Why DevTools throttling options are not changing my Lighthouse performance?
Every resource I can find about why Lighthouse performance scores are vastly different than desktop says it's because the bandwidth/throttling is set to 3G (still?), but changing the throttling settings doesn't change the performance rating for me. Am I using it wrong?
In the Lighthouse tab I select DevTools throttling (advanced):
There are no further settings after selecting the option, so I am presuming it uses whatever I have set in the Network tab, which is set for no throttling:
My other throttling options also do not affect the performance rating.
Could this be a browser bug?
Lighthouse has changed with the Update on august 2022. I found that they replaced the noThrottling option, with advanced Throttling. Unforgently the advanced Throttling setting will only affect the Trace display and not the performance score, as i understand it. THUS THE SCORE WILL ALLWAYS be affected by the simulated Throttling. Terrible product policy...
Lighthouse manual says:
The Lighthouse panel has a simplified throttling setup:
Simulated throttling remains the default setting. This matches the setup of PageSpeed Insights and the Lighthouse CLI default, maintaining cross-tool consistency.
If you click the View Original Trace button, the trace values will not match up with Lighthouse's metric results, as the original trace is prior to the simulation.
Applied throttling is available within the Lighthouse panel settings (⚙): uncheck the Simulated throttling checkbox.
In this mode, the performance data seen after clicking the View Trace button will match Lighthouses's numbers.
Find the documentation here:
- How to benchmark Swift code execution?
- Will an iterator surrounding an assert statement affect performance of production build?
- Improve Pandas and Vectorization Performance on large dataset
- Efficiently calculate an availability date from a list of project assignments and known capacity
- PrimeVue Table slow if dataset >200
- native memory leak of compose multiplatform
- Efficient solution for the same-fringe problem for binary trees
- Get Index of Sub Array in MongoDB
- Performance implications of iterating over set with --downlevelIteration - is a new array always created?
- When should I use ConcurrentSkipListMap?
- How to optimize screen real estate in a J2EE struts app?
- C++ replacement for C99 VLAs (goal: preserve performance)
- Using compound literals inside loop in C
- Performance cost of type comparisons
- What is the fastest way to read in a large yaml file containing lists of lists?
- Why loading some css and js file taking too long time?
- Quickest way to iterate over a pandas dataframe and perform an operation on a column, when what the operation is depends on the row
- Why is Query Optimizer Underestimating and Causing Data Spill to tempdb?
- Can I enforce the order of XML attributes using a schema?
- How can I optimize the performance of this numpy function
- How to move multiple elements with requestAnimationFrame() without having to create a new function for each element
- Difference between list.Count > 0 and list.Count != 0
- Check for winning tickets in lottery using SQL
- Performance in different variable scopes?
- Why this huge performance difference for an encapsulated Javascript function?
- How to add event delegation to "change" event (not click) on radio buttons?
- Variations with repetition of r integers in {0...k} that sum to u
- Why does the basic string formatting outperforms the built-in re.compile() function?
- What is the most efficient way to loop through dataframes with pandas?
- Fastest way to count occurrences of each unique element