Oh noooo, it came out it doesn't work properly at all and I was literally p*ssed by my boss and sent to Tartarus, where I got status of miserable simp and now completely cucked and stink.

That's a scam, but not usual scam. That's really enormous scam. SCAMORZAAAAAAAA!!!! as we call it in Italy. Better use real calculator from my greek friends from the real Pantheon!

Just search for SCIENTIFIC bayesian average calculator - it's made by real decent greek scientist, not these ugly posers.

Phobos is right, I was scammed

Scam as everything this author's done for lifetime. Probably personality issue.

Poor addon, doesn't calculate bayesian waighted (average) at all. There are far better alternatives here in store.

Is this some kind of emulation of real calculations with random numbers generation? Because results I get are very far from those should be.

Not good. Value is just a randomly shifted fraction, not the real centered bayesian weighted rating value. And it doesn't work with limits properly at all.

It doesn't work properly, this is not Bayesian weighted rating. Inputs validation is also buggy. No one tested that at all, poor quality.

I believe I’ve already reviewed a previous iteration of this calculator. Still searching for a good tool for my students. Unfortunately, once again, this is neither a Bayesian-adjusted average nor a true calculator. In fact, this version is even worse in terms of the mathematical foundation despite bells and whistles like tsv export and trendy black color scheme.

Yet again, this so-called "calculator" relies on arbitrary, pre-defined "magic numbers" set by the author, making it more of a custom algorithm than a universal tool. The proof? Even the first row of inputs produces unexpected, incorrectly shifted results. The only time it aligns with a Bayesian-weighted method is when extreme values (either the upper or lower bounds) are used.

Speaking of "expected values," the outputs are not even on the same scale as the inputs. Instead of maintaining the original rating scale, the results are converted into a float between 0 and 1. A proper Bayesian-weighted rating should preserve the original scale. Means if the inputs range from [1;10], the outputs should too. Ironically, correcting this requires only a single basic arithmetic operation, yet the author seems unaware of such a fundamental mathematical concept from elementary school. It raises the question: why attempt to develop a math-related add-on without understanding basic school-level math?

Overall, this is a low-quality add-on with inconsistent and incorrect results, relying on arbitrary "magic numbers." It honestly feels like a school project that was released here for publishing trolling readons. To make matters worse, it seems like the same person (likely using a second account) keeps leaving overly positive reviews instead of actually testing or improving the tool. There’s no serious approach here - it’s disappointing, even for a school project.