[tansey]

Wow, how did this get discovered?

I'm a co-founder at EffectCheck and I was working closely with Scott this weekend as he was building this. It wasn't really ready for viewing yet, but okay... :)

Please note that the top graph is currently a mixture of two sets of data. The older points were using a less sensitive and improperly calibrated HN comment model, hence why everything is drifting around near "Typical." The points from 14:47 onward are using the correct model.

For those wondering why HN was so negative from 14:47-20:47, I believe the main topics of conversation were the Bitcoin and Sega debacles. Makes sense that people were really anxious given all that news.

[palish]

I'd like to know the scientific basis of EffectCheck. The site offers no explanation. Could you point out specific research papers, or link to technical info? Thanks!

[Stwerner]

I'll let tansey go in to more detail when he gets back. But for now, heres a link to one of his comments from last time EffectCheck was on here.

http://news.ycombinator.com/item?id=2603438

[palish]

So would you say this comment is an accurate description? http://news.ycombinator.com/item?id=2603825

"It seems that the point is to introduce their special-sauce black box, with an argument to authority about its methodology. I think the correlations you ask for are where the problems will lie, in that there is a value judgement that is being hidden. If I can put myself out on a limb here, I'd say that that measurement is going to be fundamentally unscientific."

Also, this seems like an important question, if you claim to be based in science:

"How can we falsify your claims? That is, what is a test we can perform that if it went a certain way, would show that your claims are false?"

It occurs to me that human color perception is another "subjective but scientific" field. You might want to research the experiments they performed for e.g. color matching in varying illumination levels, then come up with an experiment with a similar structure.

Specifically, if EffectCheck is accurate, then it should correlate with how 100 average people would classify e.g. a sample of 1,000 twitter messages.

[impendia]

EffectCheck looks really cool, but didn't seem to have any demo/try it out feature. I suppose it's not there? Or did I miss it? Thanks.

[bane]

Impressive work, the analysis among many different emotional vectors is interesting. How do you guys work with non-English languages as well?