a Telltale position

[Guenther]

According to the arguments I've read, Vitruvius should be considered "virgin" ... But instead he is a "sailed woman"! :D

I don't know what you have read but it was always clear that it is a
derivate and the 'author' confesses this himself(may be still not emphasizing
enough the percentage of the 'original' bases)

Quote from Vitruvius site:

Vitruvius is built on the skeleton structure of the free Ippolit programs (RobboLito, Igorrit, IvanHoe) from which it inherits certain caracteristics and pecularties.

Well of course we can say there are a lot of muscles and brain too, beneath the 'skeleton'....

Guenther

[zullil]

According to the arguments I've read, Vitruvius should be considered "virgin" ... But instead he is a "sailed woman"!

I don't know what you have read but it was always clear that it is a
derivate and the 'author' confesses this himself(may be still not emphasizing
enough the percentage of the 'original' bases)

Guenther

Do you realize that you are replying to the 'author'?

Carotino = Roberto Munter = Vitruvius author

[Guenther]

According to the arguments I've read, Vitruvius should be considered "virgin" ... But instead he is a "sailed woman"! :D

I don't know what you have read but it was always clear that it is a
derivate and the 'author' confesses this himself(may be still not emphasizing
enough the percentage of the 'original' bases)

Guenther

Do you realize that you are replying to the 'author'? :)

Carotino = Roberto Munter = Vitruvius author

I did not realize this. Can you tell me then why he made the post I replied too? It makes zero sense in my universe(may be his 'English'?).

Guenther

[Adam Hair]

Please help me to understand if Ippolit is actually based on Rybka? So far I have not heard any compelling evidence, only Vas making a statement along those lines.I believe that Houdini is based on Ippolit, but am unsure if that also implies based on Rybka.

From the similarity tool:

Key:

1) Fire 1.01 (time: 27 ms scale: 1.0)
2) Houdini 1.00 (time: 29 ms scale: 1.0)
3) RobboLito 0.085g3 (time: 27 ms scale: 1.0)
4) Rybka 3 (time: 33 ms scale: 1.0)
5) Stockfish 2.11 (time: 31 ms scale: 1.0)

1 2 3 4 5
1. ----- 71.56 73.80 59.04 47.71
2. 71.56 ----- 71.67 59.37 48.34
3. 73.80 71.67 ----- 58.86 47.80
4. 59.04 59.37 58.86 ----- 49.96
5. 47.71 48.34 47.80 49.96 -----


I personally believe the percentages would be higher except for the fact that the self-simi;arity percentages Rybka 3, the Robbo/IvanHoe engines, and Houdini are only ~75%, unlike most engines that have self-similarity percentages of 95+%.

The similarity numbers between Rybka 3 and the others (Robbo, Fire, Houdini) is enough to raise my suspicion. The likelihood of these numbers, when compared to an estimate for the entire population of engines, is quite low. Mark Watkins has stated that whoever made Ippolito had a good amount of familiarity with Rybka 3. I believe Chris Conkie and others have shown that those engines show similar characteristics to Rybka 3. It is hard to have much doubt that Ippolito was born from Rybka 3. I guess the only questions are in the details.

Note to Kai Laskos:
I was wrong about the low self-similarity numbers being related to the positions. It does show up with other positions.

[Don]

I never said that Robbolito was akin to Rybka 2.3.2. I said that the easiest Rybka to disassemble was Rybka 2.3.2b (the tuning/programmable/leaked one) because the symbols had not been stripped and that it is my belief that was what was used to create Ippolit..... (please note Ippolit).

I know, but Larry extrapolated that to evals, and that is not correct.

With regard to the evals you see in my first post, what say you Miguel?

Like I implied in my post, you have a very good point here (the coincidences in PV + eval are really striking.). That should be stressed, not Qxd4 and the near neutral eval. Otherwise, it dilutes the argument.

Miguel

Chris

Ok. Just wanting to make myself clear.

I think it is helpful at this point to remember this thread Miguel.

http://www.talkchess.com/forum/viewtopi ... 01&t=34333

In this thread he said the following......

http://www.talkchess.com/forum/viewtopi ... 44&t=34333

I am here to tell you all that there is a 0% chance that Houdini is not based on Robbolito.



Chris

You don't need to convince anyone, I don't think anyone seriously contests this issue. It's well known and you can try any random position and get the same moves and even the same score if you adjust it using a formula that someone published here.

[Adam Hair]

Yes, thousands of positions have been run and shown to be considerably different. This is expected, since, whatever has happened, Rybka uses a material table that gives a very specific flavor to the evaluation. R1 is not that close to fruit in terms of move selection.

Miguel

I really think that the relationship between Rybka 1.0 Beta and Fruit 2.1, as defined by the similarity numbers, falls in a gray area. They show, as we all know, that Rybka is influenced by Fruit. There are other engines more heavily influenced (if not more), such as Alaric. But the similarity number between Rybka and Fruit is relatively high compared to ofther pairs.

Key:

1) Alaric 707 (time: 361 ms scale: 1.0)
2) Bright-0.5c (time: 187 ms scale: 1.0)
3) Fruit 2.1 (time: 290 ms scale: 1.0)
4) Gaviota 0.84 (time: 320 ms scale: 1.0)
5) Rybka 1.0 Beta (time: 171 ms scale: 1.0)
6) SlowChess Blitz WV2.1 (time: 479 ms scale: 1.0)

Code: Select all

         1     2     3     4     5     6
  1.  ----- 48.68 59.78 48.20 53.46 49.82
  2.  48.68 ----- 51.26 45.74 47.98 48.22
  3.  59.78 51.26 ----- 46.25 55.75 49.49
  4.  48.20 45.74 46.25 ----- 46.63 45.70
  5.  53.46 47.98 55.75 46.63 ----- 46.53
  6.  49.82 48.22 49.49 45.70 46.53 -----

[Sven]

It should be noted that the biggest part of this thread is more than one year old. That does not mean its contents were outdated. But people should take into account that they are mostly not replying to recent posts but to posts from the beginning of 2011.

Sven

[Adam Hair]

I never looked at the date.