7 ms·
Yeah top 20 is a little excessive because in my own tests I found that top 20 is only marginally more accurate than top 10. You can get a more academic explana
by costco 4y ago
Yeah top 20 is a little excessive because in my own tests I found that top 20 is only marginally more accurate than top 10. You can get a more academic explanation [here](https://www.tandfonline.com/doi/abs/10.1080/09296174.2011.533591?journalCode=njql20 https://www.tandfonline.com/doi/abs/10.1080/09296174.2011.53...). I was amazed too because it seemed too easy!
- sillysaurusx 4y agoFWIW, top 20 was necessary for mine. The bolding was a brilliant move. Several of my accounts were ranked 10-20, but popped out due to the bolding.
- justusthane 4y agoWhat does the bolding indicate?
- sillysaurusx 4y agoThe explanation is here: https://news.ycombinator.com/item?id=33755466 https://news.ycombinator.com/item?id=33755466 As far as I’m concerned, it’s the killer feature of the app. The top 20 results may be noisy, but the bolded results have a signal to noise ratio close to infinity.
- jsnell 4y agoThe precision of the bolded results looks like maybe 30% to me. Significantly better than the non-bolded, but nowhere near perfect precision.
- costco 4y agoFalse positives become an increasingly difficult problem the more and more potential authors you introduce. If I had wrote a fancier model it probably wouldn't be as much of a problem but what can you do.
- jsnell 4y agoYes, this wasn't a criticism of the tool. It is crazy good. But I don't think people should be making the assumption that bolded results are definite alts, which sillysaurus' comment reads like.
- sillysaurusx 4y agoHmm, that wasn’t my intent. I see this tool as a recommendation engine more than a doxxer. By “signal to noise ratio close to infinity,” I meant that if you visit one of the bolded accounts, they’ll probably sound a lot like you. It’s one of those ideas that makes the tool substantially more effective, yet never would’ve occurred to me. It’s like the simplicity of pg’s “a plan for spam” algorithm: deceptively simple, but (like scrubbing dishes with fingers) works really well.
- tekknik 4y ago> I see this tool as a recommendation engine more than a doxxer. That is absolutely all this will be used for. This is a dangerous tool that serves no real world purpose.
- costco 4y agoThe funny thing is that I thought of it while eating dinner last night :)
- loeg 4y agoI have 7 bolded names (0.53-0.62) in the top 20 list, and none are alts of mine.
- ghaff 4y agoPretty much the exact same. (I do have a throwaway account but I rarely use it and it probably hasn't been used enough to qualify.)
- morsch 4y agoI'm one of them and I can confirm. But then again that's what I'd say if I was.
- loeg 4y agoHi style-adjacent friend :-). Just briefly looking at your recent comment history, we seem to find different kinds of articles interesting, but maybe have a similar writing style.
- dragonwriter 4y agoOf my top 20, 19 are bold, all are above 0.6, and I have no alts.
- notahacker 4y agoVast majority of my top 20 were bold, except you funnily enough! None of them are me (and you were the only one I recognised and thought "yeah, I can see where it gets it from"...)
- dimmke 4y agoMy results have 5 bolded users in my top 20, and I have 0 alt accounts.