World Championship Winning Computer Chess Software Program & Downloads for Chess Databases, Analysis and Play on PC, Mac, iPad and iPhone — Visit: Hiarcs.com
I have completed playing Excalibur Ivan and Igor on Level 53 30 seconds average time. Therefore below is a closer comparison at this level between Ivan, Igor, Grandmaster, Mirage and Avenger. I have also dropped in CXG Legend so that it gives you an opportunity to compare them at this level.
TEST GAME 1 - 15TH CENTURY MASTERS
As you can see from above everyone is a keeper, not one is exactly the same playing at this level. But you still clearly see that Mirage 10 MHz and Igor 12 MHz are still almost identical at this level. The additional 2 MHz did not help Igor in this first test.
TEST GAME 2 - THE TURK
In Test Game 2 CXG Legend scored best with Igor being the best compared Excalibur. As is Test 1 every computer plays a little differently.
RATING COMPARISON CHART AFTER 2 TESTS
The above results I think clearly also show the continuing debate. The moves played are different indicating a possible different programmer or a Tune up carried out by somebody. The continuous similarity in playing strength which is consistent if you play them against each other or test them as above always repeats itself. The tuned up Horvath or different programmer debate continues because of this.
Regarding not being able to use the spreadsheet, which version of Excel are you using. Unfortunately the pop downs etc might not work with later versions but it should work just fine with Version 2010 onwards.
Nick, I'm using Excel for Mac 2011. It's possible that I botched your instructions.
I will add Game 2 for Super Connie later this week, and then add the Saitek Turbo King.
If I get time, I'll throw in the Radio Shack 2150L, to add yet more fuel to the Kaplan debate. I don't think we've quite settled the question on 2150L, which may or may not contain Kaplan's C program. An earlier debate on the forum led to a general (but not unanimous) agreement that TK contains Kaplan's D or D+ program (stronger than C). No one said much about RS 2150 - I don't own it, but it has a much larger book than 2150L, and I wonder if it mirrors/clones the Turbo King (which has the same size book).
- R.
"You have, let us say, a promising politician, a rising artist that you wish to destroy. Dagger or bomb are archaic and unreliable - but teach him, inoculate him with chess."
– H.G. Wells
"You have, let us say, a promising politician, a rising artist that you wish to destroy. Dagger or bomb are archaic and unreliable - but teach him, inoculate him with chess."
– H.G. Wells
"You have, let us say, a promising politician, a rising artist that you wish to destroy. Dagger or bomb are archaic and unreliable - but teach him, inoculate him with chess."
– H.G. Wells
"You have, let us say, a promising politician, a rising artist that you wish to destroy. Dagger or bomb are archaic and unreliable - but teach him, inoculate him with chess."
– H.G. Wells
Hi Nick - so sorry. The first Game 1 move for Turbo King and Atlanta (as Black) is ...Nf6. My apologies.
Cutting and pasting too fast regards,
- R.
"You have, let us say, a promising politician, a rising artist that you wish to destroy. Dagger or bomb are archaic and unreliable - but teach him, inoculate him with chess."
– H.G. Wells
So far, these tests lead to three observations, one of which may force me to eat crow while Steve watches:
1. Igor, Mirage and Excalbur GM are closely related, and I'm betting standard Excalibur GM is closer than Platinum.
2. Krypton Legend/Systema Challenge (an accepted Horvath program) is not, NOT a clone of Igor. The deviations are too great. Igor and Legend deviate 42 percent of the time in Nick's tests. That's well beyond the generally accepted clone horizon of 90 percent, and it suggests more than tuning.
3. Avenger is an unsolved mystery. It deviates from Legend and Ivan in significant ways. I can't agree with mclane - Avenger deviates too much to qualify as a Legend clone - moves differed 77 percent of the time in the two test games.
- R.
"You have, let us say, a promising politician, a rising artist that you wish to destroy. Dagger or bomb are archaic and unreliable - but teach him, inoculate him with chess."
– H.G. Wells
Reinfeld wrote:So far, these tests lead to three observations, one of which may force me to eat crow while Steve watches:
1. Igor, Mirage and Excalbur GM are closely related, and I'm betting standard Excalibur GM is closer than Platinum.
2. Krypton Legend/Systema Challenge (an accepted Horvath program) is not, NOT a clone of Igor. The deviations are too great. Igor and Legend deviate 42 percent of the time in Nick's tests. That's well beyond the generally accepted clone horizon of 90 percent, and it suggests more than tuning.
3. Avenger is an unsolved mystery. It deviates from Legend and Ivan in significant ways. I can't agree with mclane - Avenger deviates too much to qualify as a Legend clone - moves differed 77 percent of the time in the two test games.
- R.
No Need For You To Eat Crow..R
The Question as to who programmed the later Excaliburs could very well be a question never fully answered although I suspect it will not stop those from trying(which can only be a good thing)
for me the question was settled years ago ..but I wont repeat all of that again because i have nothing new to add to the debate other then what I have repeated so may times now that it borders on
Correction - bad math on my part (see bolded/fixed number below):
3. Avenger is an unsolved mystery. It deviates from Legend and Ivan in significant ways. I can't agree with mclane - Avenger deviates too much to qualify as a Legend clone - moves differed 23 percent of the time in the two test games.
The underlying point still feels right to me, however. A dedicated that matches moves less than 80 percent of the time with a dedicated of roughly equal strength doesn't feel like a clone.
- R.
"You have, let us say, a promising politician, a rising artist that you wish to destroy. Dagger or bomb are archaic and unreliable - but teach him, inoculate him with chess."
– H.G. Wells
Reinfeld wrote:Correction - bad math on my part (see bolded/fixed number below):
3. Avenger is an unsolved mystery. It deviates from Legend and Ivan in significant ways. I can't agree with mclane - Avenger deviates too much to qualify as a Legend clone - moves differed 23 percent of the time in the two test games.
The underlying point still feels right to me, however. A dedicated that matches moves less than 80 percent of the time with a dedicated of roughly equal strength doesn't feel like a clone.
- R.
Your statement would be acceptable for clones but for identifying same programmers it is not that simple and not really correct in all cases. Take the Horvath Test post for example that I posted and compare Test Game 1 Style 9 with Legend to the standard out of the box style 5. There are 17 move deviations out of 34 moves. That would make it a similarity of 50% coming out of the exact same computer by just changing the style. Based on your comments you would discount it as an absolute certainty that style 9 is a different programmer to style 5.
I had posted these tests so that people can get a complete picture of the difficulty of these discussions.
The Avenger plays very similar to a Horvath so therefore I agree with Thorsten on this one. Besides Igor and Mirage had more variations at level 53 because it is an average time and a more random setting than the test I posted at Level 22 (Igor) and Level 23 (Mirage). You also have to factor in the slight speed difference. The similarity of Avenger would probably be seen even clearer if someone were to play it at level 23 (assuming it has 10 MHz) or Level 22 (if 12 MHz).
PS... Since I don't have an Avenger, here is an obvious question. Does it have the built-in play styles and style parameter settings to create your own out of the hundred thousand billion possible settings?
