You're counting a whole lot of zero-acre fires. If you look at the damage caused, target shooting accounts for a good deal more than 10%. Also, target shooters make up a rather small proportion of the population and cause a vastly disproportionate number of fires.
Any target shooting outside of a gun range during a red flag warning shows a lack of common sense, and trying to excuse these people's rampant irresponsibility by saying other people sometimes act irresponsibly too shows you're the one with the political agenda.
I don't have any idea who you think you're replying to. I'm not claiming gun ownership should be outlawed, and I don't see anybody who's making that claim.
You admit "Utah needs to change their local laws concerning the time and place it is appropriate to shoot" and that's precisely what I'm saying.
Your claim that these people did nothing illegal runs afoul of the reckless burning ordinance; this was a class A misdemeanor. But that's not enough to dissuade people from destroying land and endangering others' lives, because people are too stubborn to believe their irresponsible actions really cause any risk of fire, even when 19 fires had already been started by shooters in Utah this year.
Target shooting on public land during a red flag warning should be illegal, and it's farcical that the Legislature has not only refused to put in place reasonable regulations but has barred counties and municipalities from doing so.
So destroying thousands of acres of public and private land, costing the state millions of dollars in firefighting costs, risking the lives of firefighters, and causing >9000 people to evacuate their homes and businesses doesn't really matter as long as nobody got killed and no homes were destroyed?
Even if the target shooters had the money to pay the firefighting costs (extremely unlikely), the burned lands, the threat to others' lives and property, and the loss of >9000 people's time would be worth a criminal conviction.
There have been around a dozen fires started by target shooters in Utah this year, and some were larger than this; this one gets the news because it was closer to homes.
Years ago the legislature seized power to keep counties and municipalities from enforcing anything related to shooting, and they've repealed any and all restrictions on gun use they could find. They too are responsible for the fires.
Google announced their initiative to save Algol 68 a couple years ago: they revised it a little and called it Go, in hopes that it would thereby become the Next Big Thing in software development.
It's not "perhaps" I'm "now" limiting my argument to 48fps vs higher rates; this whole conversation started because I was defending 48fps against someone who was complaining it wasn't 120. Neither is this just "slightly less outrageous" than saying 24fps is enough; anyone can distinguish 24fps from higher rates with sufficiently quick full-frame motion, but that simply isn't the case with 48fps.
The BBC "study" you cited is an unscientific failure. It involved no blind testing; the viewers knew at all times what framerate they were looking at, meaning the results are almost sure to be dominated by confirmation bias. Also, they spent a good deal of time and effort looking at individual frames, looking at slow-motion playback, and complaining about the loss of detail in an individual frame caused by having non-infinite shutter speeds, none of which has any bearing on whether there's a difference in perceptible detail when played back at normal speed. On top of that, they used naive averaging of frames from their high-speed camera to simulate the slower speeds, which is a very poor approximation; their artifical low-frame-rate videos definitely looked worse than real 50fps capture. (You have to have a properly weighted average to get anything close to the right results; see this paper for some detail.)
Look at my reply to another poster regarding the other effects which come into play when you're considering the framerate of games.
How the devil did you and several others get the impression I was saying 48fps is overkill? I was defending 48fps against somebody who was complaining that it wasn't 120fps. 24fps is very obviously not sufficient for rapid full-frame motions to appear smooth. Just moving to 30fps makes a big difference (the variable we really should be using here is frame delay, so the objective difference between 24fps and 30fps is just as big as the difference between 60fps and 120fps, i.e. 8.3 ms), but that may still appear slightly jerky in some situations. As I said, by the time you reach 48fps with proper motion blur the studies simply don't show a perceived quality difference between that and higher framerates.
I've seen the 100fps.com page before, and while it has some obvious truths it's also got a lot of half-truths, distortions, and absurd speculations. It's not really a good source for learning about perception.
The primary reason a 48fps average may not be good enough for gaming is because of variability. In a 48fps movie, every single frame comes 20.83ms after the one before it. If a game has an average of 61 FPS, it could be that sixty frame times are 15ms and the sixty-first is a very noticeable and jarring 100ms. Hardware reviewers, starting with Tech Report, are starting to catch on to the fact that the longest frame times matter much more than the average.
Two other reasons why a framerate where movies are smooth may not suffice for gaming:
Normal 3d rendered frames are like using a camera with an infinitely fast shutter. The difference between the resulting sudden motions and the natural blur in frames that have had exposure times comparable to their frame display times is the same thing as the difference between "jaggies" and smoothly antialiased edges in a single frame, except in the time domain rather than the spatial. Artificial motion blur effects try to improve on that, but often it's easier to just increase framerate (just as increasing the resolution makes the "jaggies" in a single frame less obvious).
The relationship between Input lag and frame delay in a game is somewhat complex. It would of course be possible to have a thousand or more frames per second and still have a full second of input lag, and it's possible to have input lag low enough for twitch FPSes with less than 60fps. But if a game's engine design, buffering, and the hardware input and output pathways were such that input lag is 3*frame time + 50ms, you would notice improvements in input lag from having ridiculously high frame rates well beyond the point of smooth motion.
Here's the study showing little difference between 30fps and 60fps for a first person shooter. Note the dramatic differences up to 15fps, where we begin to perceive many things as motion rather than individual frames, the still-quite-noticeable improvement in moving to a 30fps rate, where quick full-frame motions are starting to appear smooth, and the much smaller change (esp. the almost-zero perceived quality change) between 30fps and 60fps. Since the frame time difference between 30fps and 48fps is 12.5ms and the difference between 48fps and 60fps is only 4.17ms, I sincerely doubt their test setup would have showed any difference between 48fps and any higher rate (even 240+). However, their test setup was different from most games in one or maybe two of the three points I mentioned above. The frame rate limitations in the paper were artificial- the computer was capable of producing a much much higher framerate- so the frame times were probably almost exactly consistent. It's also possible that their mechanism for limiting framerate didn't give as strong a link between framerate and input lag as most games have (you'd have to look at the code to be sure).
You may think you can see it, but unless you've shown you can tell the difference in a double-blind study, that only tells us about your psychological biases rather that about your perception. (Saying you're biased is not an insult here; all of us are subject to quite a number of well-known psychological biases, and if anybody ever wants find the truth in these kinds of things they have to find ways to avoid letting these biases control their conclusions.)
All kinds of people think they can see 120fps, can hear the difference between normal audio cables and $10,000 gold-plated audio cables, etc. They can't. To believe such claims when they directly contradict all science is backwards superstition. If you're really so sure you can see the difference between 48fps and 120fps with proper motion blur, I'd bet the Randi Foundation would like to offer you their one million dollar Paranormal Challenge prize if you can prove it.
It's a fuzzy mess and the brain will not piece it together.
You obviously have no understanding of how human perception works. Go educate yourself about the response time of rods and cones and don't go around mouthing off at people just because they're better informed.
The refresh rate on a CRT and the frame rate of a movie have almost nothing in common. Your eyes can tell, somewhat painfully, that they're only getting 60 strobes of light per second with darkness the rest of the time. They can't tell that they're getting only 48 frame changes per second, especially with normal motion blur.
No study I'm aware of has ever shown people successfully distinguishing >40fps with proper motion blur from any higher framerate. Indeed, I've seen studies that showed very little difference in perceived quality between 30fps non-blurred frames and much higher framerates.
People on/. keep acting as though "48fps is good enough" is like "640k should be good enough." There's a vast difference here; human perception has very definite limitations.
I imagine having a spaceport wouldn't be all that different from having an airport, though an airport sees constant use and a spaceport would therefore seem to be less of a disturbance.
We've made some attempts to include Russia (mostly as monitors/operators at missile defense sites in Eastern Europe). Those attempts have generally been rebuffed. Plus, Russia is getting various benefits from their saber-rattling, and their behavior generally indicates that they'd be pleased as punch to see Iranian missiles falling on Israel and North Korean missiles falling on South Korea. They've consistently vetoed just about every UN effort to deal with peace and stability issues since Putin came in power. You're right that they would not be benefitted by attacks on Europe. But they simply don't believe that will happen.
I wasn't quoting myself "as a source" - instead of just repeating myself I was acknowledging that I've said it before, and I linked to my previous post because it included more details for people who were interested. For sources, most of what Rob Weir says on the AOO incubator dev list will show my point. I haven't followed it in the last few months but when I was reading it frequently stuff like this was going on all the time
The LibO download size may look bloated, but that's because their default download includes all the languages rather than having separate installers for each language. I switched to LibO 3.5 recently and my install uses ~75MB less space than my OpenOffice 3.3 install did.
I'd like to see AOO succeed. But its leadership dooms it. As I've said before:
Rob Weir, who is basically running the show and who seems like a perfectly reasonable person from his blog, acts like a caustic, sarcastic, and poorly socialized adolescent in communicating with other developers. He's alienating people right and left. People have tried to get him to stop, but he either ignores it or just acts like it's those he's offended who are to blame for any unpleasantness.
He's not the only one either. Few people who aren't on the IBM payroll want to contribute to a project with that kind of leadership. People from the open source community in general and from the LibO camp in particular are reluctant to do anything to cooperate with Weir and co. By the time AOO actually gets a release out it will likely be too late to revitalize any interest in the project.
Another reason not to bother talking about this now: given the state of the Indian justice system, it may well happen that all parties involved are deceased before this actually gets a court date.
Wrong again; you're the one who needs to look at the numbers. Over half of the deaths in the Hiroshima and Nagasaki were basically immediate and were ascribable only to the scale of the explosion and not to any radiation effects. Even among the later deaths conventional burns were a primary cause. Less than 20% of all the deaths were caused by radiation sickness etc.
Even ignoring all the non-radiation deaths, the difference between the fallout from a nuclear bomb and any possible contamination from a nuclear plant is tremendous. While it's easy to find lots of politically-motivated exaggerated numbers about Chernobyl- which due to horrible design failures and a deliberate simulation of disaster conditions was much much worse than any disaster at a sanely designed reactor will ever be- realistic estimates of the number of total Chernobyl-related deaths (including long-term increased cancer risk deaths over 26 years) are under 5000, while the atomic bombings in Japan caused ~200000 deaths within a period of a few months, ~40000 of which were due to radiation.
Trying to claim that nuclear reactor disasters are at all comparable to nuclear bombings is utterly ridiculous.
If you think the figures from the atomic bomb have any relevance to the present discussion you have no understanding of physics.
There is almost as much of a difference between a nuclear bomb and a nuclear power plant as there is between a nuclear power plant and the nuclear family. Same word, completely different implications.
Lots of people out there- this guy, Nassim Taleb, tons of economic pundits (usually not real economists), etc- are ready to blame probability theory for every mismanaged disaster situation out there. This is absurd-- even more so than "well, we had another Space Shuttle accident, therefore physics is a failed science."
The problems aren't with probability, they're with bad assumptions. One example: the Black-Scholes model for stock etc prices, which is taught as gospel in business and finance schools, makes the simplistic assumption that price changes are normally (i.e. Gaussian) distributed. That was a nice clean assumption to make for a toy model so the math would come out simply, but real-world price changes badly fail any normality tests- they're very leptokurtotic, i.e. very small and very big changes are both more common than with a normal distribution with the same variance. The result is that the model does considerably more harm than good, as it leads people to vastly underestimate some kinds of risks.
People claim that probability theory leads to people discounting unlikely contingencies- especially the combination of several individually unlikely factors - as not worth planning for. But it's not probability theory's fault that people make unwarranted frequentist assumptions, fail to take into account the increased possibility of error in multiple comparison tests, disobey Cromwell's Rule, overstate their certainty by using priors with insufficient entropy, or wrongly assume that events- esp. catastrophic events- are independent.
There are all kinds of horrid abuses of statistics out there. That means we need to do better at teaching people about probability and be more rigorous in rejecting badly done research; it doesn't mean we give up on a strong mathematical discipline which has made so many of the advances of the last hundred years possible.
Aaaaand... now I find out that not only was he vastly overstating the certainty, the whole thing was a late April Fool's joke, so the point is kinda moot. In their "fooled ya"/retraction, they admit that a solution is many many orders of magnitude off. Many of us knew that a full search was totally impossible, and the thing about stopping at an evaluation of -5.12 seems like it was just a way to fudge the claim enough to keep many of us from immediately realizing it was an April Fool's joke.
It can't spot every forced mate "well in advance of the point where it becomes a problem" - if it (or even a forced draw which doesn't begin with 3. Be2) exists anywhere in the googol of game possibilities in the King's Gambit Accepted it already is a problem. You seem to keep thinking this is about the computer managing to spot and avoid threats before a human can spot and exploit them. But it isn't. This isn't about playing a game against the computer. Nothing in the story, the comment thread, etc is. benthurston27's comment, along with every other bleeding comment in this whole bleeding article, is about the purported "solving" of the KGA.
We probably aren't getting all the details here, but the article said the "solution" stopped analyzing lines and said "win for black" if the evaluation function said black was up the equivalent of 5.12 pawns. benthurston27 was pointing out that the "Immortal Game" provides a concrete example of where the search may have tagged a line as won and moved on when the line was actually lost. Sure, Black can avoid this particular mislabeled line, but the analysis doesn't guarantee that Black can avoid all such lines. The programmer, when pressed about the fact that being up 5.12 pawns according to the evaluation function doesn't absolutely guarantee a win, said he's nevertheless 99.99999999% sure that Black has a forced win from any move other than 3.Be2 and a forced draw from 3.Be2. I think he's vastly overstating the certainty here.
And I never claimed that a chess engine can't spot a mate in 7, just that you were missing benthurston27's point:)
You probably caught on by now, but just in case, he wasn't saying the computer can't spot a mate 7 moves from where it starts its analysis, but rather that it can't spot a mate 7 moves further down a particular line than where it decided to stop its analysis.
BTW it wasn't a computer that identified black's 11th move as the crucial blunder- that was GM Robert Hübner back in the 1970s. Since that move puts black up material but causes a huge positional problem, it's harder-- especially for chess programs-- to see that it's a blunder. Even with a few minutes' thinking time on my C2D, Fruit seems to still think it's a good move that leaves black with an advantage. Unless I am much mistaken nobody's actually proven the game is lost that far back (still a rather huge game tree only 11 moves into the game) but as far as I can tell nobody- man or machine - has found a line that looks good for black under best play past that point.
Slashdot Mirror
You're counting a whole lot of zero-acre fires. If you look at the damage caused, target shooting accounts for a good deal more than 10%. Also, target shooters make up a rather small proportion of the population and cause a vastly disproportionate number of fires.
Any target shooting outside of a gun range during a red flag warning shows a lack of common sense, and trying to excuse these people's rampant irresponsibility by saying other people sometimes act irresponsibly too shows you're the one with the political agenda.
I don't have any idea who you think you're replying to. I'm not claiming gun ownership should be outlawed, and I don't see anybody who's making that claim.
You admit "Utah needs to change their local laws concerning the time and place it is appropriate to shoot" and that's precisely what I'm saying.
Your claim that these people did nothing illegal runs afoul of the reckless burning ordinance; this was a class A misdemeanor. But that's not enough to dissuade people from destroying land and endangering others' lives, because people are too stubborn to believe their irresponsible actions really cause any risk of fire, even when 19 fires had already been started by shooters in Utah this year.
Target shooting on public land during a red flag warning should be illegal, and it's farcical that the Legislature has not only refused to put in place reasonable regulations but has barred counties and municipalities from doing so.
So destroying thousands of acres of public and private land, costing the state millions of dollars in firefighting costs, risking the lives of firefighters, and causing >9000 people to evacuate their homes and businesses doesn't really matter as long as nobody got killed and no homes were destroyed?
Even if the target shooters had the money to pay the firefighting costs (extremely unlikely), the burned lands, the threat to others' lives and property, and the loss of >9000 people's time would be worth a criminal conviction.
There have been around a dozen fires started by target shooters in Utah this year, and some were larger than this; this one gets the news because it was closer to homes.
Years ago the legislature seized power to keep counties and municipalities from enforcing anything related to shooting, and they've repealed any and all restrictions on gun use they could find. They too are responsible for the fires.
Here's what pirates' attempts to move silently look like.
Google announced their initiative to save Algol 68 a couple years ago: they revised it a little and called it Go, in hopes that it would thereby become the Next Big Thing in software development.
It's not "perhaps" I'm "now" limiting my argument to 48fps vs higher rates; this whole conversation started because I was defending 48fps against someone who was complaining it wasn't 120. Neither is this just "slightly less outrageous" than saying 24fps is enough; anyone can distinguish 24fps from higher rates with sufficiently quick full-frame motion, but that simply isn't the case with 48fps.
The BBC "study" you cited is an unscientific failure. It involved no blind testing; the viewers knew at all times what framerate they were looking at, meaning the results are almost sure to be dominated by confirmation bias. Also, they spent a good deal of time and effort looking at individual frames, looking at slow-motion playback, and complaining about the loss of detail in an individual frame caused by having non-infinite shutter speeds, none of which has any bearing on whether there's a difference in perceptible detail when played back at normal speed. On top of that, they used naive averaging of frames from their high-speed camera to simulate the slower speeds, which is a very poor approximation; their artifical low-frame-rate videos definitely looked worse than real 50fps capture. (You have to have a properly weighted average to get anything close to the right results; see this paper for some detail.)
Look at my reply to another poster regarding the other effects which come into play when you're considering the framerate of games.
How the devil did you and several others get the impression I was saying 48fps is overkill? I was defending 48fps against somebody who was complaining that it wasn't 120fps. 24fps is very obviously not sufficient for rapid full-frame motions to appear smooth. Just moving to 30fps makes a big difference (the variable we really should be using here is frame delay, so the objective difference between 24fps and 30fps is just as big as the difference between 60fps and 120fps, i.e. 8.3 ms), but that may still appear slightly jerky in some situations. As I said, by the time you reach 48fps with proper motion blur the studies simply don't show a perceived quality difference between that and higher framerates.
I've seen the 100fps.com page before, and while it has some obvious truths it's also got a lot of half-truths, distortions, and absurd speculations. It's not really a good source for learning about perception.
The primary reason a 48fps average may not be good enough for gaming is because of variability. In a 48fps movie, every single frame comes 20.83ms after the one before it. If a game has an average of 61 FPS, it could be that sixty frame times are 15ms and the sixty-first is a very noticeable and jarring 100ms. Hardware reviewers, starting with Tech Report, are starting to catch on to the fact that the longest frame times matter much more than the average.
Two other reasons why a framerate where movies are smooth may not suffice for gaming:
Here's the study showing little difference between 30fps and 60fps for a first person shooter. Note the dramatic differences up to 15fps, where we begin to perceive many things as motion rather than individual frames, the still-quite-noticeable improvement in moving to a 30fps rate, where quick full-frame motions are starting to appear smooth, and the much smaller change (esp. the almost-zero perceived quality change) between 30fps and 60fps. Since the frame time difference between 30fps and 48fps is 12.5ms and the difference between 48fps and 60fps is only 4.17ms, I sincerely doubt their test setup would have showed any difference between 48fps and any higher rate (even 240+). However, their test setup was different from most games in one or maybe two of the three points I mentioned above. The frame rate limitations in the paper were artificial- the computer was capable of producing a much much higher framerate- so the frame times were probably almost exactly consistent. It's also possible that their mechanism for limiting framerate didn't give as strong a link between framerate and input lag as most games have (you'd have to look at the code to be sure).
You may think you can see it, but unless you've shown you can tell the difference in a double-blind study, that only tells us about your psychological biases rather that about your perception. (Saying you're biased is not an insult here; all of us are subject to quite a number of well-known psychological biases, and if anybody ever wants find the truth in these kinds of things they have to find ways to avoid letting these biases control their conclusions.)
All kinds of people think they can see 120fps, can hear the difference between normal audio cables and $10,000 gold-plated audio cables, etc. They can't. To believe such claims when they directly contradict all science is backwards superstition. If you're really so sure you can see the difference between 48fps and 120fps with proper motion blur, I'd bet the Randi Foundation would like to offer you their one million dollar Paranormal Challenge prize if you can prove it.
You obviously have no understanding of how human perception works. Go educate yourself about the response time of rods and cones and don't go around mouthing off at people just because they're better informed.
The refresh rate on a CRT and the frame rate of a movie have almost nothing in common. Your eyes can tell, somewhat painfully, that they're only getting 60 strobes of light per second with darkness the rest of the time. They can't tell that they're getting only 48 frame changes per second, especially with normal motion blur.
No study I'm aware of has ever shown people successfully distinguishing >40fps with proper motion blur from any higher framerate. Indeed, I've seen studies that showed very little difference in perceived quality between 30fps non-blurred frames and much higher framerates.
People on /. keep acting as though "48fps is good enough" is like "640k should be good enough." There's a vast difference here; human perception has very definite limitations.
Bring in an expert.
AARGH should have used preview. LZO Professional.
Markus Oberhumer, author of LZO, also offers LZO Professional, a commercial version not subject to the GPL.
I imagine having a spaceport wouldn't be all that different from having an airport, though an airport sees constant use and a spaceport would therefore seem to be less of a disturbance.
Throughout the world a lot of airports have wildlife preserves- especially wetlands- near them; that's the case for both of the airports closest to me. The airport and its noise make it less likely that people will drain/bulldoze the wetlands for housing developments. Bacteria in wetlandscan make short work of deicing chemicals used by the airport, which would otherwise build up to toxic levels. Bird strikes don't increase as much as you might think. There's some mutual benefits here.
We've made some attempts to include Russia (mostly as monitors/operators at missile defense sites in Eastern Europe). Those attempts have generally been rebuffed. Plus, Russia is getting various benefits from their saber-rattling, and their behavior generally indicates that they'd be pleased as punch to see Iranian missiles falling on Israel and North Korean missiles falling on South Korea. They've consistently vetoed just about every UN effort to deal with peace and stability issues since Putin came in power. You're right that they would not be benefitted by attacks on Europe. But they simply don't believe that will happen.
I wasn't quoting myself "as a source" - instead of just repeating myself I was acknowledging that I've said it before, and I linked to my previous post because it included more details for people who were interested. For sources, most of what Rob Weir says on the AOO incubator dev list will show my point. I haven't followed it in the last few months but when I was reading it frequently stuff like this was going on all the time
The LibO download size may look bloated, but that's because their default download includes all the languages rather than having separate installers for each language. I switched to LibO 3.5 recently and my install uses ~75MB less space than my OpenOffice 3.3 install did.
I'd like to see AOO succeed. But its leadership dooms it. As I've said before:
He's not the only one either. Few people who aren't on the IBM payroll want to contribute to a project with that kind of leadership. People from the open source community in general and from the LibO camp in particular are reluctant to do anything to cooperate with Weir and co. By the time AOO actually gets a release out it will likely be too late to revitalize any interest in the project.
Another reason not to bother talking about this now: given the state of the Indian justice system, it may well happen that all parties involved are deceased before this actually gets a court date.
Wrong again; you're the one who needs to look at the numbers. Over half of the deaths in the Hiroshima and Nagasaki were basically immediate and were ascribable only to the scale of the explosion and not to any radiation effects. Even among the later deaths conventional burns were a primary cause. Less than 20% of all the deaths were caused by radiation sickness etc.
Even ignoring all the non-radiation deaths, the difference between the fallout from a nuclear bomb and any possible contamination from a nuclear plant is tremendous. While it's easy to find lots of politically-motivated exaggerated numbers about Chernobyl- which due to horrible design failures and a deliberate simulation of disaster conditions was much much worse than any disaster at a sanely designed reactor will ever be- realistic estimates of the number of total Chernobyl-related deaths (including long-term increased cancer risk deaths over 26 years) are under 5000, while the atomic bombings in Japan caused ~200000 deaths within a period of a few months, ~40000 of which were due to radiation.
Trying to claim that nuclear reactor disasters are at all comparable to nuclear bombings is utterly ridiculous.
If you think the figures from the atomic bomb have any relevance to the present discussion you have no understanding of physics.
There is almost as much of a difference between a nuclear bomb and a nuclear power plant as there is between a nuclear power plant and the nuclear family. Same word, completely different implications.
Lots of people out there- this guy, Nassim Taleb, tons of economic pundits (usually not real economists), etc- are ready to blame probability theory for every mismanaged disaster situation out there. This is absurd-- even more so than "well, we had another Space Shuttle accident, therefore physics is a failed science."
The problems aren't with probability, they're with bad assumptions. One example: the Black-Scholes model for stock etc prices, which is taught as gospel in business and finance schools, makes the simplistic assumption that price changes are normally (i.e. Gaussian) distributed. That was a nice clean assumption to make for a toy model so the math would come out simply, but real-world price changes badly fail any normality tests- they're very leptokurtotic, i.e. very small and very big changes are both more common than with a normal distribution with the same variance. The result is that the model does considerably more harm than good, as it leads people to vastly underestimate some kinds of risks.
People claim that probability theory leads to people discounting unlikely contingencies- especially the combination of several individually unlikely factors - as not worth planning for. But it's not probability theory's fault that people make unwarranted frequentist assumptions, fail to take into account the increased possibility of error in multiple comparison tests, disobey Cromwell's Rule, overstate their certainty by using priors with insufficient entropy, or wrongly assume that events- esp. catastrophic events- are independent.
There are all kinds of horrid abuses of statistics out there. That means we need to do better at teaching people about probability and be more rigorous in rejecting badly done research; it doesn't mean we give up on a strong mathematical discipline which has made so many of the advances of the last hundred years possible.
Aaaaand... now I find out that not only was he vastly overstating the certainty, the whole thing was a late April Fool's joke, so the point is kinda moot. In their "fooled ya"/retraction, they admit that a solution is many many orders of magnitude off. Many of us knew that a full search was totally impossible, and the thing about stopping at an evaluation of -5.12 seems like it was just a way to fudge the claim enough to keep many of us from immediately realizing it was an April Fool's joke.
*sigh*
It can't spot every forced mate "well in advance of the point where it becomes a problem" - if it (or even a forced draw which doesn't begin with 3. Be2) exists anywhere in the googol of game possibilities in the King's Gambit Accepted it already is a problem. You seem to keep thinking this is about the computer managing to spot and avoid threats before a human can spot and exploit them. But it isn't. This isn't about playing a game against the computer. Nothing in the story, the comment thread, etc is. benthurston27's comment, along with every other bleeding comment in this whole bleeding article, is about the purported "solving" of the KGA.
We probably aren't getting all the details here, but the article said the "solution" stopped analyzing lines and said "win for black" if the evaluation function said black was up the equivalent of 5.12 pawns. benthurston27 was pointing out that the "Immortal Game" provides a concrete example of where the search may have tagged a line as won and moved on when the line was actually lost. Sure, Black can avoid this particular mislabeled line, but the analysis doesn't guarantee that Black can avoid all such lines. The programmer, when pressed about the fact that being up 5.12 pawns according to the evaluation function doesn't absolutely guarantee a win, said he's nevertheless 99.99999999% sure that Black has a forced win from any move other than 3.Be2 and a forced draw from 3.Be2. I think he's vastly overstating the certainty here.
And I never claimed that a chess engine can't spot a mate in 7, just that you were missing benthurston27's point :)
You probably caught on by now, but just in case, he wasn't saying the computer can't spot a mate 7 moves from where it starts its analysis, but rather that it can't spot a mate 7 moves further down a particular line than where it decided to stop its analysis.
BTW it wasn't a computer that identified black's 11th move as the crucial blunder- that was GM Robert Hübner back in the 1970s. Since that move puts black up material but causes a huge positional problem, it's harder-- especially for chess programs-- to see that it's a blunder. Even with a few minutes' thinking time on my C2D, Fruit seems to still think it's a good move that leaves black with an advantage. Unless I am much mistaken nobody's actually proven the game is lost that far back (still a rather huge game tree only 11 moves into the game) but as far as I can tell nobody- man or machine - has found a line that looks good for black under best play past that point.