Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

General Irish politics discussion thread

14546485051126

Comments

  • Registered Users Posts: 17,568 ✭✭✭✭VinLieger


    We don't need e-voting there's no benefits we are missing with our current system.

    Also non of my concerns are trumpian, the realities and vulnerabilities of electronic voting via machine, software or whatever have existed since well before trump adopted it as a method to excuse his inadequacies.

    LOL that you freely admit you again know nothing about blockchain therefore your argument amounts to once again "others do it".

    You still have yet to provide a valid reason to make any of the changes you seem to be suggesting.



  • Registered Users Posts: 13,679 ✭✭✭✭markodaly


    Anything that facilitates the breadth of ways people can vote is a good thing.

    I've made this point and others clear many times.

    I've also made reference to the fact that we could and should have the ability to count votes quicker as well.



  • Registered Users Posts: 13,679 ✭✭✭✭markodaly


    We don't need e-voting there's no benefits we are missing with our current system.

    No benefit at all?

    Quite a statement.


    Also non of my concerns are trumpian, the realities and vulnerabilities of electronic voting via machine, software or whatever have existed since well before trump adopted it as a method to excuse his inadequacies.

    Yet you repeat and parrot them here as a way to argue against e-voting. Its amazing you see to argue both against Trump but copyright his words to argue in effect the same thing.



  • Registered Users Posts: 6,362 ✭✭✭Brussels Sprout


    Personally I believe that the downside of one or two constituencies taking more than 3-4 days to count every election cycle is more than made up for by the transparency of the process. The chain of custody of the ballots is clear and in a lot of constituencies the count centres are open to the public (they're all open to the candidates and journalists). The cost and duration of the count is worth it for that peace of mind.

    Besides, Electronic Voting is up there with Water Charges. Something that was rushed through in a half-arsed manner and then became a political football when it all ended up in tears. It's now a settled matter in this country so really it's not even worth arguing over - the status quo is here to stay.



  • Registered Users Posts: 27,981 ✭✭✭✭AndrewJRenko


    Why is that 'a good thing'? What actual benefit would arise?

    The UK Electoral Commission tested a range of voting methods about twenty years ago, using mobile phones, teletext, internet and more - and found none of them has a significant effect on electoral turnout.

    What benefit would arise from 'breadth of ways people can vote'?

    What benefit would arise from counting votes quicker?



  • Advertisement
  • Registered Users Posts: 33,650 ✭✭✭✭Hotblack Desiato
    Restaurant at the End of the Universe


    The provision is inadequate both in terms of when one can have an abortion, and where. The number of people who choose to avail of the service is entirely irrelevant to these issues which stem from weak legislation and the failure of the HSE to ensure that holy joe consultants are not permitted to block the provision of legal services in their hospitals.

    A teacher who refused to teach religion on conscience grounds would be sacked!

    It took a while but I don't mind. How does my body look in this light?



  • Registered Users Posts: 33,650 ✭✭✭✭Hotblack Desiato
    Restaurant at the End of the Universe


    As a person who works in IT security, there is not a hope in hell I would agree with any form of computerised voting.

    It is vital that the vote is not only counted correctly but seen to be counted correctly. It is certainly not vital that the vote be counted quickly.

    It took a while but I don't mind. How does my body look in this light?



  • Registered Users Posts: 17,568 ✭✭✭✭VinLieger


    This is pathetic trolling at this stage you arent adding anything to the discussion and ignore any attempts by others to engage you while continuously trying to derail the thread



  • Moderators, Science, Health & Environment Moderators Posts: 19,254 Mod ✭✭✭✭Sam Russell


    I am utterly against electronic voting.

    However, I have outlined a secure system of copying the current ballots on to a computer file using character recognition as used by An Post to read those ballots into a file. If for any reason the ballot is unreadable, dubious, unclear, spoilt, or any reason, the image of the ballot paper is diverted to a human who makes sense of it.

    Now, at this stage, we have a computer file that derives from images of the current ballots. All have been checked to be valid by the reader, or else they are checked by a human eye.

    Now to be sure to be sure, each ballot carries a serial number to identify the ballot paper (not the voter). Now a further check - a random selection of ballots are checked against the computer file to see they are correctly copied. So now, the returning officer (and all interested individuals) is certain that the computer file is a true copy of all the votes cast.

    So now the count can start. A validated computer program will now crunch the numbers to decide the quota, and the transfers for each stage. Now how this is handled is theatre, but the result should be accurate. The algorithm either apes the current system or uses the purest form of STV for transfers using fractions of a vote.

    At the end of the day, the original ballots can be counted in the old way, just to verify the result. They have not been touched after scanning.

    I am very much in favour of electric counting.

    Computerised voting is opening up lots of ways of confusing voters, allowing too many factors to distort the result, and would lack a proper audit trail.



  • Registered Users Posts: 23,716 ✭✭✭✭Larbre34


    Our current system of observable hand counting is about as transparent as you can get.

    If the ultimate objective is faith in the system and the outcomes, then there is no appreciable benefit from switching to e-voting now.

    Especially in this post-truth world of paranoia, accusation as a political weapon and conspiracy mongering.

    If it takes 24 hours to get a verified outcome of a poll that is going to provide a 5 years mandate, so be it. Far better to be accurate than be fast. If it ain't broke, let's not fix it.



  • Advertisement
  • Moderators, Science, Health & Environment Moderators Posts: 19,254 Mod ✭✭✭✭Sam Russell


    Didn't one of the counts in the RDS go on for weeks, with counters trapped in the building (free the RDS 22). I think it was Green Party vs Progressive Democrats - (Gormley vs Michael McDowell) - forget who won in the end.



  • Registered Users Posts: 33,650 ✭✭✭✭Hotblack Desiato
    Restaurant at the End of the Universe


    That changes nothing.

    You cannot prove the computer is secure.

    You cannot prove the count is correct without manually counting every vote.

    What is the point?

    It took a while but I don't mind. How does my body look in this light?



  • Registered Users Posts: 23,716 ✭✭✭✭Larbre34


    No Counters went home every night, as they always do. I know because I did it for 20 years.

    If I recall correctly Byrne / Briscoe outstripped McDowell / Gormley for longevity of a count and ultimately went to Court for a decision. But the right decision was ultimately determined and those marathon situations are so rare we can remember them by name, even 30 years on.



  • Moderators, Science, Health & Environment Moderators Posts: 19,254 Mod ✭✭✭✭Sam Russell


    First, you can prove that the actual ballots have been copied correctly to the computer file. This can be done by randomly checking a representative sample because every ballot has a serial number, and the ballots are kept in a manner that renders then easy to find. Now if you cannot accept that this possible, then manual counting has been proved to be fallible with bundles of votes are found to be in the wrong candidate pile.

    Now the computer file created above has not been manipulated in any way, nor has a single vote been counted. It simply is a version of all the ballot boxes.

    So the votes are counted to get the total votes cast, and the quota is determined. How can that be wrong? All votes were verified in the machine reading phase, so any errors or disputed votes have been dealt with.

    Now this is not electronic voting. Votes are cast exactly as they have always been cast.

    Now the counting can proceed, and the algorithm would have checked and tested to be correct for our unusual voting system. If necessary, two or three different programs could be used and see if they give the same result. Would that satisfy you?

    Universities would be pleased to be part of such an exercise - particularly if the raw, anonymised, data is given to them for political analysis.

    Post edited by Sam Russell on


  • Registered Users Posts: 33,650 ✭✭✭✭Hotblack Desiato
    Restaurant at the End of the Universe


    A random sample is not a full check.

    Any number of programs would not satisfy me, no. Votes are fed into a black box and a result comes out. That cannot be trusted.

    The nonsense of having a manual count as a backup just underlines how pointless and useless electronic counting is in the first place.

    It took a while but I don't mind. How does my body look in this light?



  • Moderators, Science, Health & Environment Moderators Posts: 19,254 Mod ✭✭✭✭Sam Russell


    The whole manual counting system depends on randomisation. All votes in single constituency are tossed onto a table and mixed so that there is no bias in any amount of votes taken out for a surplus. Otherwise, if a candidate gets over a quota on the first count, votes taken from a single box might be different from boxes taken from a different box.

    Let us take this slowly.

    The ballot is filled in by the voter exactly as currently happens. Have you a problem with that?

    Every ballot is passed through a reader where the writing on the ballot is read, and if it makes sense, it is transferred to a computer file, where the record consists of the ballot serial number, a location code so the ballot can be found easily for any reason, and the entry for each candidate - being the preference number given by the voter. Have you a problem with that?

    Now, if a ballot cannot be read or the reading is not consistent for any reason, the image passes to a human who either refers it to a scrutineer and the ballot is pulled for manual checking, or corrects the difficulty - for example a digit looks like a 5 or an 8 but there is already a clear 5 elsewhere and there is a 7 on the paper, and all numbers up to 7 but no 8 so it is obviously an 8. Have you a problem with that?

    Now, we take a random sample of ballots, identified for checking, locate the ballot and compare it with the data in the file. Now the number checked could be 5%, 1% or 0.1% or whatever the returning officer decides. If no errors are detected, then the returning officer declares the data file correct. If significant errors are detected, they do a full manual count. Have you a problem with that?

    So now we are at your problem. You do not trust computer programs. Well, the algorithm is checked with small datasets, and checked manually to see it gets the right result. Then larger data sets, and then accepted as correct. It is possible that separate teams could devise different programs and the data fed into each one as a test and they should come out with the same answer. Have you a problem with that?

    Yet you are happy with tons of paper ballots, tied with elastic bands into bundles, being moved from one part of the count in barrows, to another part where they might end up in the wrong candidates pile. This has been discovered in recounts. Have you a problem with that?

    I assume, you get your pay from a computer program, with pay rates calculated correctly with tax deducted correctly, all by a black box. I assume your money is paid into a bank who uses a black box to look after it. I wonder if you have a credit card that is managed by a black box. Have you a problem with that?

    So in everyday life, much of your life is handled by a black box, but not when you vote. Have you a problem with that?



  • Posts: 0 [Deleted User]


    Electronic voting is fast but a nightmare.

    You have an airgapped/dumb terminal and you srill need to make sure that it's tamperproof as all votes for that terminal are stored at that terminal until removal. All local votes vulnerable.


    If votes are stored away from terminal then even more critical that the terminal is secure (constant upgrades/replacement). You now also need to ensure network security. All votes vulnerable


    Currently you need:

    Pencil

    Paper

    Stamp

    Publicly visible box

    Burley men

    Public counting


    High tech isn't always better.



  • Registered Users Posts: 33,650 ✭✭✭✭Hotblack Desiato
    Restaurant at the End of the Universe


    It's totally irrelevant that in PR-STV the result may depend on which votes are chosen as a surplus, etc. The computer could be programmed to replicate that behaviour, or calculate exact (possibly fractional) votes, etc. That's not a reason to change our counting system, everyone currently accepts there is a certain but small amount of 'noise' in the results.

    Comparing voting to salary or banking is rather more worrying. This point was often brought up during the e-voting controversy and tbh it reveals a fundamental lack of understanding of the issues.

    With payroll, banking, even autopilots, the inputs are relatively small in number and known. With voting the inputs are huge in number and cannot be 'known' as each vote must remain confidential. With the first type of system it's easy to demonstrate that you have obtained the correct result - your salary is either correct or not, your bank account balances, your plane doesn't crash.

    With electronic counting the only way we can know if the result is correct is if a full manual count is performed. So, once again, I ask what is the point? What is the benefit for the substantial cost incurred? Getting a (non-verified, until we do a full manual count) result a day earlier when we only have elections every five years or so is utterly pointless.

    It took a while but I don't mind. How does my body look in this light?



  • Moderators, Science, Health & Environment Moderators Posts: 19,254 Mod ✭✭✭✭Sam Russell


    You do not like statistics and randomized sampling.

    Look, medical research develops new treatments that point to a cure by looking at statistics and sampling.

    A major metric is the survival rates for cancer patients on the new treatment vs those on a placebo or the normal treatment. If survival rates extend with the new treatment, then it is considered a successful treatment. So going from a survival of 4 years to 4.3 years means the treatment will be given. Now if I am a patient, my interest is not getting an extra 4 months , but being cured. (If I was on the treatment, because that is not how it works, because for some, they do not survive as long as the would have, while others get cured - it is a bit of a lottery which result the patient gets.) Now, if there is a lottery effect on this new treatment, I would point out that the patient has only one ticket in that lottery.

    The reason for going for electronic counting is to reduce the element of errors affecting the result. If there is doubt in the result, the calculation can be run again with a different, but tested, algorithm. If the results are different, then the count is done manually.

    Incidentally, by being quicker, it also costs less to run, but that is incidental and should not be the reason to go for this type of counting.

    Electronic voting has many, many, pitfalls, and I for one would not like it to come in here. [Audit trails being the most obvious.]

    A small number of ballots could change the result, depending whether they happen to be included in a particular part of the count or not. Dick Spring won his seat in the 1987 election by four votes.



  • Registered Users Posts: 4,351 ✭✭✭political analyst


    I agree. Bertie Ahern gave no thought to the effect that e-voting would have on the integrity of the electoral process, i.e. security, the effect of the sudden announcement of the result on Nora Owen in a constituency in 2002 when e-voting took place on a trial basis.



  • Advertisement
  • Moderators, Science, Health & Environment Moderators Posts: 19,254 Mod ✭✭✭✭Sam Russell


    If we were to go with electronic counting, then the Presidential election should be the first, with EU elections next or perhaps local elections.

    If they worked, the Dail elections could be tested.



  • Registered Users Posts: 27,981 ✭✭✭✭AndrewJRenko


    The difference between electronic voting and all the other aspects of your life that are managed by a computer is the conflicting requirements for auditability AND anonymity. For your payroll, your credit card, and all the other aspects of your life handled by a black box, those transactions are intrinsically tied to you, albeit with different levels of confidentiality.

    Your vote CANNOT be tied to you. If your vote is tied to you, you break the fundamental secrecy of the ballot. You open up the system to vote buying and selling and duress voting. There's good reason for the secret ballot. There's no good reason to put it at risk.



  • Moderators, Science, Health & Environment Moderators Posts: 19,254 Mod ✭✭✭✭Sam Russell


    I am not considering electronic VOTING. You are right that eVoting has a lot of problems.

    I am talking about electronic COUNTING which is completely different in that the data is derived from the manual ballots. If all else fails, they can be counted as now. All that happens is that those ballots are scanned (or even input by a person) to create an accurate version in a computer file. The ballots are not identified back to a voter in any way (more than at present).

    The computer mimics the manual count, working from this datafile that is a copy of the ballots.



  • Posts: 0 [Deleted User]


    You're ralking about counting using ICR? Not something I'd have too much issue with as long as we have a publicly vetted tally



  • Moderators, Science, Health & Environment Moderators Posts: 19,254 Mod ✭✭✭✭Sam Russell


    The transfer of the data could be done by a human. OCR is accurate and used by An Post with no apparent problems. It would be easy to verify, as all that the transfer achieves is to transfer the ballot paper to a computer file. No counting is done at this stage.

    The ballot papers are always available to be checked an verified.



  • Registered Users Posts: 27,981 ✭✭✭✭AndrewJRenko




  • Moderators, Science, Health & Environment Moderators Posts: 19,254 Mod ✭✭✭✭Sam Russell


    Accuracy, reliability, speed, lower cost.

    The biggest benefit would be in the Euro elections because the ballot papers are getting to be about 50 cm long and are difficult to manage. Not only are the ballot papers huge, there thousands and thousands of them. Just handling them is a real problem. Plus the count goes on forever.

    In the case of the Euros, there were over 750,000 votes in the South constituency. Now the three quarters of a million votes had to be counted in one location and moved into piles for each of the 23 candidates. There were twenty counts before the result could be determined.

    With the system I have outlined, the votes could be scanned in many locations, and the data transferred to one location for the whole country. Just the logistics of running such a large poll with so many candidates for just one constituency is a huge task.



  • Registered Users Posts: 1,546 ✭✭✭rock22


    @Sam Russell wrote

    "Now, if a ballot cannot be read or the reading is not consistent for any reason, the image passes to a human who either refers it to a scrutineer and the ballot is pulled for manual checking, or corrects the difficulty - for example a digit looks like a 5 or an 8 but there is already a clear 5 elsewhere and there is a 7 on the paper, and all numbers up to 7 but no 8 so it is obviously an 8. Have you a problem with that?"

    Who would do this?

    Would it be legal?

    As i understand the present system, the candidates agents must examine the disputed ballot and all agree ? So this could only be done at the count centre and therefore the ballot papers would have to be brought there for checking.

    And "

    Now, we take a random sample of ballots, identified for checking, locate the ballot and compare it with the data in the file. Now the number checked could be 5%, 1% or 0.1% or whatever the returning officer decides. If no errors are detected, then the returning officer declares the data file correct. If significant errors are detected, they do a full manual count. Have you a problem with that?"

    Who does this?

    On what basis would the returning officer decide to check 0.1% or 5% ?



  • Registered Users Posts: 6,362 ✭✭✭Brussels Sprout


    Accuracy, reliability, speed, lower cost


    Is there any evidence that the current system is inaccurate or unreliable?

    As for lower cost....eventually, perhaps, but I'm sure after some exorbitant consultation fees and then actually purchasing the software and buying and maintaining the hardware the cost would be extortionately high (that's typically the way it goes with these kinds of government projects) with a surprisingly long payback period.

    Speed...yeah that one Euros ballot was physically difficult to count in 2019. Something like that could be tackled with legislation about what the minimum required amount of information is required on the ballot paper. For example, after the recent spate of abuse of elected representatives we should definitely remove their addresses from the ballot. As for election counts in general, almost all of them are typically counted within 3 days which is perfectly acceptable.

    The only real benefit I can see in switching to electronic counting, from the current system, is that is that they could calculate transfers in a more accurate/proportionate manner, rather than using the elements of randomisation that we have in our current process. That's not really enough of a reason to change the system though.



  • Advertisement
  • Registered Users Posts: 33,650 ✭✭✭✭Hotblack Desiato
    Restaurant at the End of the Universe


    Maybe the real issue is raising the bar a bit to keep the head cases out - bring back deposits?

    It took a while but I don't mind. How does my body look in this light?



Advertisement