Walt Brown is absolutely within his rights to only debate astigmatic Sagittariuses with red hair, but he should at least admit that what he’s doing is peer review. Or gatekeeping. Or censorship, if you prefer.

Do I wish to debate the claim that you were told that genes evolved from mutated viruses forty-two years ago? Sure, after you debate my claim that in 1978 my cousin said that he had a girlfriend in Saskatchewan. In other words, no. I think that your memory is faulty, and/or that you misunderstood what you were told, or that you had a bad teacher who told you something wrong. If you can provide a citation of a high school textbook from the 1970s claiming that genes arose from mutated viruses, I will of course stand corrected. But then so would the book.

I did not claim that the references to superheavy elements and magnetic monopoles came from Brown’s book. I said that Brown cites a particular lab’s experiments as proof of the piezoelectric nucleosynthesis portion of his hydroplate theory. If you are curious about this, as I was, you follow Brown’s reference (to the Proton-21 Electrodynamics Lab, in Kiev) to their website, and look through their other publications addressing their nucleosynthesis work. There you can find their claims about synthesizing stable superheavy elements (atomic weight up to ~500) and magnetic monopoles. Incidentally, they also claim that their nucleosynthesis process only produces _nonradioactive_ elements (“without alpha, beta, or gamma activities”). This is perhaps because they are seeking funding to continue their work, and wish to promote it as a way to process nuclear waste into safe, nonradioactive products (“In this case, similarly to nature, the products of laboratory nucleosynthesis contain practically no α-, β-, or γ-active isotopes, which opens the possibility of using the discovered physical phenomenon for the reprocessing of radioactive and toxic wastes.” It’s also in their ‘About Us’ statement: “Scientific-research Electrodynamics Laboratory (EDL) was established in 1999 by a group of private investors within a scope of a venture project on creation a safe and effective technology for radioactive waste utilization. The project was based on an innovative and original conception on initiating an extreme conditions for nucleosynthesis process ignition in a super-dense cold substance”). Sadly, this rather contradicts Walt Brown’s claim that nucleosynthesis in this fashion generated _all_ of the Earth’s radioactive elements. You can find the papers yourself at:

link to proton-21.com.ua

I would recommend their “Superheavy nuclei research”, “Full-Range Nucleosynthesis in the Laboratory”, and “Experimental observation and analysis of action of light magnetic monopoles on multilayer surfaces” to begin with. I know you won’t be put off by the fact that they self-publish and/or publish in journals like Infinite Energy.

Regarding biological weapons – sure, they often backfire on the ones that use them. My point was that your statement “I could not have designed a better weapon” is absurdly wrong. To make a disease into an effective weapon, you would want it to be highly contagious, highly virulent over a tactically useful time period, and to not persist for long in the environment once the enemy has been incapacitated. You would also want to have effective preventatives or treatment for your own military and civilian populations. For HIV, the only criterion that it meets is that doesn’t persist in the environment like, say, anthrax or tetanus. HIV makes for an absolutely miserable biological weapon.

Anyhow, on to the comet argument. I’m tempted to ask you for the third time to answer whether you think Ron Wyatt found both Arks, Noah’s grave, Jesus’ blood, etc., and whether Joseph Smith translated golden plates with miraculous stones before proceeding, but since I find the math here so interesting I’ll continue to wait for your answer on those questions and just proceed.

First, I believe that Brown’s method of projecting cometary orbits backwards in time without actually modeling their gravitational interactions with planets and anchoring the model to recorded observations is unphysical, and akin to predicting how long it would take you to drive the length of Manhattan Island based on how long it takes you to drive from your garage to the end of your driveway. Still, we can stipulate to all of Brown’s starting conditions and see whether his claimed result matches what would happen if his starting assumptions were applied rigorously.

To recap, Brown observes that cometary orbits are variable, and that for Halley and Swift-Tuttle (ST), multiple observations over many years allow us to define a mean period as well as a sample standard deviation from that mean for each comet. Brown then oddly assigns a _constant_ period to both comets, and projects these constant periods back in time to find the point at which they come closest to one another: call that the ‘closest approach’ (actually, he searches for the closest approach over a particular window of time). Once he has found the point of closest approach between Halley and ST perihelia over his search window (which will always fall on a unique date, given that the starting dates and periods for each comet are _fixed_ in this analysis), he then claims that that closest approach must lie between the 27th Halley orbit and the 22nd ST orbit. He then attempts to reintroduce the period variability that he had previously discarded by applying the standard deviations of the mean orbital periods specifically to the dates of the 27th Halley and 22nd ST perihelia to generate that peak around 5300 BC. His trick of shifting the starting dates of the backwards projection without varying the periods of the comets does nothing to correct this fundamental error, which lies in ignoring the variability of the periods in his first analysis.

Here’s the error in a nutshell. Brown is saying the following: IF these comet orbits are CONSTANT, THEN the shortest sum-of-squares wait between perihelia is a FIXED DATE, between FIXED PERIHELIA (27th Halley, 22nd ST). Then he performs an unrelated calculation, of the following form: IF these comet orbits are VARIABLE, following a particular statistical distribution, THEN the dates of two GIVEN perihelia follow the following DISTRIBUTIONS, and the MEAN of those distributions follows this THIRD DISTRIBUTION: centered on 5300 BC with a standard deviation of ~100 years.

The _appropriate_ way to perform this test, given the same starting assumptions, would be to ask a different, but fundamentally more accurate question: IF these comet orbits are VARIABLE according to a particular distribution, WHEN is the closest approach between perihelia over the search window? There is no general analytical solution to this sort of question; it is like a three-body problem in gravitational systems. You can, however, _simulate_ the behavior of comets with variable periods, and you can find a date-of-closest-approach for those _simulated_ orbits. If the simulation is run many thousands of times, a scatterplot of those dates-of-closest-approach can be generated. If Brown’s analysis is correct, the scatterplot should show a nice tight Gaussian peak around the 5300 BC mark. When simulated, however, we see that because the cometary periods are variable, the closest approach of the comets does not always lie between the 22nd ST and 27th Halley perihelia. Frequently (indeed, in simulation, most of the time) it lies between other pairs of perihelia, and the distribution of simulated closest-approach dates is actually quite a complex shape.

In principle, the simulation works like this: you are given a start date for each comet, a starting period for each comet, and a means to modify the starting period using the known standard deviation of that comet’s mean period. You step off the perihelion dates sequentially by subtracting the starting period from the starting date, then modifying the starting period according to the rule, subtracting again to get the next perihelion date, then modifying the modified period, subtracting again, etc. The length of the period takes a one-dimensional random walk around its starting value. This is the simulation method that Brown uses himself in his ‘method C’ on his ‘Calculations that show comets began near Earth’ website page. Brown applies it to find the distribution of dates for a given perihelion for a single comet (ie ‘what does the distribution of dates for a predicted 19th Halley perihelion look like?) but he does not extend the method to both comets simultaneously.

Once you have generated a unique list of simulated perihelia for each comet using the above steps, you can compare those lists and look for a unique date of closest approach using the sum-of-squares method. Each set of simulations will give you one such date. Some of these closest approaches will be very close, even closer than Brown’s 5300 BC approach; some will not be particularly close, but each simulation will give you one clos_est_ approach date in the same way that Brown’s spreadsheet does in Step 2 on the page mentioned above. As an interesting side note and speaking just from memory, I believe that Brown’s closest approach is at about the 70th percentile of the approaches found through simulation – that is, about 30% of the simulations show tighter approaches than Brown’s – but I’d have to dig through a bunch of old files to find the exact number. I might well be misremembering.

This process is then repeated as many times as you wish, building a list of closest approach dates. If you generate a few tens of thousands of such runs, and plot out the closest approach dates, you see that there is no nice Gaussian distribution of dates around 5300 BC, which is what Brown’s analysis supposes. The actual distribution defies simple description, but there are a few discrete peaks starting around 1400 BC that quickly shorten and spread out as you go further back in time, such that they essentially merge into a constant baseline out beyond 1900 BC. Go farther back than that, and every year has roughly the same likelihood of being a closest approach year. Intuitively, this makes sense because the distributions of predicted perihelia dates for the two comets are so spread out that far back that they overlap themselves to an immense degree; if you see comet Halley in a given year that far back, it’s only slightly more likely that that’s the nth perihelion than the n+1, n-1, n+2, or n-2th perihelion. They all overlap significantly.

If you like, I can provide you with a copy of a commented program that can be run on a free statistical software package, and you can check out the results yourself. It might be best if we can reach agreement about the principle of the program before delving into the specifics. To wit: would you agree that given Brown’s starting assumptions of starting date, period, and sample standard deviation of period change, a simulation program could be written, _as Brown did_, that would generate a list of simulated comet perihelia for each comet (ie dates for the 1st, 2nd, 3rd, etc. perihelia before a given start date)? Further, that one such simulated list for Halley could be compared to a list from Swift-Tuttle to arrive at a simulated date of closest approach? And that repeating this simulation process thousands of times would produce a distribution of simulated dates of closest approach? And that if Brown is correct in his analysis, this method should duplicate his results and show a Gaussian peak at around 5300 BC with a standard deviation of about 100 years?

By the way, I believe that Brown’s argument against this simulation method will be that not every simulation run produces a closest approach that is tighter than his 5300 BC closest approach. And that it’s a pain to execute in Excel. Which facts are true, but irrelevant to whether or not it is a valid method. Such findings are only a problem if one presupposes that there _must_ be a distribution of closest convergence dates that is very tight, well ordered, and which falls in the range of dates that Brown is already willing to accept as Biblically plausible. The null hypothesis is that there is not such a distribution.

As I recall, he ran nine (!) simulations of the type I describe in Excel, over a smaller search space than he allowed for his original fixed-period analysis, and rejected the technique because the results did not match his, and none of the nine simulations produced a convergence that was as close as his 5300 BC convergence. Which I don’t find surprising, or troubling. Simulating the two comets one time and not ending up with a tight closest convergence could just be bad luck in the draw of the variables that vary the comet periods, even if his hypothesis was true. If that’s just bad luck, though, then repeating the process 500,000 or a million or ten million times should give you a much more accurate picture. And guess what: if ten million simulations say that it’s rare for you to _ever_ have a really tight convergence over your search window, that’s telling you something important. And that thing is not that the closest convergences that you find are the only ones worth considering. Further, if the 10 million closest convergences that you _do_ find don’t cluster around a particular date, that’s telling you something too.

In reviewing my calculations, I see that Brown’s 5300 BC closest convergence was actually in the 92nd percentile of simulated convergences, not the 70th. My memory was faulty, and I apologize for the error.