expert based top-down blocklists are a little more effective than large community based bottem-up blocklists (less is more versus big is beautiful)
  4. Found the Achilles heel of both approaches
    • Les is more weakness: Expert based top-down blocklist are less effective than community based bottem-up lists against ad and tracking networks with a presence of less than 20 websites in the Alexa top 200.000 (the graph showing 0-20 as uBlockOrigin being the winner, in the 20-200 prevalence graph Ghoster overtook uBlockOrigin and had a clear advantage in the 200+ prevalence results).
    • Big is beautiful weakness: To fight the anti-adblock counter measures the community based ABP-syntax Adblockers develop increasingly complex syntax (like CSS and Scriptlet injection). The advanced CSS and super advanced Scriptlet sytax options require real IT-knowledge. Requiring more knowledge to write advanced blocking rules reduces the number of people who can contribute (they simply don't have that tech-knowledge). This probably negatively impacts the collective intelligence and rule writing (man)power of the community and as a result the (future) effectiveness of the blocklists made by this community.
  5. Best answer to research question
    The most effective Adblocking extensions probably are:
    1. Ghostery (winner with a nose length, representing the expert small blocklist players, but also using technology of the future, see advice)
    2. uBlockOrigin (currently the champion of the brute-force - community based - ABP-syntax blockers)
    3. PrivacyBadger (runner up and thought leader on heuristics based approach of the EFF)
  6. Conclusion and advice
    Heuristics based blocking is probably the best way forward, considering that:
    • Two studies showed that machine learning based adblocking is not around the corner yet (they need to be trained and training them with Easylist actually degraded their ability to determine most likely good or most likely bad with 40 to 60 percent depending on the ML-model)
    • My little field test in a small langauge area (Dutch language) with websites not in the Alexa top 200.000 showed that Ghostery was able to find new unlisted trackers (even in video ads hidden on a content delivery network). New Privacy badger also showed that it needs less training (than previous) versions) now it uses a blacklist (block all) - grey list (block cookies) and whitelist (allow) from the start. So the combo small expert list + behavior heuristics seems the best way forward (at this moment with current state of technology).
 
Last edited:
Top