SCCT II Prize Tournament's Leader is still:Attacker 4 - Congratulations to Adriean Costinescu
For Current Standings:
It's too early to make any conclusions,but anyway currently Attacker v4.0 is playing even slightly better with a score:65 %
The point I am making here is the difference between 1 and 2nd place was huge last time, 146.5 for Attacker 3 to 128 for Iconoclast. This time, difference till now is very close.
Nelson and I were discussing your book the other day, and we were both impressed. A job well done! Keep up the good work :)
How does it go.. "In the land of the blind..."
For the foreseeable future, the best B90 books will win, the worst will end up at the bottom, and the rest will occupy somewhere in between.
The anti-b90 movement seems to be slowly gaining strength.
Sunny's book is, in my view, the star of the tournament because he is doing so well with 1.d4 openings.
Some games are played, and the worst performing book is kicked out, then some other games are played, and the worst performing book is kicked out again.
This way we'd get rid of the B90 slaves that are the main cause of Attacker winning. It's like seeing a lot of people jumping off a cliff, but other people keep following them anyway.
Sorry...but here i dont agree with you,for me it was really interesting too see that Attacker's performance is based on B90 opening or not
Thats why i have decided to create a ratinglist (without any B90 opening game of the current tour played so far)
And as we see currently (even without B90 opening) again Attacker has the best score:
Program Elo + - Games Score Av.Op. Draws
01 Attacker 4 : 3324 71 63 36 62.5 % 3235 63.9 %
02 Yograj 6.4 : 3310 73 71 49 60.2 % 3239 46.9 %
03 Sick Panda : 3310 66 61 41 61.0 % 3232 63.4 %
04 Go Go Angel : 3286 90 88 27 57.4 % 3234 55.6 %
05 Jagged Ice : 3274 62 61 46 55.4 % 3236 63.0 %
06 Fruit 2.327 : 3265 67 66 40 53.8 % 3239 62.5 %
07 EmreMert : 3247 66 65 38 51.3 % 3238 65.8 %
08 Franven 1.3 : 3244 70 70 42 50.0 % 3244 57.1 %
09 Stimulus Package : 3242 54 54 60 50.0 % 3242 63.3 %
10 TheGuard 4.3 : 3239 64 64 48 50.0 % 3239 58.3 %
11 Iconoclast II : 3237 62 62 60 49.2 % 3243 51.7 %
12 BarmikiM 3 : 3236 73 73 40 50.0 % 3236 55.0 %
13 Leschats 2.3 : 3236 76 76 34 50.0 % 3236 58.8 %
14 Midtown 1.0 : 3236 50 50 60 49.2 % 3242 68.3 %
15 Zeeman 2.6 : 3222 62 63 38 47.4 % 3240 68.4 %
16 Refugium : 3214 59 60 60 45.8 % 3243 55.0 %
17 Tornado 3 : 3214 52 53 60 45.8 % 3243 65.0 %
18 Megadeath : 3213 57 58 60 45.8 % 3242 58.3 %
19 CompMaster 2010 RE : 3210 71 72 43 45.3 % 3242 53.5 %
20 Tron : 3197 68 71 34 44.1 % 3238 64.7 %
21 Macedonia : 3185 68 70 52 41.3 % 3246 48.1 %
22 1337chess v5 : 3182 89 91 32 42.2 % 3236 46.9 %
Note:the rating list does not include any single game of B90 Sicilian Najdorf Opening
One thing more,Attacker is one of the strongest books that i have ever tested so far
And i hope we will see many books near to Attacker playing level strenght or even stronger in SCCT tournaments...
Out of 660 games so far, 28% are Najdorf lines. Here are some statistical notes:
Attacker (the leader) in the Najdorf 6 wins, 6 draws, 1 loss playing white, 5 wins, 5 draws, 1 loss playing black: 11+ 11= 2- (69%). However we mustn't minimize his other, non-Najdorf games: 6 wins, 11 draws, 1 loss with white, 5 wins, 12 draws, 1 loss with black: 11+ 23= 2- (62%). This is a fantastic record, 22 wins against 4 losses with 34 draws.
Three players specialized in Pirc: Iconoclast, Midtown and Macedonia. Macedonia got slaughtered with only 28% and Midtown/Iconoclast were at 45-48%, mildly disappointing for both, I'm sure.
Iconoclast and Midtown also specialized in Larsen (1.b3) openings. (Brandon and I did not coordinate this in any way, though great minds do think alike.) 16+ 32= 11- combined, 54%. So far Midtown has gotten 55% vs. Iconoclast's 53%.
Fruit and to a lesser extent Jagged Ice played B22 Sicilian (2.c3) and scored 73% with it, an eye-opening result.
Refugium specialized in B24-B25 Sicilian (2.g3) openings and is 48% so far with it. Maybe with a little more work?
French openings got badly treated, black getting 34%. Seven different books played it, though none exclusively.
Leschats was badly outclassed with Najdorf, 2 wins, 18 draws and 13 losses with both colors, 33% success. What is he thinking?
Sick Panda was dominant with white (13+ 15= 1- 71%) but less so with black (3+ 25= 3- 50%). Still, 16 wins against 4 losses? Who wouldn't take this?
1.d4 openings: 51%. 1.e4 openings: 58%. Only two tries with 1.Nf3 and 1.c4 apiece. Worthy candidates for the next tournament!
Book exits >= 0.50 with white: Attacker 13, Sick Panda 12, Go Go Angel 9. Iconoclast and one other book had zero. Most on the receiving end of these bad book exits: Megadeath 13, TheGuard 10, Jagged Ice 9. All players experienced this at least once, though the fewest were recorded by Attacker (1). Games where white achieved this objective were 51 wins, 50 draws and 5 defeats for white (72%). Obviously when you have 13 excellent book exits against just one bad one you're playing well.
In general, Attacker is no fluke. This is one of the best, if not the best, opening book that has ever played in a tournament.
A Healthy Panda that isn't happy with a draw as black would be very interesting to see.
Do you get to choose the engine and engine settings in these things?
Yes, but is this requirement written in stone somewhere? Suppose I have a book that is optimized for an engine that is widely acknowledged to be a little worse than R3? Using a book optimized for one engine with another engine can be a disaster.
>Yes, but is this requirement written in stone somewhere?
Yes, in The Ten/15 Commandments ;-) :
Suppose that Rybka 4 is released and due to the knowledge removed to make Remote Rybka more shiny Rybka 4 just sucks at the Najdorf, so that several books perform better with Rybka 3 than with Rybka 4.
Shouldn't the book authors be allowed to keep participating with Rybka 3 so all their work in this line isn't wasted instead of being forced to use Rybka 4 because it performs better than R3 with generic books?
In the same vein, (fe.) a book tuned for Zappa Mexico II should be allowed to participate with that engine (what if it's best? Doesn't that mean people should be using Zappa Mexico II to test their books and tuning to it?, using Rybka 3 currently is more arbitrary than anything).
Many thanks your interesting comments and detailed information :)
But as i mentioned before,everything can happen in the next rounds...i wish once more good luck to all Book Authors and let's win the best book !
846 of 924 games are played so far,in other words to the FINAL are left only 78 games !!!
A few comments about latest tournament:
-The previous SCCT I Champion ''Attacker '' still leads,very close to be again a Winner,but guaranteed to be at least 2nd place
-Sick Panda is currently 2nd,one of the strongest public books,which is fighting for the 1st place,just 2.5 points behind the current Leader
-Yograj 6.4 opening book is currently 3rd,one of my favorite books,normally it started very well,but unfortunately lost some points in the last rounds
-CompMaster 2010 RE is currently 4 place,very tactical book,has big chance to be 3rd and especially in the last rounds started to show its real strength
-Stimulus Package and Tron opening books are new participants in SCCT and the both books have big chances to be 3rd in the current tour
-Franven 1.3 book is a serious candidate to take 3rd place,note that it has 1 game less played than rest participants
-Go Go Angel is a very dangerous opening book,still has a big chance to be 3rd place too
-Jagged Ice is just 9MB,the smallest book,but its performance is almost equal to huge 200 MB books
-Zeeman 2.6,a successful book and currently is 13 place (the winner of Pionerchess),but here it seems the participants are much stronger
-EmreMert is 17th,not bad performance for a book without any optimization of hand or auto tuning (the book author is very new in book making)
-Iconoclast II is a strong book,personally i expected much more from it,but anyway for a higher performance i think it needs some optimizations/tuning
-Leschats 2.3 is a quality opening book,but i think its performance is fallen down due to its almost all strongest lines are blocked
> Yograj 6.4 opening book is currently 3rd,one of my favorite books
Thanks Sedat....Book authors can change lines and play another stronger variations in this book. I will share the book after tour.
Nelson, Yograj was suppose to play 1. d4 as one of the line but I forgot to remove ?? mark from the book before sending to Sedat, else results surely wold have been different.
Iconoclast II: well-regarded American gelding broke well, could not get inside, pulled up lame in the backstretch, hobbled in well back of the pack.
Attacker 4: Romanian favorite showed early speed, cruised wire to wire, displayed a strong finishing kick ending several lengths ahead.
>broken legs.....put out to pasture
More like they get euthanized right on the race track. If a race horse completely breaks its leg it probably is going to die long before it get's put out to pasture. Horse racing is fun but I always have to set aside about a hundred bucks every spring-summer to cover gambling losses.
>This is funny, almost like a horse-racing analysis.
Even better then that you can read your commentary while listening to this
http://www.youtube.com/watch?v=H9b2zB3HwOM - I know this has got to be your kind of music, you probably grew up listening to this guy on the radio.
Now, the race is on and here comes "Pride" up the backstretch
"Heartaches" are a-going to the inside
"My tears" are holding back
They're tryin' not to fall
"My heart" is out of the runnin'
"True love" is scratched for another stake
The race is on and it looks like "heartaches"
And the winner loses all!
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill