Atlanta/Magellan ...Redux

[spacious_mind]

Here are the results again with Senator included:

------ATLANTA------ ------MAGELLAN------ ------SENATOR------
13. … c5 2.0 13. … c5 2.0 13. … c5 2.0
14. … Nc6 2.5 14. … Nb7 0.0 X 14. … Nb7 0.0 X
15. … Nc6 3.0 15. … Bc6 1.5 X 15. … Bc6 1.5 X
16. … Nxd5 3.0 16. … Nxd5 3.0 16. … Nxd5 3.0
17. … Nd4 2.6 17. … Nd4 2.6 17. … Nd4 2.6
18. … Bh3 2.2 18. … Bh3 2.2 18. … Bh3 2.2
19. … Bxg2 3.0 19. … Bxg2 3.0 19. … Bxg2 3.0
20. … Nxc4 3.0 20. … Nxc4 3.0 20. … Nxc4 3.0
21. … Bxb2 3.0 21. … Bxb2 3.0 21. … Bxb2 3.0
22. … Rxb2 3.0 22. … Rxb2 3.0 22. … Rxb2 3.0
23. … Qg4 3.0 23. … Qg4 3.0 23. … Qg4 3.0
24. … Rb8 3.0 24. … Rb8 3.0 24. … Rb8 3.0
25. … h6 1.2 25. … g5 0.0 X 25. … g5 0.0 X
26. … Qh5 2.1 26. … Qd7 2.6 X 26. … Qh5 2.6 XXX
27. … Qe5 3.0 27. … Qe5 3.0 27. … Qe5 3.0
28. … Rxb1 2.0 28. … Rxb1 2.0 28. … Rxb1 2.0
29. … Qc3 3.0 29. … Qc3 3.0 29. … Qc3 3.0
30. … a3 2.0 30. … a3 2.0 30. … a3 2.0
31. … Qb2 3.0 31. … f5 0.0 X 31. … f5 0.0 X
32. … h5 3.0 32. … h5 3.0 32. … h5 3.0
33. … h4 2.0 33. … h4 2.0 33. … h4 2.0
34. … gxh4 2.0 34. … gxh4 2.0 34. … gxh4 2.0
35. … Qa1+ 2.6 35. … Kg7 3.0 X 35. … Kg7 3.0 X
36. … Qh2+ 3.0 36. … Qh2+ 3.0 36. … Qh2+ 3.0
37. … Qg5+ 1.1 37. … Qg5+ 1.1 37. … Qg5+ 1.1
38. … Qd4+ 1.1 38. … Kf7 2.6 X 38. … Kf7 2.6 X
39. … Qh2 0.0 39. … Qh2 0.0 39. … Qh2 0.0
40. … Qg7 1.0 40. … Qg7 1.0 40. … Qg7 1.0

Magellan comparison to Atlanta = 7 deviations from 28 = 75.00% Match or 25.00% Mismatch
Senator comparison to Atlanta = 6 deviations from 28 = 78.57% Match or 21.43% Mismatch
Senator comparison to Magellan = 1 deviation from 28 = 96.43% Match or 3.57% Mismatch

My Magellan and Senator definitely show all the indications of being clones. Atlanta still has to prove it.

I think these tests are perfectly suitable for clone testing because if you can repeat 90% percent moves under these conditions there has to be very little doubt about it. But at the same time I would still stick to my guns that any machines below 90% under these conditions should be listed separately so that everyone can enjoy and see how a small difference like this would impact the ratings.

Moving on to Test Game 1 and keeping Senator and MM6 in order to confirm validations of other tests.

Best regards,

[Steve B]

My Magellan and Senator definitely show all the indications of being clones. Atlanta still has to prove it.

I think these tests are perfectly suitable for clone testing because if you can repeat 90% percent moves under these conditions there has to be very little doubt about it. But at the same time I would still stick to my guns that any machines below 90% under these conditions should be listed separately so that everyone can enjoy and see how a small difference like this would impact the ratings.

a 96% match rate between your Senator and Magellan modules seems to be a failry good indication that your Magellan is running slower then the normal 20 Mhz and a fairly good indication that these tests are NOT reliable for clone detection

It is well known that the only difference between the Senator and Magellan modules is the Magellan hash tables
Senator is rated a good 100 Pts lower then Magellan by Wiki and a good 100 Pts lower then the Atlanta/Magellan by Selective Search
I guess this difference does not show itself in 30 Sec .per move games??
the problem here is that a casual observer would walk away thinking the Senator and Magellan are clones..that is..identical..in every aspect of game play
when we know that in a longer time control the hash tables would show a significant difference
I take no position about these games being good or bad for determining strength against humans as I am not very interested in that but I am interested in their Clone detecting usefulness
as you are now prepared to say that these tests( with pondering ON,at 30 sec. per move) are good clone detectors...



I wonder..will you petition Wiki to show Senator and Magellan as one computer given your conclusion they are probably clones?


I think Not Regards
Steve

[spacious_mind]

I wonder..will you petition Wiki to show Senator and Magellan as one computer given your conclusion they are probably clones?


I think Not Regards
Steve

Hi Steve,
Answer is no absolutely not would I petition to join any computer! I have games where the 16 MHz Cougar repeats 100% a TC2100 game which has 10 MHz. Longer time periods and even at the same time period of 30 seconds there will be situates where the faster speed will find a better move and therefore justify it's separate ELO. I have been a proponent of ELO Split and not ELO bundling! Within reason that is.
The Test does very definitely show that both Senator and Magellan make the same bad move choices. Atlanta did not, so in my opinion my Atlanta program has other differences that sets it apart. Magellan/Senator showed the same trend, Atlanta so far did not want to go down the same path.
Speed I am sure that the Magellan has 20 MHz and I will open it after the Tests again to confirm it. My theory is that more likely the manufacturer did not really pay much attention what ROM they put into the machines so yours Magellan could be a different version to my Senator Magellan. In other words I might have a Senator Magellan and you might have an Atlanta Magellan.

Best regards

[Steve B]

I wonder..will you petition Wiki to show Senator and Magellan as one computer given your conclusion they are probably clones?


I think Not Regards
Steve

Hi Steve,
Answer is no absolutely not would I petition to join any computer! I have games where the 16 MHz Cougar repeats 100% a TC2100 game which has 10 MHz. Longer time periods and even at the same time period of 30 seconds there will be situates where the faster speed will find a better move and therefore justify it's separate ELO. I have been a proponent of ELO Split and not ELO bundling! Within reason that is.
The Test does very definitely show that both Senator and Magellan make the same bad move choices. Atlanta did not, so in my opinion my Atlanta program has other differences that sets it apart. Magellan/Senator showed the same trend, Atlanta so far did not want to go down the same path.
Speed I am sure that the Magellan has 20 MHz and I will open it after the Tests again to confirm it. My theory is that more likely the manufacturer did not really pay much attention what ROM they put into the machines so yours Magellan could be a different version to my Senator Magellan. In other words I might have a Senator Magellan and you might have an Atlanta Magellan.

Best regards

Nick...you said these tests are good clone detectors when we know that
the Senator and Magellan are not clones
I have never heard of this Magellan(Senator) Vs.Magellan(Atlanta) ..business
you think perhaps the manufacturer was not very careful and so released Magellan's that might be 100 Pts lower then other Magellan's if they were more careful?
perhaps ..I cannot say...
what I can say is that no where does the literature of the day mention this..anywhere
Mephisto/Saitek Catalogues always showed the same Estimated ratings and specs for Atlanta's and Magellan's and different estimated ratings for the Senator..
Selective Search Mag(a UK distributor) always mentioned Atlanta's and Magellan's as having the same exact programs and Specs..
Niggemann (a European Distributor)always advertised them with the same exact ratings and Specs
here in the US never a word about these "different" Magellan's
the list is endless really

what I am finally starting to realize is somehow you have a Magellan ..you call yours a "Senator" Magellan..which is different then the NORMAL Magellan released for sale (who knows perhaps they were only sold in Europe that way or perhaps someone modified your Magellan before you acquired it or perhaps it was a mistake in one of the production runs that Saitek/Mephisto caught and corrected)
I think I now understand why I achieved a 100% match and you did not

perhaps the Wiki entry should be slightly modified to place an * next to the entry mentioning that the ratings pool is impacted by the inclusion of 16 Mhz Magellans and Magellan(Senators) and Magellan(Atlanta's)?
anything is better then the separate listing now which to be honest is a bit misleading..unintentionally of course.. but still not really 100% accurate

Cmon ..you have pull with the Wiki folks
a request from me to make a clarification like that would be put on hold until the cows come home
but from you..it would be dealt with ASAP


Lets Have A Wiki entry with MORE information rather then LESS Regards
Steve

[spacious_mind]

I wonder..will you petition Wiki to show Senator and Magellan as one computer given your conclusion they are probably clones?


I think Not Regards
Steve

Hi Steve,
Answer is no absolutely not would I petition to join any computer! I have games where the 16 MHz Cougar repeats 100% a TC2100 game which has 10 MHz. Longer time periods and even at the same time period of 30 seconds there will be situates where the faster speed will find a better move and therefore justify it's separate ELO. I have been a proponent of ELO Split and not ELO bundling! Within reason that is.
The Test does very definitely show that both Senator and Magellan make the same bad move choices. Atlanta did not, so in my opinion my Atlanta program has other differences that sets it apart. Magellan/Senator showed the same trend, Atlanta so far did not want to go down the same path.
Speed I am sure that the Magellan has 20 MHz and I will open it after the Tests again to confirm it. My theory is that more likely the manufacturer did not really pay much attention what ROM they put into the machines so yours Magellan could be a different version to my Senator Magellan. In other words I might have a Senator Magellan and you might have an Atlanta Magellan.

Best regards

Nick...you said these tests are good clone detectors when we know that
the Senator and Magellan are not clones
I have never heard of this Magellan(Senator) Vs.Magellan(Atlanta) ..business
you think perhaps the manufacturer was not very careful and so released Magellan's that might be 100 Pts lower then other Magellan's if they were more careful?
perhaps ..I cannot say...
what I can say is that no where does the literature of the day mention this..anywhere
Mephisto/Saitek Catalogues always showed the same Estimated ratings and specs for Atlanta's and Magellan's and different estimated ratings for the Senator..
Selective Search Mag(a UK distributor) always mentioned Atlanta's and Magellan's as having the same exact programs and Specs..
Niggemann (a European Distributor)always advertised them with the same exact ratings and Specs
here in the US never a word about these "different" Magellan's
the list is endless really

what I am finally starting to realize is somehow you have a Magellan ..you call yours a "Senator" Magellan..which is different then the NORMAL Magellan released for sale (who knows perhaps they were only sold in Europe that way or perhaps someone modified your Magellan before you acquired it or perhaps it was a mistake in one of the production runs that Saitek/Mephisto caught and corrected)
I think I now understand why I achieved a 100% match and you did not

perhaps the Wiki entry should be slightly modified to place an * next to the entry mentioning that the ratings pool is impacted by the inclusion of 16 Mhz Magellans and Magellan(Senators) and Magellan(Atlanta's)?
anything is better then the separate listing now which to be honest is a bit misleading..unintentionally of course.. but still not really 100% accurate

Cmon ..you have pull with the Wiki folks
a request from me to make a clarification like that would be put on hold until the cows come home
but from you..it would be dealt with ASAP


Lets Have A Wiki entry with MORE information rather then LESS Regards
Steve

I agree Steve you are not going to read about a different version anywhere. However manufacturers using a different version ROM is not uncommon and this was never advertised, it was always found out through someone digging and testing.

I know my Atlanta, Senator and Magellan are Originals no one has ever put their fingers inside these except me.

Since we are both showing a different behavior in the Magellan's ( I am assuming this since you are talking about your past match and not having tried this new test) that I can only conclude that the Manufacturer used a couple of different ROM's in the Magellan. Since mine at first attempt leaned in the direction of the Senator I can only assume that their connection is closer than their connection to my Atlanta. This is just one Test. Let's see what happens over the 5 Tests.
If you have the time to play along then perhaps we can get to the bottom of this, We both know that we are going to get some variations and some built in randomness but overall our computers should be either quite closely aligned or there are some different ROM versions playing a factor here.

Best regards,

Nick

[Steve B]

I agree Steve you are not going to read about a different version anywhere. However manufacturers using a different version ROM is not uncommon and this was never advertised, it was always found out through someone digging and testing.

Since mine at first attempt leaned in the direction of the Senator I can only assume that their connection is closer than their connection to my Atlanta. This is just one Test. Let's see what happens over the 5 Tests.

Yes but publications like Computer Chess Reports(US) and Selective Search magazine(UK) always acted as Watch Dogs and tried to expose issues like this

i dunno Nick
starting to sound like you have a Senator in a Magellan casing
better check that you have the 512 kb Ram while your in there
and to think it was this Magellan that started the momentum to have the computers separated on Wiki



anyway...for the sake of this discussion I will grant you that there might be a statistically significant number of so-called "Magellan(Senator's)" out there
for what ever reason..
the point is...
if your tests hold up and you are still showing a cloning match between this version of the Magellan you have and the Senator..
will you ask for the Wiki entry to be modified with an * and an explanation the same way you campaigned for the original combined wiki entry to be split apart with no mention of any of this?

Best Regards
Steve

[spacious_mind]

I agree Steve you are not going to read about a different version anywhere. However manufacturers using a different version ROM is not uncommon and this was never advertised, it was always found out through someone digging and testing.

Since mine at first attempt leaned in the direction of the Senator I can only assume that their connection is closer than their connection to my Atlanta. This is just one Test. Let's see what happens over the 5 Tests.

Yes but publications like Computer Chess Reports(US) and Selective Search magazine(UK) always acted as Watch Dogs and tried to expose issues like this

i dunno Nick
starting to sound like you have a Senator in a Magellan casing
better check that you have the 512 kb Ram while your in there
and to think it was this Magellan that started the momentum to have the computers separated on Wiki



anyway...for the sake of this discussion I will grant you that there might be a statistically significant number of so-called "Magellan(Senator's)" out there
for what ever reason..
the point is...
if your tests hold up and you are still showing a cloning match between this version of the Magellan you have and the Senator..
will you ask for the Wiki entry to be modified with an * and an explanation the same way you campaigned for the original combined wiki entry to be split apart with no mention of any of this?

Best Regards
Steve

Hi Steve,
I did not have a Senator at the time of the tests a few years ago otherwise I would have thought of comparing the two previously.

Try this position with your Atlanta and Magellan if you have time and let me know what they play. It is Black to move. Setting 30 seconds per move.

[fen]4r3/p4Ppk/1pp5/3p1p2/3P1Pq1/BP2r1Pb/P2Q3P/R5K1 w - - 0 27[/fen]

As for reporting to Info, let's see what happens in the end before we decide what is correct. As I mentioned I raised the questions but others also did their tests before it was changed. My influence alone was not enough to change the decision. As it should be.

Besides we have no proof until other Magellan owners step up and replay the same game with theirs. Who knows it might be yours that is the Freak! j/k

Also without other Senator confirmations it might be the Senator that has a Magellan software too ?

Anyway if you can just set up that position with Magellan and Atlanta and let Black play at 30 seconds per move I would be interested in reading what move Black plays from the above position.

Best regards

[Steve B]

a trick position?

i tried the position on the Senator,Magellan and Atlanta
all 3 computers will play Rxe1+ in about 7 seconds And mistakenly announce a mate in 5
ignoring the fact that White can promote to a Knight with + on move 4
obviously a bug that appears in all three computers

so what does this tell you?

Wondering Regards
Steve

[spacious_mind]

a trick position?

i tried the position on the Senator,Magellan and Atlanta
all 3 computers will play Rxe1+ in about 7 seconds And mistakenly announce a mate in 5
ignoring the fact that White can promote to a Knight with + on move 4
obviously a bug that appears in all three computers

so what does this tell you?

Wondering Regards
Steve

Hi Steve,

My Magellan, Atlanta and Senator also play the same move So does Milano Pro. Never tested the Master.

The other computers below Milano Pro do not have this bug ie. Gk2100 and co.

It tells me that the core software in all these computers are of the same family which has an ingrained bug. It also tells me that my Atlanta, Magellan and Senator at least in this test are behaving as they should identically to yours.

So I think finding right answers is not that easy.

Best regards,

Nick

[Steve B]

so the Atlanta and Magellan have the same" core" software
well thats a start


ill take what i can get regards
Steve

[spacious_mind]

so the Atlanta and Magellan have the same" core" software
well thats a start


ill take what i can get regards
Steve

Okay!! lol

[spacious_mind]

a trick position?


Wondering Regards
Steve

Hi Steve,

No it was genuine game that I played in February 2009 between Cosmos 24 MHz and Milano Pro.

[Event "Fritz Tests 30s/move 20Mhz"]
[Site "Hoover, Alabama, USA"]
[Date "2008.02.09"]
[Round "4"]
[White "Saitek Cosmos 24Mhz"]
[Black "Mephisto Milano Pro"]
[Result "1/2-1/2"]
[ECO "D31"]
[WhiteElo "2105"]
[BlackElo "2158"]
[Annotator "Mind,Spacious"]
[PlyCount "70"]
[EventType "match"]
[EventRounds "4"]
[EventCountry "USA"]

1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. e3 e6 {Saitek Cosmos 24Mhz out of book} 5. Bd3
{Mephisto Milano Pro out of book} b6 6. Nf3 Bb4 7. cxd5 exd5 8. Qb3 Bd6 9. O-O
Bg4 10. Nh4 Ne4 11. Bxe4 Qxh4 12. g3 Qh5 13. Nb5 O-O 14. Bxh7+ Kxh7 15. Nxd6
Bh3 16. Qd3+ f5 17. Qd1 Qg6 18. Re1 Qxd6 19. e4 Qg6 20. e5 Re8 21. b3 Re6 22.
f4 Nd7 23. Ba3 Rae8 24. Qd2 Qg4 25. Re3 Nf6 26. exf6 Rxe3 27. f7 Re1+
28. Rxe1 Rxe1+ 29. Qxe1 Qf3 30. f8=N+ 1/2-1/2


Hard to believe that this was over 4 years ago. That position mystified me as I could not believe that Milano Pro could not find the winning move. So I used this as a Test position to test the other Morschies.

Regards,

[Steve B]

Cosmos 24 Mhz??


please dont tell me you also modify and tune your computers

is there no Justice?

Sigh Regards
Steve

[spacious_mind]

Cosmos 24 Mhz??


please dont tell me you also modify and tune your computers

is there no Justice?

Sigh Regards
Steve

No I wouldn't know how to do it I have two left hands. I bought the Cosmos in Germany after the 2008 WM from the person who beat my RS2250XL in the B - Semi Finals.

I have 5 tuned machines out of a total of xxx machines and each one is a double. I found a couple interesting for strength comparisons so I do use these 5 a lot.

Besides the more are tuned the less originals are left in the world we should be happy!! Our collection value goes up!

Best regards,

[spacious_mind]

I still have a few games to go before I can post results. But just to update you. Test Game 1 Magellan and Senator 100% identical moves. Test Game 2 Magellan and Senator 100% identical moves.

After Test Game 1 I decided to open both Magellan and Senator and compared. Both have 20 MHz and same ROM. Only differences are that Magellan has an additional ROM which I can only assume is the 512K Hash. One other difference there is this round tube which is Black with the Senator and Black and Blue with the Magellan. The Circuit Board clearly states Magellan/Senator in other words it is intended to be used on both. The game ROM is a more modern square rom which settles the fact that it cannot be taken out and swapped by someone else. The technology is totally different to the old ROM's on the Modules ie Lyon, Vancouver, Almeria etc. which were all interchangeable.

I can only assume two things:

1) by 30 seconds per move the Hash has almost zero influence and may kick in at longer times.
2) The Hash does not function on all Magellan modules.
3) There are two versions, but I kind of doubt that for now.

Still have to play Test Game 3 and 4 with Magellan and 1-4 with Atlanta so this is only a preliminary report without Atlanta comparison.

ps. I also ran the self test and both passed their internal check. No problems with the modules.

Regards

Nick