Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: When is the loss of raw data a relevant event? Lightyears between

Author: Rolf Tueschen

Date: 16:44:29 10/03/02

Go up one level in this thread


On October 03, 2002 at 19:24:54, Omid David wrote:

>On October 03, 2002 at 15:15:18, Rolf Tueschen wrote:
>
>>In CTF I wrote about the scandal of the German scientist Schön. And now I have
>>explained why the data loss of Schön is somewhat more relevant than the bad luck
>>Bob Hyatt had with his disk crash. But since Vincent attacked Bob over a period
>>of some days, I think it's worthwile to understand the differences of the two
>>cases. Are my conclusions ok so far? Thanks for the interest.
>>
>>Here is the article from CTF, where Bob never appeared.
>>
>>QUOTE
>>
>>Subject: Re: Fossils and Bob's parallel computers are different to Schön's
>>"loss"
>>
>>Posted by Rolf Tueschen (Profile) on October 03, 2002 at 15:03:34:
>>
>>In Reply to: Fossil remains *Bad excuses* posted by Matthew Hull on October 03,
>>2002 at 10:07:45:
>>
>>
>>On October 03, 2002 at 10:07:45, Matthew Hull wrote:
>>
>>>On October 02, 2002 at 18:33:22, Rolf Tueschen wrote:
>>>
>>>>On October 02, 2002 at 18:02:03, Axel Schumacher wrote:
>>>>
>>>>>
>>>>>
>>>>>>On October 02, 2002 at 09:01:33, Rolf Tueschen wrote:
>>>>>>
>>>>>>>On October 02, 2002 at 08:53:36, Rolf Tueschen wrote:
>>>>>>>
>>>>>>>>A couple of months ago I wrote about the suspicion against the young German
>>>>>>>>researcher Jan Hendrik Böhm. Now the commission found him guilty. It's a sad
>>>>>>>>case of cheating in science.
>>>>>>>>
>>>>>>>>http://www.srijith.net/trinetre/archives/000029.shtml
>>>>>>>>
>>>>>>>>Please make your own thoughts.
>>>>>>>>
>>>>>>>>Rolf Tueschen
>>>>>>>
>>>>>>>Here is a better site:
>>>>>>>
>>>>>>>http://www.lucent.com/news_events/researchreview.html
>>>>>>
>>>>>>In the German reviews I found one who at least left one door open in defense for
>>>>>>the past alegged new Nobel Prize Laureate.
>>>>>>
>>>>>>I try to translate:
>>>>>>
>>>>>>The earlier "highly competent and sharp examination system" seems no longer to
>>>>>>function in case of the Multi-Publications.  The earlier in the USA very sharp
>>>>>>system of referees for the invoiced works is sloppy now. Ingolf Ruge from
>>>>>>Fraunhofer Institute thinks that Schön did not do it intentiously but became a
>>>>>>victim of the multi-publishing, following the method, don't believe in
>>>>>>statistics, you didn't cheat yourself.
>>>>>>
>>>>>>So far the quote.
>>>>>>
>>>>>>Rolf Tueschen
>>>>>
>>>>>
>>>>>Maybe he did it unintentionally. But is it then stupidity?? Difficult to
>>>>>believe, since he is certainly a very clever guy. So, the main accusation was,
>>>>>that he included the same graphs in different papers. "In April, a small group
>>>>>of physicists noticed that graphs in three unrelated papers appeared identical
>>>>>down to what should have been random noise.". Hmmm...strange..I'm a scientist
>>>>>myself and I never ever send the same graphs to different journals. I know
>>>>>exactly what is included in these papers. I really don't know how that can
>>>>>happen. Of course this will show up very soon after publishing. If I'm
>>>>>interested in a special topic I read usually all the papers from a group. Very
>>>>>strange case indeed.
>>>>>But maybe the biggest suspicion arose from the "lost" data: "According to the
>>>>>report, most of the evidence of those original experiments has vanished. Schön's
>>>>>transistors were all damaged or destroyed during the course of the original
>>>>>experiments, and attempts to replicate them failed. Likewise, Schön had deleted
>>>>>the raw data he had originally gathered in his experiments, because "his
>>>>>computer lacked sufficient memory".
>>>>>On the other hand. This can happen. I keep all my original data as far as I can,
>>>>>but if I have to look trough some data, 5 or 6 years ago it will get difficult,
>>>>>I'm sure I could not find all data.
>>>>>Anyway; he admitted that he did a lot of things wrong. But he insists, that
>>>>>everything was experimental data. We will probably never know.
>>>>>Nevertheless, I think it is a difficult situation for well known scientist.
>>>>>Everyone expects super-duper-great results from you. All the time. Maybe some
>>>>>scientist just break down by that pressure and try everything to please the
>>>>>public. I expect a quite high percentage of misconduct in science left
>>>>>unnoticed.
>>>>>Axel
>>>>
>>>>Axel, thanks,
>>>>
>>>>I want to add just this. In one of the quotes I read something he has allegedly
>>>>said now. To make it clear: the main reason for his condemnation by the
>>>>committee was not the point of the incredible number of equal diagrams with
>>>>several different descriptions. The main reason was the impossibility to
>>>>reproduce his results. Ok, and now what he said: (quote by heart) I had these
>>>>results and from now on I will try, no matter how difficult it might be, to
>>>>produce them again. - -
>>>>
>>>>Well for me that sounds strange. Simply because if he already had the results
>>>>once before, he should not insinuate that it might be most difficult. He is
>>>>either reacting on the committee, but that is then a contradiction to his own
>>>>pretended past, or he is arguing against his own concsience and he has let slip
>>>>a sort of unconscious confession. Because he won't believe that he must now wait
>>>>forever for the reproduction if he had it done already before.
>>>>
>>>>What do you think about my logic?
>>>>
>>>>As for the diagrams I think that it's strange that 19 collegues of
>>>>Schön,allproven innocent, should never have read these articles with the false
>>>>diagrams? And if they had read, why they didn't discover something?
>>>>
>>>>A third point of his data security.
>>>>
>>>>It is almost not to believe that exactly the main data of his most important
>>>>discovery should have been lost. I mean it's so basic, that some day someone
>>>>might ask you to show them. So all this is a complete violation of all known
>>>>necessary standards of science. It's a difference if some young student loses by
>>>>chance the data and the first version of his paper - out of his car or such, or
>>>>if a future alleged Nobel Prize Laureate loses his main data before others had
>>>>the chance to reproduce anything.
>>>>
>>>>Rolf Tueschen
>>>
>>>
>>>Rolf,
>>>
>>>This kind of thing is very common.  In fact, it is standard procedure in
>>>radiometric fossil dating.  But no one questions the results because those
>>>results are "politically correct".
>>>
>>>But dont take my word for it.  Check into it yourself.  Find the published
>>>papers and look.  You will find that none of the results are re-producable or
>>>verifiable.  Everyone is expected to take their results on faith, and that's
>>>exactly what everyone does.  Professors, text books, documentaries, journals,
>>>everyone just believes it.  It is the biggest fraud in all of "science".
>>>
>>>The most important fossil remains are stored with such high security and
>>>controlled access in steel vaults, that virtually no one is allowed to examine
>>>them.  Anyone wanting to check the work of the original discoverers and their
>>>published results will not be permitted to do so.  One may well ask, "what's the
>>>big secret?"
>>>
>>>Regards,
>>
>>I don't think that is a good example. Because Schön had to create new
>>experimental technology to get his "effects" while the method to calculate the
>>age of such fossils is world-widely known and accepted IMO. So a specific result
>>is not depending on the particular machines being used. But in the case of Schön
>>look at this:
>>
>>He lost or destroyed
>>
>>(1) his raw data
>>
>>(2) the exact experimental design
>>
>>(3) the proof for his effects he pretended of having found
>>
>>That is a bit too much.
>>
>>Let's give that a little importance with a comparison to the Vincent attack on
>>Bob Hyatt.
>>
>>Here are the differences and similarities.
>>
>>(1) Bob lost his raw data by a disk crash.
>>
>>(2) Bob is no one who wanted to prove by such results that he is a qualified new
>>expert. Because he's already for decades an expert.
>>
>>(3) His experiments are by no means a secret or difficult to reproduce. Almost
>>everybody with enough ressources could repeat it. Either with 2 processors or 4
>>or 8 etc.
>>
>>(4) Bob didn't find something really unexpected. Or would you say that the fact
>>that more processors have a potential win in speed is something relatively
>>unexpected?
>>
>>So, I would conclude that the data loss of Bob is nothing relevant in comparison
>>to the loss of Schön.
>>
>>Rolf Tueschen
>
>
>A researcher is expected to keep track of his experiments, or be able to
>reproduce them. In computer chess, the latter usually isn't possible, since the
>very frequent engine updates and the difference between different versions of
>the engine (not mentioning hardware upgrades) doesn't allow an exact
>reproduction of the original experiments.
>
>So in computer chess it's very important for a researcher to save his log files,
>since they are his only proof of his results' accuracy. The data stored in
>computers can be lost due to many reasons, so backups are necessary to protect
>this valuable data (I personally store hundreds of thousands of log files on a
>daily basis). But even then, there are cases where the data is lost and the
>backup is unreadable (as was the case in Bob's DTS article).
>
>Omid.

Ok, but here I wanted to compare Schön case to our discussion about the other
case.

Rolf Tueschen



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.