Timo Klaustermeyer - finish testing because of poor feedback

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
FriedmannC
Posts: 273
Joined: Fri Feb 10, 2012 6:58 pm
Location: SUCEAVA, ROMANIA

Timo Klaustermeyer - finish testing because of poor feedback

Post by FriedmannC » Mon Mar 19, 2012 12:53 pm

Timo announced me he would stop testing engines because of poor feedback and lack of interest of people in discussing the games he provided FOR FREE! And honestly I agree with him! He did a wonderful work, and I wish him all the best in his career!

TimoK
Posts: 97
Joined: Sun Jan 03, 2010 11:28 am
Location: Hamburg

Re: Timo Klaustermeyer - finish testing because of poor feed

Post by TimoK » Mon Mar 19, 2012 1:43 pm

Hi Catalin,

this thread really wasn't necessary in my opinion. This sounds like it's armageddon when I stop testing. 99,9% of the people would think: So what? :lol:

Further more you misunderstood me, I think. I didn't say I would stop testing engines completely. You asked me in a PM to continue my public testing with more engines. I answered I won't do any live broadcast and public covering of the tests any more. After all the feedback was too little (a bit better here, but very poor in the German CSS-Forum) to warrant any further public testing in the future. The effort is too high and only very few people were interested (or were at least giving some feedback).

I will continue to fiddle with computer chess because it's one of my favourite hobbies. Maybe I will provide some results of the tests I'll do in the future, let's see.

Best regards
Timo

User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 10:14 pm
Location: Munster, Nuremberg, Princeton

Re: Timo Klaustermeyer - finish testing because of poor feed

Post by Rolf » Mon Mar 19, 2012 2:08 pm

FriedmannC wrote:Timo announced me he would stop testing engines because of poor feedback and lack of interest of people in discussing the games he provided FOR FREE! And honestly I agree with him! He did a wonderful work, and I wish him all the best in his career!
Since we are a forum for open and free discussions, let me disagree with your concern. I could follow the whole topic which started in the German CSS, where many known testers are talking. Mainly in German of course. So, perhaps you might have missed it.

IMO - it's just my impression and critic - the whole idea of Timo is a hoax. Because IMO he was concerned about engines like Komodo because of its only 1-core actual state or in Junior 13 because of its weak play in TC which is common nowadays. So, Timo and others tweaked the usual testing precedure by crippling intentionally well designed engines like Houdini (probably the best) so that the other mentioned could still have a chance. That is as if you would plan to test a F1 spider car with a bicycle and therefore the spider must run on three tires to make it a fair challenge.

Why should one cripple intentionally the best programs? There is only one single possible answer - the now really weak former top programs of commercial authors have an interest to let it appear as if they still had a strong competitor in the race which is absolutely not true.

In case of Komodo it's the problem with the MP that is still not ready. So Timo shows interest in core-1 matches.

Everybody is free in his ideas and decisions, but I thought that I should lead your attention to that sort of actual hoax.

It's a cheat to the thousands if not millions of human chessplayers who want to train with the actually best possible engine and that is actually Houdini, not that it would be really stronger than Rybka, but Vas is keeping his novelties safe because he knows that most others only wait that they can use his tricks. But anyway a chessplayer therefore will analyse with Houdini. For sure not with Junior 13 as the hord of fans wants to insinuate in CSS with their welcome addressings.
-Popper and Lakatos are good but I'm stuck on Leibowitz

FriedmannC
Posts: 273
Joined: Fri Feb 10, 2012 6:58 pm
Location: SUCEAVA, ROMANIA

Re: Timo Klaustermeyer - finish testing because of poor feed

Post by FriedmannC » Mon Mar 19, 2012 2:11 pm

Hi Timo, I'm sorry if I made my statement sound in that way. Honestly, I didn't mean it. I'm glad your project is not over 8-)
All the best,
Catalin

TimoK
Posts: 97
Joined: Sun Jan 03, 2010 11:28 am
Location: Hamburg

Re: Timo Klaustermeyer - finish testing because of poor feed

Post by TimoK » Mon Mar 19, 2012 2:34 pm

Rolf wrote: IMO - it's just my impression and critic - the whole idea of Timo is a hoax. Because IMO he was concerned about engines like Komodo because of its only 1-core actual state or in Junior 13 because of its weak play in TC which is common nowadays. So, Timo and others tweaked the usual testing precedure by crippling intentionally well designed engines like Houdini (probably the best) so that the other mentioned could still have a chance. That is as if you would plan to test a F1 spider car with a bicycle and therefore the spider must run on three tires to make it a fair challenge.
So your opinion is that all 1-core testing like the excellent IPON rating list is a hoax because MP engines are "crippled" there?
Why should one cripple intentionally the best programs? There is only one single possible answer - the now really weak former top programs of commercial authors have an interest to let it appear as if they still had a strong competitor in the race which is absolutely not true.
Of course, my whole intention of the test was to push interest in commercial engines like Komodo! :lol: Don and Larry are daily transferring thousands of dollars to my bank account because of the excellent advertisement I'm doing for them... :roll:
It's a cheat to the thousands if not millions of human chessplayers who want to train with the actually best possible engine...
Where did I say I want to provide tips to human chessplayers about which software they should use? I provided the hyperlinks to my tests in 2 computer chess fora. I don't think there are millions of human chessplayers reading here or in the CSS-Forum and taking my tests as a benchmark on which software they should take for training. My tests were turned on computer chess experts who simply know that a Houdini with 6 threads would of course score better than a Houdini with just 1 thread. And for the Deep Junior 13 test I was telling everybody that my goal was to find out if there is any sort of TC that leads to a better scoring for Junior (because some people keep insisting that Junior is an engine that needs more time to be a good competitor). So no false pretences at all - everyone could read my intentions even on the test web page!

I don't know you, but I'm strongly under the impression that you are seriously suffering from paranoia if you want to ascribe any commercial interest to my testing. AFAIK you have a lifetime ban on the CSS-Forum - now I know why the former moderators had to do that.

noctiferus
Posts: 364
Joined: Sun Oct 04, 2009 11:27 am
Location: Italy

Re: Timo Klaustermeyer - finish testing because of poor feed

Post by noctiferus » Mon Mar 19, 2012 2:51 pm

I must confess my laziness in not having thanked you before for what I consider a great and significant effort on your side.
I'm not a good chess player, so you couldn't expect clever comments from me, but I find extremely interesting to look at these high-level matches.

Martin did something alike some time ago: IMHO, the only difference is that there was an interesting way of kibitzing and commenting during the games.
What you are doing is anyway great in producing interesting matches.
Please, go on with it.

Ciao
Enrico

Adam Hair
Posts: 3201
Joined: Wed May 06, 2009 8:31 pm
Location: Fuquay-Varina, North Carolina

Re: Timo Klaustermeyer - finish testing because of poor feed

Post by Adam Hair » Mon Mar 19, 2012 6:52 pm

Rolf,

Why do you, and others, take it for granted that the non-public version of Rybka is stronger than Houdini? Is there any basis for this?

Adam

User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 10:14 pm
Location: Munster, Nuremberg, Princeton

Re: Timo Klaustermeyer - finish testing because of poor feed

Post by Rolf » Mon Mar 19, 2012 7:27 pm

TimoK wrote:
Rolf wrote: IMO - it's just my impression and critic - the whole idea of Timo is a hoax. Because IMO he was concerned about engines like Komodo because of its only 1-core actual state or in Junior 13 because of its weak play in TC which is common nowadays. So, Timo and others tweaked the usual testing precedure by crippling intentionally well designed engines like Houdini (probably the best) so that the other mentioned could still have a chance. That is as if you would plan to test a F1 spider car with a bicycle and therefore the spider must run on three tires to make it a fair challenge.
So your opinion is that all 1-core testing like the excellent IPON rating list is a hoax because MP engines are "crippled" there?
Why should one cripple intentionally the best programs? There is only one single possible answer - the now really weak former top programs of commercial authors have an interest to let it appear as if they still had a strong competitor in the race which is absolutely not true.
Of course, my whole intention of the test was to push interest in commercial engines like Komodo! :lol: Don and Larry are daily transferring thousands of dollars to my bank account because of the excellent advertisement I'm doing for them... :roll:
It's a cheat to the thousands if not millions of human chessplayers who want to train with the actually best possible engine...
Where did I say I want to provide tips to human chessplayers about which software they should use? I provided the hyperlinks to my tests in 2 computer chess fora. I don't think there are millions of human chessplayers reading here or in the CSS-Forum and taking my tests as a benchmark on which software they should take for training. My tests were turned on computer chess experts who simply know that a Houdini with 6 threads would of course score better than a Houdini with just 1 thread. And for the Deep Junior 13 test I was telling everybody that my goal was to find out if there is any sort of TC that leads to a better scoring for Junior (because some people keep insisting that Junior is an engine that needs more time to be a good competitor). So no false pretences at all - everyone could read my intentions even on the test web page!

I don't know you, but I'm strongly under the impression that you are seriously suffering from paranoia if you want to ascribe any commercial interest to my testing. AFAIK you have a lifetime ban on the CSS-Forum - now I know why the former moderators had to do that.
Thanks for the feedback. It's a known problem on the net that such sloppy talks raise aggressions and anger in some people because they feel attacked or insulted. However if that should be meant as a new science, perhaps you could inform me about the faculty where you learned such things. Meaning that one must be crazy if one dared to ask questions or to criticize you when you are 'just' doing some experiments without (sic!) any intentions at all, especially when the experts know anyway all the results you could get! It's also interesting me how you diagnose paranoia in people you admittedly dont even know. BTW your afaik guess is wrong.

Back to Junior. Wouldnt you also prefer to test a much stronger Junior so that you had no obligation to cripple say Houdini to make it weaker so that Junior 13 had better chances? Why dont you take Houdini's Q off the board, because that might give Junior 13 or Komodo 4 a fair chance!?
Just joking a bit.

You are correct that IPON is IMO also a hoax without offering the moves of the games.
-Popper and Lakatos are good but I'm stuck on Leibowitz

User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 10:14 pm
Location: Munster, Nuremberg, Princeton

Re: Timo Klaustermeyer - finish testing because of poor feed

Post by Rolf » Mon Mar 19, 2012 7:50 pm

Adam Hair wrote:Rolf,

Why do you, and others, take it for granted that the non-public version of Rybka is stronger than Houdini? Is there any basis for this?

Adam
It's my belief that Vas is a different person than Robert Houdart.
However if not then he did enough to stay on top. If I had Vasik's talents I would have placed that egg of H. too to confuse all my critics! Excuse me, Houdini, for that joke. However I could also imagine that H. was designed by Hyatt. R.H. :D
-Popper and Lakatos are good but I'm stuck on Leibowitz

TimoK
Posts: 97
Joined: Sun Jan 03, 2010 11:28 am
Location: Hamburg

Re: Timo Klaustermeyer - finish testing because of poor feed

Post by TimoK » Mon Mar 19, 2012 11:31 pm

Thanks for the feedback. It's a known problem on the net that such sloppy talks raise aggressions and anger in some people because they feel attacked or insulted.
That may be a known problem, but if you read your first posting again from a more objective point of view you must confess that most people would feel attacked and insulted by it. You used some unpleasant expressions / verbalizations there - who wouldn't feel insulted when a person that one doesn't know says wrong and incredible defamatory things about ones intentions like you did?
Meaning that one must be crazy if one dared to ask questions or to criticize you when you are 'just' doing some experiments without (sic!) any intentions at all, especially when the experts know anyway all the results you could get!
Asked me questions / criticized me? Well, that's a brilliant understatement of what you did! You ascribed me to have commercial intentions with my tests and blamed me to cheat millions of human chess players intentionally because of "crippeling" Houdini in my tests! That sort of stuff you call "asking questions / criticizing"? C'mon you were intentionally provoking my harsh reaction to your post. Since I learnt in the past few hours that your profession is psychiatry you may like to dare people by your special way of asking questions / criticizing. But since I have a different profession I won't let the discussing go deeper in a direction you may want to provoke - so for me the discussion is over at this point.
Back to Junior. Wouldnt you also prefer to test a much stronger Junior so that you had no obligation to cripple say Houdini to make it weaker so that Junior 13 had better chances? Why dont you take Houdini's Q off the board, because that might give Junior 13 or Komodo 4 a fair chance!?
This paragraph shows that you didn't even take a closer look at the test or my webpage. Otherwise you would have noticed that Junior and Houdini both used 6 threads when playing against each other. Only in the matches against Komodo 4 and Fritz 13 Deep Junior 13 was playing with only one thread because his opponents can also use only one. That was a fair comparison IMO. You have a different opinion, that's your godgiven right. So what?

Post Reply