On your argument, I don't see how using a certain file format suddenly opens up rights to all of the spreadsheet software. It should be possible to separate the file handling from the spreadsheet itself, say by use of a converter module.
One would reckon you are correct. But no court has found this interpretation to be correct or incorrect, which would be definitive. The question is, who wants to be the first to put it to the test? Yes, this is FUD, but I think it is legitimate FUD until microsoft more completely clarifies such subtleties.
Don't believe the hype. Don't believe the FUD. There are real reasons to complain about M$. This isn't one of them.
Would it be fair to say that a license that has been tested in court is better defined than a license that hasn't? Would it be fair to say that microsoft does not have a history of open specifications? I think both would be fair to say. If microsoft wanted a well defined open license, they should have used one that is already available (GPL, etc.) rather than making up one of their own that is, by any one's definition, ambiguous.
IANAL, but lets look closer at the promise.
If you file, maintain or voluntarily participate in a patent infringement lawsuit against a Microsoft implementation of such Covered Specification, then this personal promise does not apply with respect to any Covered Implementation of the same Covered Specification made or used by you.
To an average English speaker, this clause appears to mean, for example, that if a developer patents an algorithm for (implementation of) a better spreadsheet and such spreadsheet uses the open specification, then they can't sue microsoft if microsoft "borrows" their algorithm without the developer's losing his right to implement the specification.
That clause covers a heck of a lot of territory and effectively gives microsoft right to use any implementation of their specification. Of course the intelligent person's response should be "so don't use the specification if you don't like it"--and that would be the point of TFA and the FUD surrounding it.
Ah, yes. This makes great sense, but the announcement should have read "one of the cores defective", which would be more correct. The word disabled suggests purposeful disabling, which is misleading--but perhaps the announcement was a victim of marketing language chicanery.
I'm not going to certify myself as a lunatic by arguing that we will get the "AI fantasy-future" and the other things you mentioned any time before the sun burns out, but I just read the Heubner paper and his metric for innovation is patently weak. First, he is using a metric based on culling "technology events" from a non-peer reviewed book of written by a third party (Hellemans and Bunch). Second, he is using only one metric instead of several independent metrics that one could compare. Stronger would be to assemble measurements from independent peer reviewed publications, or to collect measurements himself. He would also need to develop a more compelling definition of the term "technological event" (more compelling, that is, than "Hellemans and Bunch said so"), and to compare these now well-defined "technological events" across fields (e.g. manufacturing, engineering, physical sciences, life sciences). We could then assess the validity of this definition by comparing numbers to our expectations. For example, we expect a recent spike in technological events in the biological sciences. Were I a reviewer for this paper, it would have a tough time making it into any journal in its present form. Notice also that Heubner doesn't even have an affiliation with any academic institution.
Obviously, responding to a flamebait post is probably going to get me in karma trouble, but I can't help it. Now, what you have said may have merit, but it is how you said it that got you modded down. Let me propose to put it another way.
Personally, I believe that, just as there are sounds that drive teens off, there are also sounds that can drive adults off. I think such things can be hard-wired. For example, a cat's visual cortex can be heavily patterned by early stimuli. It is well known that early patterning is usually irreversible and manifests itself throughout development.
One would be remiss to assume the same patterning would not function similarly for the auditory centers. I would postulate that your disdain of rap and hip-hop comes from the type of music you listened to as a young child--which was probably what your mom was listening to (maternal patterning). Perhaps you were a child of the 70's, like me, and sacrificed your sense of rhythm for the melodic sounds of the Bee Gees or Paul McCartney. Perhaps you were young during the 80's and your musical sensibilities were destroyed by the likes of Spandau Ballet, Cindy Lauper, and, god forbid, Duran Duran. Now remember, these groups would be the ones your Mom was listening to--you would probably not listen to such groups by your own choice.
The natural question, of course, becomes: what the hell were the hip-hoppers and rap kids' moms listening to?
I don't get the rest of TFA. Its all about how you shouldn't loose emails. But the example they have chosen is good reason why you should loose emails.
Defendent Akbar Butt was glad some common sense saved his saved his rear and is glad to put the episode behind him. In recent weeks, Akbar Butt's life has hit rock bottom. But, to quote Akbar Butt, he "is glad the Justice System was not sitting on the job on this one". In comment, Akbar the Great said that Akbar Butt better watch his ass in the future.
I'm thinking the way to do this is for each photographer to register a longish (64 bit) and unique key with a registration authority. This will be his ID. Then, take this key and, for each image, apply it holographically to the noise of the image, in some scale free and non-deterministic manner, such that, if the image is resampled, one can still reconstruct the key unambiguously. The math doesn't come to me immediately, but it would probably be more-or-less trivial to implement, considering the redundancy available with a 7 megapixel image or similar. I smell a patent.
I got around not having a valid registration of vista: I select the NI (not installed) Mode. This mode comes with every non registered version of vista, but is not well known. The benefits are that you get unlimited access to the web and your files, your computer runs faster, your software choices are unlimited, and you don't have to put up with annoying adware. Since I've switched to NI mode, I've been more productive and had more time to spend on./ because I spend less time dealing with the vista bugs.
Here is one reference. Original references are usually much less alarmist than the stupid news stories created by journalists who don't understand what they are reporting. This is corn ethanol, which is known to be an inefficient source of energy, so the Science article comes as no great surprise--though it does contradict an earlier report in PNAS. The journalism mistakenly groups all biofuels with corn here (unless the article irresponsibly leaves out other references). Independent studies would need to be done for every biofuel source to warrant the sweeping generalizations of the Seattle Times article.
IE is logically causing most of the problem, as that is probably the program most used to go to websites. I'm not randomly hating on windows here: Remember, windows is the most used OS, so the IE browser is the most used program that fetches web pages. This is pure logic and is the same reasoning defenders of windows use when they account for the preponderance of viruses on windows machines.
This argument falls on deaf ears in neocon circles. These people will do anything and give up any personal freedoms (even guns) in the name of religious hatred and xenophobia.
I think the burden of proof is on you regarding this point. I'm curious to see your source for this figure. Just one professional/academic (i.e. not popular press) source for this figure will suffice.
Lets see, according to wikipedia, major outages (like we've seen here) have been in 1929 (earthquake), 2005 (failure), 2007 (no, that was shown to be malicious), 2006 (earthquake), and these in 2008 (which I'm not counting as "historical). Do you have better numbers or are you going to guess its more? So here, I erred in favor of more breaks. Using these numbers and simple binomial statistics, the likelihood that the current breakage rate is by natural causes is 1.1776544649214165e-14. That's so much lower than the Bayesian analysis, its ridiculous. That's simple counting statistics--so the breakage is correlated.
The reason I say this is that I find it hard to believe that of the thousands of undersea cables already laid, that only one of them fails per year. Like I said, in 2006, in the span of two days six of them failed.
This was due to seismic activity. So your hypothesis is seismic activity? We would know about this already, so lets stop using 2006 as some sort of example. It was an outlier with known natural causes, which is tantamount to chance and is included in the 1/365 assumption as a single event. I don't think it should be included as multiple events, because it had a single documented cause: an earthquake.
What if there was a change in the maintenance operations of these cables? What if these cables all share a common engineering flaw that is just now becoming apparent? What if seismic events are involved? What if a biological organism is involved?
Each of these hypotheses can be ruled out. First, we would know about a widespread change in maintenance operations by now, as such change would likely be documented. Second, by what mechanism could different cables in different parts of the world fail all together because of a common engineering flaw? Especially since, in 5 out of 6 cases, the failure was a break. No engineering flaw in history has the same clustering as we have seen here. Name one flaw and I'll find the numbers and show that such clustering, even for that flaw, would be outrageously unlikely. Seismic events would be detected and we would know about them. Biological organism? Even you know you are stretching here.
What if, it turns out, that these cable failures follow seasonal or temporal patterns (e.g. reflecting shipping patterns, undersea currents, or maintenance schedules)?
Any such patterns would have already been established and would be reflected in the historical failure rate, which I am taking, absent of any better data than wikipedia describes, as 1/365 per day. Also, as an example, the first breaks were near Alexandria in the Mediterranean sea, one was in the Arabian sea (Dubai) and one was on the east side of the Indian Ocean (Malaysia). A distance the width of Australia separates the closest of any two of these. We can now rule out correlated changes in shipping and currents.
No, you can't explain it any other way. You use a simple model which basically divides the plethora of possible explanations into "normal yearly failures" and "malice".
The plethora of possible explanations are (1) accidental/random/unintentional--which would be subject to historical patterns or obvious changes affecting the patterns, and (2) intentional, which would by definition be malice. So yes, the model is simple.
Fair would be the least number of clicks from start to finish, as this is what the majority of machines would be running in the world, and so the results would give an estimation of real world performance (not ubergeek world, but real world). If more people chose windows to attack because they thought it would be easiest, then that would also be a reflection of real world. I'd also stipulate that the install CDs would have to checksum with those available from bestbuy (or the politically correct equivalent). Several different linux installs would probably need to be tested as well, as these would vary.
Hmm, I think you are a little too full of yourself.
You will notice that every hint of malicious intent was greeted with "tin foil hat" type references after the 3rd cut. After that cut, I posted and, if you will read, several idiots contested that the cables were getting cut by blind chance, even though probability strongly suggested otherwise. Of course, the critical number has been reached and even public pundits who have to "play it safe" are beginning to publicly doubt the likelihood of chance. We passed the probability for blind luck a long time ago at #3. Now we are up to #5 (6 if you count the one that was "disabled"), and the need to defend my previous posts against idiots who put more faith into their intuition than solid counting statistics has given me an attitude. Now, please end your insulting comments, or I will insult you back.
One cable was cut around Jan 23, before this became widely known. So 5 cables cut, one "disabled".
But yes, if you are implying that only five cuts and one disabling is not correlated, you are hopelessly holding on to the illusion of chance. Give it up. More will be cut, I've predicted it two times in a row, once after the after 4th cut and once after the 5th cut. I don't have time to do Bayesian right now, but, for some raw binomial statistics, the probability that this is caused by chance is: 3.223829097722378e-12. See my previous posts for assumptions.
Waiting for idiots to come along and tell me how I'm full of shit, just so I can post again when the next cable is cut.
1/365 as the rate of failure. He just made this number up. Without any actual data about the failure rates of undersea cables, the entire exercise is pointless.
I didn't make this up. The failure rate of these cables is about 1 per year. This is documented.
0.01 as the rate of failures due to "malicious intent." Again, he just made this up. You cannot do a rational analysis with made-up numbers.
What number would you use? This is a very conservative number, in an attempt to not skew they statistics towards malice, as I will explain below.
Despite the fact that in his discussion he talks about the failure rate of a single undersea line, his variable indicates that all of the thousands of undersea cables are likely to have been tampered with in a hundred year period.
Wrong. The variable assumes that every hundred years, one might expect one and only one of the thousands of cables to be tampered with and disabled because of malice. Not every cable, just one cable if the many thousands in one hundred years.
First of all, he assumes that cable failures are independent events and are randomly distributed. But is there a good reason to do this?
The model is testing whether the failures are independent events.
What if they aren't?
Good question, now you are starting to get the idea.
Computing a probability based upon made-up numbers usually just gives you the answer that you assumed to be true all along.
If you were to come up with some conservative guesses for these numbers, you would see that Bayesian prediction would show a strong likelihood that the cable disabling is somehow correlated. We can't blame weather, ship anchors, or an earthquake. Absence of any natural phenomenon and absence of blind luck, we are left with only a person or people.
It's programmers didn't just program four datapoints of what they thought spam would look like and then call it a day.
You are confusing data points with parameters. Bayesian analysis can be done with three parameters. It can also be done with more parameters. The data we are using is (1) past experience (e.g. failure rate due to accident) and (2) reasonable guesses: (a) estimation of how often one and only one of the thousands of cables will be cut as a result of malice, and (b) estimation of how easy it would be to cut a cable if you really wanted to.
Quite simply, in the absence of objective data, I wouldn't [compute a probability]. Computing a probability based upon made-up numbers usually just gives you the answer that you assumed to be true all along. You might as well follow your intuition, at that point, and not kid yourself with fake probabilities and statistics.
Wrong again. (1) You have computed a probability in your head, but you are writing it off to "intuition". This is how the human brain works. If pressed, you would provide your own "guess" for the likelihood of malice in terms of a number. My guess from your argumentativeness is that number would be low. Whether you like it or not, you have a number in your head. (2) Again, if you make reasonable guesses either way, you will find that these statistics begin to converge on a high likelihood for malice (that is if you can't explain the correlation any other way). That's the point of the exercise. Any type of analysis designed to show or not to show a correlation with incomplete data is going to require some assumptions. Handling these assumptions gracefully and powerfully is why I chose Bayesian analysis. Other methods exist, but they don't do as good of a job. (3) If someone concludes, without calculation, that no correlation (and hence malice) exists, they are still "kidding themselves with fake probabilities and statistics."
The results are not based simply on 5 guessed numbers, which shows your misunderstanding. We have 1 observation per day for the last umpteen years whether or not a cable has been cut or disabled on that day--that's a lot of observations. This is where we get that the historical average is one cable accidentally disabled 1 out of every 365 days. This is counting statistics taken on best knowledge. Surely you don't have a problem with that number. If the media is hyping cable disabling these days, we could then wonder what the media is up to with the hype--because it would be a serious break with its previous reporting pattern.
We can also look at similar systems and see the frequency at which they get sabotaged historically. The most critical number we are using, in terms of the statistics, is assuming that a cable would get maliciously disabled every 100 years. We don't or can't know this number, and so it is part of our assumption. But, if we are to calculate any odds whatsoever, we have to make some assumptions about the system--this is standard practice. I presume you don't like that number. If not, what number would you use? Provide that number and, if it is reasonable, I'll re-run the calculations. My guess is you will still be in disagreement, which would be good for you because that would be your first step towards your Fields medal winning work to discredit Bayesian statistics. You might want to change fields to math if you do this.
Finally, our last number is also an assumption about how easy it is to cut cables. We assume that a properly enabled entity could cut cables at 0.5 per day. Is it easy to cut cables? Well, if they are getting cut accidentally at their current rate, they must be, so someone (or some people) with intent could cut them pretty fast. So this number must support both sides of the argument (If its hard to cut them, then why are they getting cut so fast? If its easy to cut them, then someone could cut them pretty fast.), so I left it at 0.5. If you think cable disabling would be easier or harder, give your number and a valid reason and I'll re-run the analysis.
You could also do binomial statistics to get similar numbers.
My Bayesian calculator says you won't think about these numbers or the type of statistics you would use and will not provide your own carefully constructed analysis and so you are full of shit. Is my Bayesian calculator right yet again?
The analysis says that your 4 shits are part of a pattern, especially if you usually take maybe one (or like most people average of less than one) shit per day. Bayesian analysis would say that there is a better (e.g. 99.9%) probability that you have diarrhea. Yes, it would be good for medical diagnostics--which it is actually used for. Bayes does not say whether the government is involved or whether you have a virulent strain of E. coli in your gut or that you haven't shit in a week and finally took some laxatives. But the analysis would suggest with good accuracy that you have some abnormal and probably pathological shitting going on. Kind of like the cables.
Slashdot Mirror
One would reckon you are correct. But no court has found this interpretation to be correct or incorrect, which would be definitive. The question is, who wants to be the first to put it to the test? Yes, this is FUD, but I think it is legitimate FUD until microsoft more completely clarifies such subtleties.
Would it be fair to say that a license that has been tested in court is better defined than a license that hasn't? Would it be fair to say that microsoft does not have a history of open specifications? I think both would be fair to say. If microsoft wanted a well defined open license, they should have used one that is already available (GPL, etc.) rather than making up one of their own that is, by any one's definition, ambiguous.
IANAL, but lets look closer at the promise.
To an average English speaker, this clause appears to mean, for example, that if a developer patents an algorithm for (implementation of) a better spreadsheet and such spreadsheet uses the open specification, then they can't sue microsoft if microsoft "borrows" their algorithm without the developer's losing his right to implement the specification.
That clause covers a heck of a lot of territory and effectively gives microsoft right to use any implementation of their specification. Of course the intelligent person's response should be "so don't use the specification if you don't like it"--and that would be the point of TFA and the FUD surrounding it.
Ah, yes. This makes great sense, but the announcement should have read "one of the cores defective", which would be more correct. The word disabled suggests purposeful disabling, which is misleading--but perhaps the announcement was a victim of marketing language chicanery.
Enable that other core!
I'm not going to certify myself as a lunatic by arguing that we will get the "AI fantasy-future" and the other things you mentioned any time before the sun burns out, but I just read the Heubner paper and his metric for innovation is patently weak. First, he is using a metric based on culling "technology events" from a non-peer reviewed book of written by a third party (Hellemans and Bunch). Second, he is using only one metric instead of several independent metrics that one could compare. Stronger would be to assemble measurements from independent peer reviewed publications, or to collect measurements himself. He would also need to develop a more compelling definition of the term "technological event" (more compelling, that is, than "Hellemans and Bunch said so"), and to compare these now well-defined "technological events" across fields (e.g. manufacturing, engineering, physical sciences, life sciences). We could then assess the validity of this definition by comparing numbers to our expectations. For example, we expect a recent spike in technological events in the biological sciences. Were I a reviewer for this paper, it would have a tough time making it into any journal in its present form. Notice also that Heubner doesn't even have an affiliation with any academic institution.
Obviously, responding to a flamebait post is probably going to get me in karma trouble, but I can't help it. Now, what you have said may have merit, but it is how you said it that got you modded down. Let me propose to put it another way.
Personally, I believe that, just as there are sounds that drive teens off, there are also sounds that can drive adults off. I think such things can be hard-wired. For example, a cat's visual cortex can be heavily patterned by early stimuli. It is well known that early patterning is usually irreversible and manifests itself throughout development.
One would be remiss to assume the same patterning would not function similarly for the auditory centers. I would postulate that your disdain of rap and hip-hop comes from the type of music you listened to as a young child--which was probably what your mom was listening to (maternal patterning). Perhaps you were a child of the 70's, like me, and sacrificed your sense of rhythm for the melodic sounds of the Bee Gees or Paul McCartney. Perhaps you were young during the 80's and your musical sensibilities were destroyed by the likes of Spandau Ballet, Cindy Lauper, and, god forbid, Duran Duran. Now remember, these groups would be the ones your Mom was listening to--you would probably not listen to such groups by your own choice.
The natural question, of course, becomes: what the hell were the hip-hoppers and rap kids' moms listening to?
I don't get the rest of TFA. Its all about how you shouldn't loose emails. But the example they have chosen is good reason why you should loose emails.
Defendent Akbar Butt was glad some common sense saved his saved his rear and is glad to put the episode behind him. In recent weeks, Akbar Butt's life has hit rock bottom. But, to quote Akbar Butt, he "is glad the Justice System was not sitting on the job on this one". In comment, Akbar the Great said that Akbar Butt better watch his ass in the future.
I'm thinking the way to do this is for each photographer to register a longish (64 bit) and unique key with a registration authority. This will be his ID. Then, take this key and, for each image, apply it holographically to the noise of the image, in some scale free and non-deterministic manner, such that, if the image is resampled, one can still reconstruct the key unambiguously. The math doesn't come to me immediately, but it would probably be more-or-less trivial to implement, considering the redundancy available with a 7 megapixel image or similar. I smell a patent.
Um, try some linguistic analysis. He bought real player 8 years ago (past-tense), but says supporting it is (present tense) dumb.
I got around not having a valid registration of vista: I select the NI (not installed) Mode. This mode comes with every non registered version of vista, but is not well known. The benefits are that you get unlimited access to the web and your files, your computer runs faster, your software choices are unlimited, and you don't have to put up with annoying adware. Since I've switched to NI mode, I've been more productive and had more time to spend on ./ because I spend less time dealing with the vista bugs.
Here is one reference. Original references are usually much less alarmist than the stupid news stories created by journalists who don't understand what they are reporting. This is corn ethanol, which is known to be an inefficient source of energy, so the Science article comes as no great surprise--though it does contradict an earlier report in PNAS. The journalism mistakenly groups all biofuels with corn here (unless the article irresponsibly leaves out other references). Independent studies would need to be done for every biofuel source to warrant the sweeping generalizations of the Seattle Times article.
There should be a law.
The article cites no references nor names any of the "eminent" scientists. I smell political propaganda.
IE is logically causing most of the problem, as that is probably the program most used to go to websites. I'm not randomly hating on windows here: Remember, windows is the most used OS, so the IE browser is the most used program that fetches web pages. This is pure logic and is the same reasoning defenders of windows use when they account for the preponderance of viruses on windows machines.
Macs are better than PCs. Window$ sux!!!! Vote Ron Paul! Know the truth about 911. Jesus saves.
--
LaskoVortex (n) the vortex where you squander your mod points.
This argument falls on deaf ears in neocon circles. These people will do anything and give up any personal freedoms (even guns) in the name of religious hatred and xenophobia.
Well, at least I know the idiot mod who can't recognize humor isn't going to spend that point where it could really do some damage.
Lets see, according to wikipedia, major outages (like we've seen here) have been in 1929 (earthquake), 2005 (failure), 2007 (no, that was shown to be malicious), 2006 (earthquake), and these in 2008 (which I'm not counting as "historical). Do you have better numbers or are you going to guess its more? So here, I erred in favor of more breaks. Using these numbers and simple binomial statistics, the likelihood that the current breakage rate is by natural causes is 1.1776544649214165e-14. That's so much lower than the Bayesian analysis, its ridiculous. That's simple counting statistics--so the breakage is correlated.
This was due to seismic activity. So your hypothesis is seismic activity? We would know about this already, so lets stop using 2006 as some sort of example. It was an outlier with known natural causes, which is tantamount to chance and is included in the 1/365 assumption as a single event. I don't think it should be included as multiple events, because it had a single documented cause: an earthquake.
Each of these hypotheses can be ruled out. First, we would know about a widespread change in maintenance operations by now, as such change would likely be documented. Second, by what mechanism could different cables in different parts of the world fail all together because of a common engineering flaw? Especially since, in 5 out of 6 cases, the failure was a break. No engineering flaw in history has the same clustering as we have seen here. Name one flaw and I'll find the numbers and show that such clustering, even for that flaw, would be outrageously unlikely. Seismic events would be detected and we would know about them. Biological organism? Even you know you are stretching here.
Any such patterns would have already been established and would be reflected in the historical failure rate, which I am taking, absent of any better data than wikipedia describes, as 1/365 per day. Also, as an example, the first breaks were near Alexandria in the Mediterranean sea, one was in the Arabian sea (Dubai) and one was on the east side of the Indian Ocean (Malaysia). A distance the width of Australia separates the closest of any two of these. We can now rule out correlated changes in shipping and currents.
The plethora of possible explanations are (1) accidental/random/unintentional--which would be subject to historical patterns or obvious changes affecting the patterns, and (2) intentional, which would by definition be malice. So yes, the model is simple.
Fair would be the least number of clicks from start to finish, as this is what the majority of machines would be running in the world, and so the results would give an estimation of real world performance (not ubergeek world, but real world). If more people chose windows to attack because they thought it would be easiest, then that would also be a reflection of real world. I'd also stipulate that the install CDs would have to checksum with those available from bestbuy (or the politically correct equivalent). Several different linux installs would probably need to be tested as well, as these would vary.
You will notice that every hint of malicious intent was greeted with "tin foil hat" type references after the 3rd cut. After that cut, I posted and, if you will read, several idiots contested that the cables were getting cut by blind chance, even though probability strongly suggested otherwise. Of course, the critical number has been reached and even public pundits who have to "play it safe" are beginning to publicly doubt the likelihood of chance. We passed the probability for blind luck a long time ago at #3. Now we are up to #5 (6 if you count the one that was "disabled"), and the need to defend my previous posts against idiots who put more faith into their intuition than solid counting statistics has given me an attitude. Now, please end your insulting comments, or I will insult you back.
One cable was cut around Jan 23, before this became widely known. So 5 cables cut, one "disabled".
But yes, if you are implying that only five cuts and one disabling is not correlated, you are hopelessly holding on to the illusion of chance. Give it up. More will be cut, I've predicted it two times in a row, once after the after 4th cut and once after the 5th cut. I don't have time to do Bayesian right now, but, for some raw binomial statistics, the probability that this is caused by chance is: 3.223829097722378e-12. See my previous posts for assumptions.
Waiting for idiots to come along and tell me how I'm full of shit, just so I can post again when the next cable is cut.
Now that we are getting the middle east out of the picture.
I didn't make this up. The failure rate of these cables is about 1 per year. This is documented.
What number would you use? This is a very conservative number, in an attempt to not skew they statistics towards malice, as I will explain below.
Wrong. The variable assumes that every hundred years, one might expect one and only one of the thousands of cables to be tampered with and disabled because of malice. Not every cable, just one cable if the many thousands in one hundred years.
The model is testing whether the failures are independent events.
Good question, now you are starting to get the idea.
If you were to come up with some conservative guesses for these numbers, you would see that Bayesian prediction would show a strong likelihood that the cable disabling is somehow correlated. We can't blame weather, ship anchors, or an earthquake. Absence of any natural phenomenon and absence of blind luck, we are left with only a person or people.
You are confusing data points with parameters. Bayesian analysis can be done with three parameters. It can also be done with more parameters. The data we are using is (1) past experience (e.g. failure rate due to accident) and (2) reasonable guesses: (a) estimation of how often one and only one of the thousands of cables will be cut as a result of malice, and (b) estimation of how easy it would be to cut a cable if you really wanted to.
Wrong again. (1) You have computed a probability in your head, but you are writing it off to "intuition". This is how the human brain works. If pressed, you would provide your own "guess" for the likelihood of malice in terms of a number. My guess from your argumentativeness is that number would be low. Whether you like it or not, you have a number in your head. (2) Again, if you make reasonable guesses either way, you will find that these statistics begin to converge on a high likelihood for malice (that is if you can't explain the correlation any other way). That's the point of the exercise. Any type of analysis designed to show or not to show a correlation with incomplete data is going to require some assumptions. Handling these assumptions gracefully and powerfully is why I chose Bayesian analysis. Other methods exist, but they don't do as good of a job. (3) If someone concludes, without calculation, that no correlation (and hence malice) exists, they are still "kidding themselves with fake probabilities and statistics."
Now, if you p
The results are not based simply on 5 guessed numbers, which shows your misunderstanding. We have 1 observation per day for the last umpteen years whether or not a cable has been cut or disabled on that day--that's a lot of observations. This is where we get that the historical average is one cable accidentally disabled 1 out of every 365 days. This is counting statistics taken on best knowledge. Surely you don't have a problem with that number. If the media is hyping cable disabling these days, we could then wonder what the media is up to with the hype--because it would be a serious break with its previous reporting pattern.
We can also look at similar systems and see the frequency at which they get sabotaged historically. The most critical number we are using, in terms of the statistics, is assuming that a cable would get maliciously disabled every 100 years. We don't or can't know this number, and so it is part of our assumption. But, if we are to calculate any odds whatsoever, we have to make some assumptions about the system--this is standard practice. I presume you don't like that number. If not, what number would you use? Provide that number and, if it is reasonable, I'll re-run the calculations. My guess is you will still be in disagreement, which would be good for you because that would be your first step towards your Fields medal winning work to discredit Bayesian statistics. You might want to change fields to math if you do this.
Finally, our last number is also an assumption about how easy it is to cut cables. We assume that a properly enabled entity could cut cables at 0.5 per day. Is it easy to cut cables? Well, if they are getting cut accidentally at their current rate, they must be, so someone (or some people) with intent could cut them pretty fast. So this number must support both sides of the argument (If its hard to cut them, then why are they getting cut so fast? If its easy to cut them, then someone could cut them pretty fast.), so I left it at 0.5. If you think cable disabling would be easier or harder, give your number and a valid reason and I'll re-run the analysis.
You could also do binomial statistics to get similar numbers.
My Bayesian calculator says you won't think about these numbers or the type of statistics you would use and will not provide your own carefully constructed analysis and so you are full of shit. Is my Bayesian calculator right yet again?
Still waiting on those numbers...
The analysis says that your 4 shits are part of a pattern, especially if you usually take maybe one (or like most people average of less than one) shit per day. Bayesian analysis would say that there is a better (e.g. 99.9%) probability that you have diarrhea. Yes, it would be good for medical diagnostics--which it is actually used for. Bayes does not say whether the government is involved or whether you have a virulent strain of E. coli in your gut or that you haven't shit in a week and finally took some laxatives. But the analysis would suggest with good accuracy that you have some abnormal and probably pathological shitting going on. Kind of like the cables.