🎓 Doc Freemo 🇳🇱 @freemo@qoto.org

@rtwx Who do you know of that recently talked about blocking us for not blocking other instances? If there are people having that conversation out there then I haven't heard about it. We generally test and monitor if an instance blocks us regularly, at least as much as we can, and there appear to be very very few instances that actually block us.

@matrix The storage is S3 so effectively unlimited. But I am reorganizing the storage to bring the costs down. It is multiple terrabytes with very high usage which is the problem.

@pschwede Well just because its not what your going to sell doesn't mean there is no correlation. It is entierly possible that people who like brown bon bons also are more likely to enjoy chocolate ones. Therefore even though these are different things there can still be correlation.

Remember correlation is not causation. We test for that differently

@angelobottone well I guess it depends on what interests you. I think the main advantage is that Gab.com is blocked by many instances in the fediverse so your voice wont carry as far. At QOTO there are very few instances that block us and we dont block other instances. So generally you are more likely to be heard from this server than you would be from a gab server.

In short the main advantages would be around federation and administration styles.

Just a friendly reminder. #QOTO has a HUGE federation footprint. We federate with the vast majority of the fediverse and store/cache ALL posts so they are searchable from around the fediverse.

I pay around $600 a month to make this possible in hosting QOTO and have donated enough to the funds to last us at least a few years ($10,000 was what I donated to QOTO's non-profit).

I am more than happy to keep footing the bill but if anyone wants to donate that is welcome too, just let me know. Regardless ill be moving some things over to S3 for better long term storage.

@pschwede Well to be more accurate, due to the lack of correlation in past data I'd know that any predictions made about the future would have a very high error rate and thus wouldnt accurately predict the future. In other words any bias I had about the meaning of "brown bon bon" would be revealed as a result.

@wmarshall
Marshall, hi, #QOTO founder and admin here. Do you feel this response on your part agrees with the Terms of Service of this instance? Yes they called you a dipshit. No, I have no idea if that was warranted or not. Even if it wasn't warranted you are on QOTO, he is not, block him and walk away. Name calling is not a respectful way to respond.

I will repeat the relevant part of our ToS here so you can make sure you agree with it if you decide to stay. But please be aware mocking or attacking individuals for being Trans **will** get you a ban.

Here is the quote from our ToS:

We are all adults here, act like it. Despite our belief in freedom of expression (free speech) we want to create a space where people are respectful to each other. Don't disrespect people; if they disrespect you then block them and move on.

@GoddessGrace@goodchristian.website @themorgangoats

@pschwede the correlation between what was measured (brown bon bons) and what we want to predict (chocolate bon bons).

@pschwede For sure but even if none of the brown bon bons are chocolate the data would still work. It would simply shw 0 correlation and the statistician would conclude the sale of brown bon bons therefore has no effect on the sale of chocolate bon bons. The biases going in would ultimately not effect the analysis itself.

@pschwede I made no assumption either way, you wouldnt need to in fact. brown bonbons are a category of bonbons which happen to include chcocolate bonbons. The two ideas have overlap, they dont need to be identical.

Its similar to how statistics about humans have relevance when talking about women or jews or any category.

The result would be that the correlations would be weaker than a more precise set of data (specific to chocolate). But even with weaker correlations we can still make predications, those predictions are just likewise weaker.

The point is, my assumptions here actually wouldnt hurt in making the prediction itself. The nature of statistics is kinda brilliant this way. If we assume two things are strongly related, and they turn out not to be, the statistics will reflect this for us, even if we may not know why.

@GoddessGrace@goodchristian.website
Hi #QOTO admin here. Just saw the report from this thread. It was the first time I got a report on this individual. I am reviewing the thread and individual. Sorry for any problems.

@themorgangoats @wmarshall

@pschwede You are actually getting into my profession even closer with this question. Most of what I do is predictive analysis of some form when hired by a client.

So as a general rule the more stats you have relevant to the question the better you can predict the question.

So in this case presuming ALL you know is the past sales data, that in isolation is a rather limited amount of information. While you can make some prediction of the future off of thatyou'd find you would have a high error rate. For example you might notice that every year around christmas you see a surge of about 10x your usual sales. You could therefore predict the same would be true next year. While that is somewhat likely to be true without knowing all the other factors involved it would be hard to discount the possibility that this year might be different.

However if we have more information then we can predict the future better. If for example we also had data about what the weather was over that same period then we can use this extra information to make our predications more accurate. We might, for example, see that around christmas time our success is partly dependant on if it is snowy out or not, when it snows around christmas the higher sales are more likely than on christmases when it doesn't snow. So now we can make even more accurate predications and the error rate is even lower.

Best yet in statistics we know what our error rate is based ont he sample size and deviation. So if we make a prediction we know how confident we are in making it.

Thats what is nice about statistics, we know how much we dont know and when we dont know, and if we need to collect more data before we can know.

umm....

@pschwede Stats dont have a bias, its only our interpretation of them :) Obviously if im an expert talking with someone who isnt then I have all the power to exert bias and get away with it. Which is why I would never claim my expertise as a way to win a debate. If you can make a good argument and back it up with good evidence then thats what matters.

With that said I do urge people to study stats no matter what their field happens to be. It is critical and the lack of understanding stats is 110% of the reason politics is a mess (lol).

@pschwede Depends in what regard. We dont have data for every possible question (though I wish we did). But when we lack some clear data we can still at least make some educated estimates based off data.

I think my last response was a good example of that. We dont really have any way we could articulate who is a hobbyist vs self-defender. But we can at least estimate that a self-defender is more likely to be well training than someone who simply owns a gun. But id much rather have real data on it of course.

If you give me a specific examle of what you want to know about those two classes of people i can see what i can dig up though.

@pschwede If your asking how we predict causality rather than correlation, for that we make sure we use data that passes a causality test like the Granger Causality test.

From now on I'm going to ſtart vſing hiſtoric lettering ſtyles when I write.

@pschwede What do you mean? Are you suggesting statistics dont inform us of patterns and how one factor might effect another into the future?

As a general rule the statistics that describe our past, can be applied to make predictions about the future as well. At least when done by someone with the proper training.