Trump 2nd term prediction thread
So, looks like Trump not only smashed the electoral college, but is looking on track to win the popular vote, which seems to be an unexpected turn of events, but a clear sign of the current temperature in the country and perhaps the wider world.
Would be interested to hear views on how his 2nd term will pan out from both sides of the aisle - major happenings, what he's going to get done, what he's not going to get done, the impact of his election on the current conflicts in Ukraine and Gaza, whether his popularity will remain the same, wane, or increase, etc.
A bit of an anemic OP, I know, just interested to hear people's thoughts now that the election uncertainty is over.
What point? Whose point? My main objection was that whatever point you were making wasn't following the logical track of the conversation. That it was wrong was secondary. Are you drunk? I hope for your sake that you are, because if not, this whole discussion is not the best reflection of your comprehension abilities.
lol guy who didn’t know what a Gaussian is makes something up and calls it stats 102. What you said is obviously wrong. No statistician in the world will say you you should just assume non random data converges to population means.
That’s the whole magic of random data. It’s really the only subset we can prove reliably converges to population means. Again, this is stats 101. There is 0 justification for using other samples as stand ins and the generally moronic objection of “how can 500 peo
you are wrong and you know it.
unless and until you can describe the skew in the non-random sampling, you have to assume the sampling distributes like the original distribution.
there is absolutely no reason to believe men with a degree don't vote like men in all other subdemographics do, IE more for trump than women with the same characteristics do
Lol maybe tell pollsters about this? They spend so much effort trying to get random samples. Their jobs would be trivial if they could just get any sample and just assume it’ll converge to the population means. You’re a complete clown.
unless and until you can describe the skew in the non-random sampling, you have to assume the sampling distributes like the original distribution.
Incredible to see someone state this confidently. Luciom, you may think you are exercising your critical thought, or using common sense, or I have no idea what you think you might be doing. What you are actually doing is signaling very clearly that you have no idea what you are talking about to anyone that knows the first thing about statistics.
Which is incredibly consistent with pretty much everything else you try to speak confidently about.
Lol maybe tell pollsters about this? They spend so much effort trying to get random samples. Their jobs would be trivial if they could just get any sample and I just assume it’ll converge to the population means. You’re a complete clown.
Right? My sample of blue haired students with tongue piercings in Mass leads me to confidently conclude that democrats will win every election with 100% of votes in perpetuity.
I give you credit for a good deal of resilience, after all you've come this far; but you'll always been on the energy sapping side of every argument because when gibberish is posted the refutation is completely ignored and more gibberish will soon follow, and the effort to type a few broad brush generalisations is ~ an order of magnitude less than that to refute them.
Lol maybe tell pollsters about this? They spend so much effort trying to get random samples. Their jobs would be trivial if they could just get any sample and I just assume it’ll converge to the population means. You’re a complete clown.
jfc they know different segments of the population vote differently, they have the data from the past (and ofc they don't know if that holds, or not, and how much, in the present).
you don't know men with a degree vote differently than (being a man) U (having a degree) [which you know how they skew] .
I give you credit for a good deal of resilience, after all you've come this far; but you'll always been on the energy sapping side of every argument because when gibberish is posted the refutation is completely ignored and more gibberish will soon follow, and the effort to type a few broad brush generalisations is ~ an order of magnitude less less than that to refute them.
I am well aware of the debunking bullshit to spewing it ratio. My responses to gibberish generally become as terse as it becomes frequent. Doesn't sap much energy to say "**** you"; might even replenish some, in fact.
Right? My sample of blue haired students with tongue piercings in Mass leads me to confidently conclude that democrats will win every election with 100% of votes in perpetuity.
keep going by yourselves with absurdities as much as you want.
we do know how having tongue pierces, and blue hair, and living in an ultra blue state skew voting propensities.
but keep doing, what do you have to lose? you have no integrity anyway to begin with
keep going by yourselves with absurdities as much as you want.
we do know how having tongue pierces, and blue hair, and living in an ultra blue state skew voting propensities.
but keep doing, what do you have to lose? you have no integrity anyway to begin with
Oh look, it's our resident good faith posting champion accusing others of lacking integrity because he can't admit he's wrong. Quelle surprise.
So, we know that some criteria skew voting propensities. In fact, out of all the meaningful criteria you can think of, the majority of it probably skews voting propensities. Yet we have to accept that the criterion "having a STEM degree" does not skew the male/female split because trust me bro.
jfc they know different segments of the population vote differently, they have the data from the past (and ofc they don't know if that holds, or not, and how much, in the present).
you don't know men with a degree vote differently than (being a man) U (having a degree) [which you know how they skew] .
This is barely English now, but I'll just quote myself
No statistician in the world will say you you should just assume non random data converges to population means.
That's 100% true and nothing to do with politics. If you can't see why there is no point in continuing. You're not even bluffing well....like you can't go from not knowing what a Gaussian is to getting into an argument about statistics 3 days later lol.
lol its not hard to be right all the time. it is hard being right all the time. but I tried to tell you guys.
Look, I can't be ****ed with this. Clearly you can't follow a logical discussion. You made a point about a toy example, that point was wrong, now you're arguing that the toy example doesn't apply in the real world. Clearly, no **** Sherlock, that's why it's a toy example and its intention was to illustrate a very specific point by using extreme values. And I really can't be ****ed expending any more effort unweave your chaotic ramblings since you clearly think you're god's gift to logical reason
Welcome commrade to the land of misfit Marxists! It’s not him, it’s all of us.
This is barely English now, but I'll just quote myself
That's 100% true and nothing to do with politics. If you can't see why there is no point in continuing. You're not even bluffing well....like you can't go from not knowing what a Gaussian is to getting into an argument about statistics 3 days later lol.
For him everything got to do with politics…..
keep going by yourselves with absurdities as much as you want.
we do know how having tongue pierces, and blue hair, and living in an ultra blue state skew voting propensities.
but keep doing, what do you have to lose? you have no integrity anyway to begin with
If I told you there were groups called "Young Earth Creationist Mothers" or "Women Truck Drivers of Alabama", would you assume these groups skew left because their membership is 100% female?
If not, why not? And why does this reasoning apply to those groups but not to STEM degrees?
If I told you there were groups called "Young Earth Creationist Mothers" or "Women Truck Drivers of Alabama", would you assume these groups skew left because their membership is 100% female?
If not, why not? And why does this reasoning apply to those groups but not to STEM degrees?
I assume (or hope) he is saying that the women still are expected to skew left of the men in these groups. The fact there are no men just makes it uninteresting
simialrly with STEM degrees. i.e it's about the skew among those with/without within that subgroup
If I told you there were groups called "Young Earth Creationist Mothers" or "Women Truck Drivers of Alabama", would you assume these groups skew left because their membership is 100% female?
If not, why not? And why does this reasoning apply to those groups but not to STEM degrees?
I would assume "women truck drivers of Alabama" to skew to the left of "men truck drivers of Alabama", wouldn't you without other data?
Remember it's about STEM degree vs non STEM degree not STEM degree vs the whole population
I would assume "women truck drivers of Alabama" to skew to the left of "men truck drivers of Alabama", wouldn't you without other data?
Remember it's about STEM degree vs non STEM degree not STEM degree vs the whole population
What we don't know is if women truck drivers of Alabama skew more left than mathematicians of Alabama. It's insane to think you do know with 0 data. But I'm just repeating myself. If you don't understand that you can't just assume non random samples converge to population means, you'll effectively be off by 10+ constantly as you were when you tried to ignorantly use 2016 exit polls vs real data.
What we don't know is if women truck drivers of Alabama skew more left than mathematicians of Alabama. It's insane to think you do know with 0 data. But I'm just repeating myself. If you don't understand that you can't just assume non random samples converge to population means, you'll effectively be off by 10+ constantly as you were when you tried to ignorantly use 2016 exit polls vs real data.
But the bold isn't what we were looking at. What we were looking at is mathematicians in Alabama vs people with other degrees in Alabama (actually we did for the whole nation).
If you know NOTHING of a degree, if you don't know how it skews culturally, by age, by religion, by ethnicity whatever, if you have zero info, you would expected degree X holders to vote like other degree holders.
Uncertainty will exist on that but starting from "how do other degree holders vote" *would still be the best estimator you have*. It's not about "non random and population mean", it's that if you don't know ANYTHING about the structure of the non-randomness, the fact that it is non random does not change your priors.
Again, it's 200 balls in a jar, pick a sample and assess the color distribution, and you pick those on the bottom instead of "randomly". Being at the bottom is non-random but unless you know of a mechanism that skews the color for being at the bottom, it doesn't matter.
When you have data on degree holders voting trump (say) 40-60, your automatic best predictor for all subsets of degree holders will be ... 40-60, until and unless you add other data to that.
You can't claim the 40-60 has NO relevance for the estimate of trump voters of a subset of degree holders (!!!).
Now if you gather data telling you that a specific degree is made up exclusively of committed atheists, and you don't know anything else about it. While others aren't. Will that move your 40-60 for that degree against Trump?
If the answer is YES, you agree with me and what you wrote is false and not applicable to what we are looking for.
If you wrote no, you mean that the knowledge the totality of degree holders in "discipline K" being atheist provides ZERO information about their voting propensities. Which is kinda insane, and absurd.
You posit a mechanism "could exist" by which even if atheists overwhelmingly vote democrat, being both an atheist and degree K holder completly changes patterns that exist everywhere else in society. While the possibility of that is non-zero of course, claiming the knowledge they are atheist doesn't matter at all (which is what you claim), doesn't inform us at all, is frankly ridicolous beyond comprehension.
So back to STEM degree holders, knowing that they skew male exceptionally more than all other degree holders, that will skew them pro trump more than other degree holders, unless other mechanisms are identified that more than compensate for that skew that exists in the whole of society.
So you need to find other measurable elements that skew pro democrat, which STEM degree holders have more than all other degree holders, to try to counter that undeniable pro trump biased linked to male-ness
Luciom,
eca is correct. You would not necessarily expect non-random data to converge to population means. And you certainly wouldn't assume that to be the case.
If you know absolutely nothing about the non-random process yes. If you do, that's what you use to correct for the non-randomness.
The "non-random" process has to be treated as random if you have absolutely no information about it.
Imagine there are 10k people selected at random across a population. Someone comes and divide them arbitrarily in 10 1k groups. You pick the group and have 0 idea of the criteria used by the guy who divided the group, and start measuring those 1k people.
Your uncertainty about the measurements of the 10k people , and the population they got picked from, will increase, but you can still use the data from the 1k people you measured as informative more than 0 about the original population.
