Well it means that this site isn't for learning, but socializing, and general bellyaching.
Ah, but you can learn a great deal from watching people socialize, and even more from observing what they choose to bellyache about and how they go about it. Given that you get to see both sides of every issue do this, you can actualy learn a great deal here.
So, tens of thousands of concurent users on a single web application (9,000 to 36,000). The first thing I'd do is check those numbers. That sounds high (e.g., larger than anything Citibank, or Intel is running). If the numbers are acurate, you may want to reconsider making it a web app.
But the real question isn't how many users, it's how much load they'll impose. Assuming that the number of users is correct, it sounds like a POS situation (but 120 registers at one site?), so maybe one full transaction cycle a minute from each user worst case, or 500 per second.
Do you have any idea how many of these will be pure read (customer found, in the example) and how many will be read-fail-write (new customer)? Remember to de-rate writes, since they will have to enter/verify the customer data.
Depending on the deatils, it may make sense to replicate the data at each site (for fast queries), and only propagate updates.
What was once a "high volume" app may be nothing for modern equipment. You're talking about on the order of 1K concurrent users (300 sites * several users per site).
If "use" means manually typing data into forms, viewing mostly static pages, etc. this isn't really a very "high volume" application, and a single decent server should handle it.
If, on the other hand, "use" means constantly running complex queries against a billion item data set, you're doomed.
P.S. Clever use of the text describing the link may help you control how much trafic you get, from low ("M. Moore Nude!") to high ("SCO caught robbing courthouse").
the fact that your party base would rather Stay at Home then vote if it ment them going through rain, or waiting in line.
I saw an interesting graph of line-length vs. "major party of precinct" vs. "party affiliation of election official"; it appeared that lines were significantly shorter if the precinct tended to vote for the party of the person in charge of the process. Of course, this doesn't have to mean anything ominous (it could, for example, have been due to subconscious bias in allocation of resources) but my experience with political types makes me doubt their innocence.
Oh, and (at least in the graph I saw) there didn't seem to be more of a problem with the red team or the blue team; in that sense, at least, it was "fair".
If the tax load was reversed (bulk staying local, state getting about the same, and the Federal government getting the scraps), we'd have at least a more accountable government - and the temptation for misusing other peoples money should be less when you know those other people.
Oooo...sort of like it says in the constitution.
I like that idea too!
-- MarkusQ
P.S. It's interesting to note that the exact same issue (regarding the states rights to conditionally vote on constitutional amendments) came up with both the equal rights amendment and the federal income tax amendment, but the offical result is that we have the federal income tax but we don't have the equal rights amendment. What's that all about?
Uniform regulations? Perhaps.
Paper ballots? Yes.
Unambiguous paper trail? Definitly!
Better Candidates? Oh, would I love that!!!
But the number one thing we could do to fix the whole mess would be to move the budget to the tax form. When you fill out your tax form, you also say where you want the money to go. But the total you allocate can not exceed (or will be prorated to match) what you paid.
The real problem with our democratic system? It lets whoever we ellect decide how to spend other people's money--sometimes up to hundreds of billions of dollars of other people's money. Expecting everyone (or anyone!) to stay honest in such a situation is just foolish.
Do you only lie to the exit-pollers if you've voted electronically?
Bingo. That's the $64 question. If we are to accept this theory as an explanation of the disparity between the exit polls and the official counts, we have to accept it in its full glory:
Theory: Bush supporters systematically lied to exit pollsters and claimed to have voted for Bush, but only after voting on specific types of voting machines in specific states (the so-called swing states). Other Bush supporters were (on average) as honest as the general population.
It seems a little far fetched when you work it out...
You are not being asked to believe that Bush supporters have been lying. You are being asked to believe that Bush supporters simply refused to talk to the polsters.
No, the claim is that he (and, by inference, many more like him) flat out lied. See the subject line of the post that started this thread. Other posters on this thread have tried to rationalize it by citing the behaviour of Kerry supporters (implying that the posters are Bush supporters) but that is beside the point. Since the whole issue here is "why did the exit polls show a strong lead for Kerry, and the actual results showed a win for Bush", the theory only makes since if it were Bush supporters doing the lying.
If we are expected to believe that Kerry supporters lied to exit pollsters (by falsly claiming to have voted for Bush), we should have seen polls predicting an even stronger win for Bush than the actuals, which is not what happened.
The only way to explain the known facts with a theory of "voters lying to pollsters" is that the vast majority of the lying was done by Bush supporters, which I find a little hard to accept.
The previous poster had a valid point. The networks use of exit polls to call the election DOES affect other peoples vote and their turnout. If it is bad for him to LIE to an exit pollster in order to, in your opinion, "influence" the outcome of an election, then why don't you have a problem with what the networks do? They DEFINITELY influence an election.
Since (according to you) I'm an ignorant fool, I don't expect you to take my point too seriously, but here it is:
It is perfectly acceptable to try to influence others by telling them the truth, the whole truth, and nothing but the truth.
It is wrong to try to influence others by knowingly lying to them.
If this is what you'd call foolishness, I'm glad I'm not what you'd consider a wise guy.
I'm not saying that the first count should be perfect. I'm saying that insisting that the process goes quickly should not take priority over insisting that everyone can see that the process was conducted fairly.
In an election such as this one, where advocates of both major parties were clearly trying to do whatever it took to win, and accusing each other of all sorts of illegal tricks in the days leading up to the election, we should invest the time to make it absolutely clear to everyone that no one got away with anything crooked. That means looking closely enough that we catch anyone who has up to this point gotten away with it, and nailing them. It also means digging into the data enough that everyone can see that the final result is fair and accurate, even if it takes a while.
Found out who she was polling for, and contacted them about the sample bias as well.
In general, drawn the fact of the bias while the process was still going on.
It is quite possible that there was a legitimate reason for the bias (perhaps she was polling for someone who was not looking to predict the outcome but rather to assess the sentements of a particular demographic. Or perhaps she was there to cross check a suspected bias in another survey. Or...)
In any case, at this point, whining because you didn't get a chance to lie to her is just...silly.
Do I want a pollster to have the truth? Not especially. Because what counts is the election, NOT the EXIT POLL! If I vote Kerry, and tell the pollster I voted Badnarik, his data is slightly slanted, but MY VOTE doesn't change because I lied to a pollster.
But you run the risk of having a legitimate election questioned (one of the tests we use when assessing the fairness of elections over seas is, did the results square with the poll watcher's / exit poller's tally?--if 90% of the voters say they hate Bob Despot and want him out, but he gets 99% of the vote, we tent to say there was fraud). You also run the risk of having actual fraud go undetected.
And finally, you have to face the fact that lying to lose is stupid, lying to win is wrong, and lying in the hope that you will neither win or lose is pointlessly nilistic.
why should California know how Florida voted until after the election is over? what is the electorate being informed of? that their candidate may or may not be winning so they are more/less inclinded to try and make it to the polls?
Why do people in the USA need to know about earthquakes in Japan? Why do we need to know about internal political matters of Sudan? Why do people in Iraq need to know the motives of the foreign (mostly incomprehensibly english-speaking) invaders? Why does anyone need to know about Enron's internal accounting practices?
The problem with questions like this is that it presupposes that you are more qualified to decided what they need/want to know than they are. If you don't want to know such things, don't watch/listen to them.
-- MarkusQ
To quote an old saying "The truth only comes as an enemy to those who have forgotten how to welcome it as a friend."
why does the news media "project winners" before all of the polls are closed? If, according to exit polls a candidate is winning/losing, it might affect how people in states where the polls are still open vote.
In other words, I do it to try and screw up their process
So you are one of the people I referred to as "trying to rig an election," albeit on a very small scale; you are hoping that by lying you can influence how (or if) one or more of your fellow citizens vote. In my book, that's no more honorable than intentionally giving bad directions to the polling place, or tossing the registrations of people you disagree with. For you (it seems) winning is more important than honesty.
But at least now I understand your motivations.
-- MarkusQ
P.S. In answer to your rhetorical question, the news media projects outcomes based on the theory that a well informed electorate is a key component of an effective democracy.
It doesn't matter if it changes the results or not. We need a fair and open examination of all of the issues, regardless of any sort of party nonsense. The way to insure trust in a process is to audit the hell out of it. Track down every error, even if it's only pennies, account for every discrepancy, and make the whole process completely open to public scrutiny.
We owe it to ourselves, and to each other; we owe it to the candidates and their supporters who may be being slandered and (if any of them are actually guilty) we owe it to any cheaters to shine some light on their accomplishments as well.
If we plan to export freedom and justice against entrenched politics and religious biases around the world, we'd better make them our priorities at home as well.
I don't understand this attitude at all. Why would you lie to exit pollsters? Do you lie to your doctor when you go in for a checkup? Do you lie to the waiter about what you want to order in a restraunt?
Having accurate exit polls is to the advantage of everyone--everyone, that is, except people trying to rig an election. They are the only ones who benefit from trashing the exit polls. Are you trying to help them?
For that matter, why is it that we are expected to believe not only that lying is rampant, but that it is much, much more common for the sort of people who place high importance on "moral values" to lie? Remember, it's not as if a bunch of Kerry supporters are supposed to have lied and said they supported Bush, is it? It the conserviative, upright rural Bush supporters who think moral values are very important that are supposed to have lied en mass. Does that make sense?
He spends money with more abandon than any liberal in history
He lies to the American public
He lauches unprovoked attacks into third world countries
He can't figure out how to win them once he's started them
He shows no respect for the constitution
He shows no respect for civil liberties
He uses legal loopholes and questionable logic to rationalize going against pretty much every American Ideal, from "innocent until proven guilty" and "seperation of church and state" to "the right to a fair and speedy trial"
He routinely places the good of corporations over the good of individuals
etc., etc.
Before any Democrats reading this get to smug, ask yourself: was Kerry the best you could come up with? "I have a plan" and "Wrong, Wrong, Wrong"?
Both parties have any number of sensible, credible people they could have run...and the real problem is they ran Bush and Kerry. The race was so close for so long mostly because neither one of them was worth voting for, except as a way to keep the other from winning.
Ah, the joys of getting to the truth in a digital age. I did not provide a link to the final "corrected" 1.5% figure (and qualified it with an IIRC) since I was unable to find one on a quick search. I have not been able to find one with a deeper search, though I have found the raw data on which the corelation with counting method was based. There is kind of a quantry researching things on the internet (boy, is that an understatement!). Things frequently go away / get edited, but apart from google cache, there isn't a good way to retain the information with its provenance. I keep local copys of many interesting thing on my hard drive, but mostly when the information (e.g. technical documentation) is useful without proof of authenticity.
To your other point, it would be interesting to see if there is a correlation between the exit polling descrepancies and where the counties are geographically. This would be a good cross check/sanity chek on the counting method correlation. Another interesting analysis that some people are doing is looking for correlation by candidate. If some people lie to exit pollsters (one theory) why would the supporters of one particular candidate in a specific region (e.g. Bush supporters in Ohio) be much more likely to lie than voters in general?
I still don't buy it. It's like saying that if ATMs gave receipts mugging ten thosand people and demanding ten dollars each would be safer than embezzling a hundred thousand dollars. But even so, the receipt dosn't have to contain a printed record of the vote (just as there's no reason for your ATM receipt to contain your account number and PIN)--all that's needed is a trail linking the votes counted to the votes cast that can be validated by the voter.
As for who's held accountable, once fraud can be demonstrated, there's well defined chain of resposibility (some on made the machines, some one certified them, etc.) and we already have ways for dealing with exactly this sort of problem--which is why it's safe to put your money in the bank, for example.
I'm sorry to say that there was no widespread fraud in all e-voting states.
How odd. I'm sorry to say that I think (just from looking at the data) there was fraud in the swing states (by the Republicans) and potentially in some congresional races (by the Democrats). Were you hoping that the fraud would be more wide spread or something?
It's just not possible. There are thousands of people involved, thousands of pieces of equipment, many, many, many election and other government officials at all levels in extremely disparate jurisdictions with different ways of doing things, with no way for any central entity to reach these machines after the fact.
Couldn't you use the same logic to claim that nothing bad ever happens? There couldn't be insurgents in Iraq, because there would have to be too many people involved? There couldn't have ever been racism or slavery because it would have involved people with different ways of doing things? For that matter, you could argue that the 9-11 hijackers couldn't have pulled off their plan, etc.
I'll agree that your argument could be used to claim that they would have a hard time going unnoticed, but the very fact that we're all discussing it here already proves that it didn't go unnoticed, so no argument is needed on that count.
In the Democratic Primary in NH, early exit polls showed Howard Dean and John Kerry in a dead heat. Yet when the votes were actually counted, Kerry won by a wide margin.
Well, that example doesn't help your case much. The sequence was:
Exit polls (as we all remember) showed a dead heat
Slashdot Mirror
Ah, but you can learn a great deal from watching people socialize, and even more from observing what they choose to bellyache about and how they go about it. Given that you get to see both sides of every issue do this, you can actualy learn a great deal here.
All it takes is an open-yet-skeptical mind.
-- MarkusQ
So, tens of thousands of concurent users on a single web application (9,000 to 36,000). The first thing I'd do is check those numbers. That sounds high (e.g., larger than anything Citibank, or Intel is running). If the numbers are acurate, you may want to reconsider making it a web app.
But the real question isn't how many users, it's how much load they'll impose. Assuming that the number of users is correct, it sounds like a POS situation (but 120 registers at one site?), so maybe one full transaction cycle a minute from each user worst case, or 500 per second.
Do you have any idea how many of these will be pure read (customer found, in the example) and how many will be read-fail-write (new customer)? Remember to de-rate writes, since they will have to enter/verify the customer data.
Depending on the deatils, it may make sense to replicate the data at each site (for fast queries), and only propagate updates.
-- MarkusQ
First step, do the math.
What was once a "high volume" app may be nothing for modern equipment. You're talking about on the order of 1K concurrent users (300 sites * several users per site).
If "use" means manually typing data into forms, viewing mostly static pages, etc. this isn't really a very "high volume" application, and a single decent server should handle it.
If, on the other hand, "use" means constantly running complex queries against a billion item data set, you're doomed.
So where do you fall in this spectrum?
Coming up next...where's the bottleneck?
-- MarkusQ
Post a URL here and we'll help.
-- MarkusQ
P.S. Clever use of the text describing the link may help you control how much trafic you get, from low ("M. Moore Nude!") to high ("SCO caught robbing courthouse").
I saw an interesting graph of line-length vs. "major party of precinct" vs. "party affiliation of election official"; it appeared that lines were significantly shorter if the precinct tended to vote for the party of the person in charge of the process. Of course, this doesn't have to mean anything ominous (it could, for example, have been due to subconscious bias in allocation of resources) but my experience with political types makes me doubt their innocence.
Oh, and (at least in the graph I saw) there didn't seem to be more of a problem with the red team or the blue team; in that sense, at least, it was "fair".
-- MarkusQ
Oooo...sort of like it says in the constitution.
I like that idea too!
-- MarkusQ
P.S. It's interesting to note that the exact same issue (regarding the states rights to conditionally vote on constitutional amendments) came up with both the equal rights amendment and the federal income tax amendment, but the offical result is that we have the federal income tax but we don't have the equal rights amendment. What's that all about?
Uniform regulations? Perhaps.
Paper ballots? Yes.
Unambiguous paper trail? Definitly!
Better Candidates? Oh, would I love that!!!
But the number one thing we could do to fix the whole mess would be to move the budget to the tax form. When you fill out your tax form, you also say where you want the money to go. But the total you allocate can not exceed (or will be prorated to match) what you paid.
The real problem with our democratic system? It lets whoever we ellect decide how to spend other people's money--sometimes up to hundreds of billions of dollars of other people's money. Expecting everyone (or anyone!) to stay honest in such a situation is just foolish.
-- MarkusQ
-- MarkusQ
Bingo. That's the $64 question. If we are to accept this theory as an explanation of the disparity between the exit polls and the official counts, we have to accept it in its full glory: It seems a little far fetched when you work it out...
-- MarkusQ
No, the claim is that he (and, by inference, many more like him) flat out lied. See the subject line of the post that started this thread. Other posters on this thread have tried to rationalize it by citing the behaviour of Kerry supporters (implying that the posters are Bush supporters) but that is beside the point. Since the whole issue here is "why did the exit polls show a strong lead for Kerry, and the actual results showed a win for Bush", the theory only makes since if it were Bush supporters doing the lying.
If we are expected to believe that Kerry supporters lied to exit pollsters (by falsly claiming to have voted for Bush), we should have seen polls predicting an even stronger win for Bush than the actuals, which is not what happened.
The only way to explain the known facts with a theory of "voters lying to pollsters" is that the vast majority of the lying was done by Bush supporters, which I find a little hard to accept.
-- MarkusQ
Since (according to you) I'm an ignorant fool, I don't expect you to take my point too seriously, but here it is:
- It is perfectly acceptable to try to influence others by telling them the truth, the whole truth, and nothing but the truth.
- It is wrong to try to influence others by knowingly lying to them.
If this is what you'd call foolishness, I'm glad I'm not what you'd consider a wise guy.-- MarkusQ
I'm not saying that the first count should be perfect. I'm saying that insisting that the process goes quickly should not take priority over insisting that everyone can see that the process was conducted fairly.
In an election such as this one, where advocates of both major parties were clearly trying to do whatever it took to win, and accusing each other of all sorts of illegal tricks in the days leading up to the election, we should invest the time to make it absolutely clear to everyone that no one got away with anything crooked. That means looking closely enough that we catch anyone who has up to this point gotten away with it, and nailing them. It also means digging into the data enough that everyone can see that the final result is fair and accurate, even if it takes a while.
-- MarkusQ
Good point. But what you should have done was
- Gone up to her and told her how you voted.
- Challenged her on her sample bias.
- Found out who she was polling for, and contacted them about the sample bias as well.
- In general, drawn the fact of the bias while the process was still going on.
It is quite possible that there was a legitimate reason for the bias (perhaps she was polling for someone who was not looking to predict the outcome but rather to assess the sentements of a particular demographic. Or perhaps she was there to cross check a suspected bias in another survey. Or...)In any case, at this point, whining because you didn't get a chance to lie to her is just...silly.
-- MarkusQ
-- MarkusQ
But you run the risk of having a legitimate election questioned (one of the tests we use when assessing the fairness of elections over seas is, did the results square with the poll watcher's / exit poller's tally?--if 90% of the voters say they hate Bob Despot and want him out, but he gets 99% of the vote, we tent to say there was fraud). You also run the risk of having actual fraud go undetected.
And finally, you have to face the fact that lying to lose is stupid, lying to win is wrong, and lying in the hope that you will neither win or lose is pointlessly nilistic.
-- MarkusQ
Why do people in the USA need to know about earthquakes in Japan? Why do we need to know about internal political matters of Sudan? Why do people in Iraq need to know the motives of the foreign (mostly incomprehensibly english-speaking) invaders? Why does anyone need to know about Enron's internal accounting practices?
The problem with questions like this is that it presupposes that you are more qualified to decided what they need/want to know than they are. If you don't want to know such things, don't watch/listen to them.
-- MarkusQ
To quote an old saying "The truth only comes as an enemy to those who have forgotten how to welcome it as a friend."
So you are one of the people I referred to as "trying to rig an election," albeit on a very small scale; you are hoping that by lying you can influence how (or if) one or more of your fellow citizens vote. In my book, that's no more honorable than intentionally giving bad directions to the polling place, or tossing the registrations of people you disagree with. For you (it seems) winning is more important than honesty.
But at least now I understand your motivations.
-- MarkusQ
P.S. In answer to your rhetorical question, the news media projects outcomes based on the theory that a well informed electorate is a key component of an effective democracy.
It doesn't matter if it changes the results or not. We need a fair and open examination of all of the issues, regardless of any sort of party nonsense. The way to insure trust in a process is to audit the hell out of it. Track down every error, even if it's only pennies, account for every discrepancy, and make the whole process completely open to public scrutiny.
We owe it to ourselves, and to each other; we owe it to the candidates and their supporters who may be being slandered and (if any of them are actually guilty) we owe it to any cheaters to shine some light on their accomplishments as well.
If we plan to export freedom and justice against entrenched politics and religious biases around the world, we'd better make them our priorities at home as well.
-- MarkusQ
I don't understand this attitude at all. Why would you lie to exit pollsters? Do you lie to your doctor when you go in for a checkup? Do you lie to the waiter about what you want to order in a restraunt?
Having accurate exit polls is to the advantage of everyone--everyone, that is, except people trying to rig an election. They are the only ones who benefit from trashing the exit polls. Are you trying to help them?
For that matter, why is it that we are expected to believe not only that lying is rampant, but that it is much, much more common for the sort of people who place high importance on "moral values" to lie? Remember, it's not as if a bunch of Kerry supporters are supposed to have lied and said they supported Bush, is it? It the conserviative, upright rural Bush supporters who think moral values are very important that are supposed to have lied en mass. Does that make sense?
-- MarkusQ
Ever hear of the space shuttle? With slightly smarter external tank management it should be able to do the job nicely.
-- Markus
You want some rational arguments against Bush?
- He spends money with more abandon than any liberal in history
- He lies to the American public
- He lauches unprovoked attacks into third world countries
- He can't figure out how to win them once he's started them
- He shows no respect for the constitution
- He shows no respect for civil liberties
- He uses legal loopholes and questionable logic to rationalize going against pretty much every American Ideal, from "innocent until proven guilty" and "seperation of church and state" to "the right to a fair and speedy trial"
- He routinely places the good of corporations over the good of individuals
- etc., etc.
Before any Democrats reading this get to smug, ask yourself: was Kerry the best you could come up with? "I have a plan" and "Wrong, Wrong, Wrong"? Both parties have any number of sensible, credible people they could have run...and the real problem is they ran Bush and Kerry. The race was so close for so long mostly because neither one of them was worth voting for, except as a way to keep the other from winning.-- MarkusQ
Ah, the joys of getting to the truth in a digital age. I did not provide a link to the final "corrected" 1.5% figure (and qualified it with an IIRC) since I was unable to find one on a quick search. I have not been able to find one with a deeper search, though I have found the raw data on which the corelation with counting method was based. There is kind of a quantry researching things on the internet (boy, is that an understatement!). Things frequently go away / get edited, but apart from google cache, there isn't a good way to retain the information with its provenance. I keep local copys of many interesting thing on my hard drive, but mostly when the information (e.g. technical documentation) is useful without proof of authenticity.
To your other point, it would be interesting to see if there is a correlation between the exit polling descrepancies and where the counties are geographically. This would be a good cross check/sanity chek on the counting method correlation. Another interesting analysis that some people are doing is looking for correlation by candidate. If some people lie to exit pollsters (one theory) why would the supporters of one particular candidate in a specific region (e.g. Bush supporters in Ohio) be much more likely to lie than voters in general?
-- MarkusQ
I still don't buy it. It's like saying that if ATMs gave receipts mugging ten thosand people and demanding ten dollars each would be safer than embezzling a hundred thousand dollars. But even so, the receipt dosn't have to contain a printed record of the vote (just as there's no reason for your ATM receipt to contain your account number and PIN)--all that's needed is a trail linking the votes counted to the votes cast that can be validated by the voter.
As for who's held accountable, once fraud can be demonstrated, there's well defined chain of resposibility (some on made the machines, some one certified them, etc.) and we already have ways for dealing with exactly this sort of problem--which is why it's safe to put your money in the bank, for example.
-- Markus
How about two out of three? We all know about Diebold's connections, and ought to remember ES&S being owned by Republican Senator Chuck Hagel & the fuss when it turned out he'd fibbed about it. How odd. I'm sorry to say that I think (just from looking at the data) there was fraud in the swing states (by the Republicans) and potentially in some congresional races (by the Democrats). Were you hoping that the fraud would be more wide spread or something? Couldn't you use the same logic to claim that nothing bad ever happens? There couldn't be insurgents in Iraq, because there would have to be too many people involved? There couldn't have ever been racism or slavery because it would have involved people with different ways of doing things? For that matter, you could argue that the 9-11 hijackers couldn't have pulled off their plan, etc.
I'll agree that your argument could be used to claim that they would have a hard time going unnoticed, but the very fact that we're all discussing it here already proves that it didn't go unnoticed, so no argument is needed on that count.
-- MarkusQ
Well, that example doesn't help your case much. The sequence was:
- Exit polls (as we all remember) showed a dead heat
- Actual results showed Kerry winning by a wide margin
- The exit polls were retroactively adjusted to match the actuals (just like last tuesday).
- Some people noticed that Kerry's lead over Dean was highly correlated with how the votes were counted:
- When the dust settled Kerry won by (IIRC) 1.5%--close to the exit poll's "dead heat", but by then he'd been spiked by a microphone and no one cared.
So, in short, before you accuse people of wearing a tin-foil hat, you should consider that you may be drinking the kool-aid.VotingTech......Margin
Diebold..........58.1%
ES&S.............35.0%
Hand..............4.7%
-- Markus