Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Engine ELO

Author: Joshua Haglund

Date: 07:06:56 01/30/05

Go up one level in this thread


On January 30, 2005 at 09:24:29, gerold daniels wrote:

>On January 30, 2005 at 01:32:56, Joshua Haglund wrote:
>
>>I think chess engines should not be given an approximate rating. I think they
>>should be given an ELO RANGE, and be ranked by the average of the ELO RANGE.
>>There shouldn't be an engine on the list unless it has played more then 5000
>>games. These games are consisting of different books, hardware, settings, etc.,
>>hence, giving the engine an ELO RANGE.
>>
>>30.01.2005
>>2937569 games
>>
>>Engine Popularity
>>
>>#	Engine	        N Games	Elo Avg.
>>1.	shredder 8	1527113	2608	±21
>>2.	fritz 8	        1061589	2543	±28
>>3.	deep fritz 8	1032565	2612	±38
>>4.	deep fritz 7	439209	2633	±52
>>5.	hiarcs 9	317815	2549	±30
>>6.	shredder 7.04	258551	2551	±29
>>7.	fritz 7	        228719	2578	±37
>>8.	junior 8	103815	2508	±22
>>9.	x3d fritz	76805	2560	±28
>>10.	deep junior 9	56381	2596	±26
>>11.	deep junior 8	51025	2540	±25
>>12.	junior 9	45999	2553	±26
>>13.	chess tiger 15	45253	2514	±25
>>14.	fritzinbahrain	37793	2501	±16
>>15.	deep juniorem	28771	2642	±31
>>16.	deep junior8.zx	22453	2535	±24
>>17.	gambit tiger2.0	21613	2503	±23
>>18.	ruffian 2.1.0	19755	2510	±8
>>19.	the king 3.33	18969	2528	±15
>>20.	shredder 7	18733	2510	±34
>>21.	pro deo 1.0	17649	2468	±11
>>22.	3d schach 4.0	16811	2519	±9
>>23.	fritz 5.32	14259	2400	±7
>>24.	fruit 2.0	13425	2537	±21
>>25.	aristarch 4.50	12305	2472	±7
>>26.	gandalf 6.0	11787	2508	±16
>>27.	shredder 7.se	11099	2510	±18
>>28.	junior 7	11037	2474	±13
>>29.	ruffian 1.0.5	10861	2485	±21
>>30.	chess tiger 14	10281	2514	±18
>>31.	hiarcs8 bareev	9267	2483	±20
>>32.	schachgenie 1.0	8763	2548	±9
>>33.	list 512	8693	2448	±4
>>34.	the king 3.23	8179	2469	±9
>>35.	crafty 19.01	6803	2356	±6
>>36.	fritz for fun 3	6751	2470	±5
>>37.	deep sjeng 1.6	6265	2449	±2
>>38.	the king 3.23r1	6047	2547	±3
>>39.	hiarcs 8	5739	2460	±3
>>
>>Engine ELO Range
>>
>>Shredder 8     ELO 2587-2629
>>Fritz 8        ELO 2515-2571
>>Deep Fritz 8   ELO 2574-2650
>>Deep Fritz 7   ELO 2581-2685
>>Hiarcs 9       ELO 2519-2579
>>Shredder 7.04  ELO 2522-2580
>>Fritz 7        ELO 2541-2615
>>Junior 8       ELO 2482-2530
>>X3d Fritz      ELO 2532-2588
>>Deep Junior 9  ELO 2573-2624
>>Deep Junior 8  ELO 2515-2565
>>Junior 9       ELO 2527-2579
>>Chess Tiger 15 ELO 2489-2539
>>FritzInBahrain ELO 2485-2517
>>Deep Junior8EM ELO 2611-2673
>>Deep Junior8ZX ELO 2514-2559
>>Gambit Tiger 2 ELO 2480-2526
>>Ruffian 2.1.0  ELO 2502-2518
>>The King 3.33  ELO 2513-2543
>>Shredder 7     ELO 2476-2544
>>Pro Deo 1.0    ELO 2457-2479
>>3d Schach 4.0  ELO 2510-2528
>>Fritz 5.32     ELO 2393-2407
>>Fruit 2.0      ELO 2515-2558
>>Aristarch 4.5  ELO 2465-2479
>>Gandolf 6.0    ELO 2492-2524
>>Shredder 7SE   ELO 2492-2528
>>Junior 7       ELO 2461-2487
>>Ruffian 1.0.5  ELO 2464-2506
>>Chess Tiger 14 ELO 2496-2532
>>Hiarcs 8Bareev ELO 2463-2503
>>Schachgenie1.0 ELO 2539-2557
>>List 5.12      ELO 2444-2452
>>The King 3.23  ELO 2460-2478
>>Crafty 19.01   ELO 2350-2362
>>Fritz for Fun3 ELO 2465-2475
>>Deep Sjeng 1.6 ELO 2447-2451
>>The King3.23r1 ELO 2544-2550
>>Hiarcs 8       ELO 2457-2463
>>
>>Engine Rankings
>>#	Engine	    Elo Range Avg.
>>
>>1.	deep juniorem	2642
>>2.	deep fritz 7	2633
>>3.	deep fritz 8	2612
>>4.	shredder 8	2608
>>5.	deep junior 9	2596
>>6.	fritz 7	        2578
>>7.	x3d fritz	2560
>>8.	junior 9	2553
>>9.	hiarcs 9	2549
>>10.	schachgenie 1.0	2548
>>11.	the king 3.23r1	2547
>>12.	fritz 8	       	2543
>>13.	shredder 7.04	2551
>>14.	deep junior 8	2540
>>15.	fruit 2.0	2537
>>16.	deep junior8.zx	2535
>>17.	the king 3.33	2528
>>18.	3d schach 4.0	2519
>>19.	chess tiger 15	2514
>>20.	chess tiger 14	2514
>>21.	shredder 7.se   2510
>>22.	shredder 7	2510
>>23.	ruffian 2.1.0	2510
>>24.	gandalf 6.0	2508
>>25.	junior 8	2508
>>26.	gambit tiger2.0	2503
>>27.	fritzinbahrain	2501
>>28.	ruffian 1.0.5	2485
>>29.	hiarcs8 bareev	2483
>>30.	junior 7	2474
>>31.	aristarch 4.50	2472
>>32.	fritz for fun 3	2470
>>33.	the king 3.23	2469
>>34.	pro deo 1.0	2468
>>35.	hiarcs 8	2460
>>36.	deep sjeng 1.6	2449
>>37.	list 512	2448
>>38.	fritz 5.32	2400
>>39.	crafty 19.01	2356
>>
>>For the list above, you can see the top 5 look out of order. That is not the
>>case. The average of the program's ELO RANGE shows it potential on strength and
>>consistency.
>>
>>Top 5
>>#	Engine	   ELO RANGE  ELO RANGE Avg.
>>1. Deep Junior8EM  2611-2673     2642
>>2. Deep Fritz 7    2581-2685     2633
>>3. Deep Fritz 8    2574-2650     2612
>>4. Shredder 8      2587-2629     2608
>>5. Deep Junior 9   2573-2624     2596
>>
>>You might ask why is Deep Junior 8EM so high on the list. I think a lot of it
>>had to do with the hardware and books used when it played it's 28,000 + games.
>>
>>You might ask why is Fritz 7 ranked higher than Fritz 8? That is because at
>>Fritz 7's ELO RANGE is a higher and the ELO RANGE average makes it stronger.
>>It's just up to you to provide the right book, hardware, settings, etc. Then you
>>might be at the higher end of the ELO RANGE.
>>
>>You also might ask, why is Shredder 8 in 4th? Even with 450,000 more games than
>>any other engine it's potential never became as high of the others. With the
>>right book, hardware, settings, etc., it could be stronger then the others most
>>of the time. it's ELO RANGE average is still lower then the rest, as of now :)
>>
>>Finally, you might ask, what if all the programs played at it's highest value in
>>the ELO RANGE? Then hypothetically, Deep Fritz 7 would win playing at 2685.
>>
>>If everything was ran on a uniform system, Same hardware, book, settings, then
>>we could give it an approximate rating. Since we all know it's not, an ELO RANGE
>>should be the solution.
>>
>>I hope to see this used more often.
>>
>>Joshua Haglund
>>toneewa@yahoo.com
>
>good morning joshua its hard enough with the present testing tell how strong
>engine is. the testers now are doing a good job of tell about how strong an
>engine is.i prefer to keep it simple as possible.i have a fast computer and a
>slow computer and get about the same results with both. i get diff.results with
>diff.books.i get diff.results with times controls. with 100 games tested on same
>computer the two engines i tested get the same results as most of the other
>tests posted here.
>
>gerold.

Good Morning Gerold,

It's simple, just keep playing, at playchess.com and the server does the rest of
the calculations.

http://playchess.com/stats/engineranking.htm

Updated daily.

Joshua Haglund
toneewa@yahoo.com



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.