Coherent 0.2.21: 3701 ± 51 over 10 anchors
Listen to this post6 min
Created by AI from the text of this post and read by the synthetic voice “Sulafat”; not a word-for-word version, and it may contain errors.
Loading…
Created by AI from the text of this post and read by the synthetic voice “Sulafat”; not a word-for-word version, and it may contain errors.Position and speed stay in this browser and are sent nowhere.
The recording could not be played. Open the audio file
Project measurement, not in the register: an anchoring against foreign engines with a listed rating. The figures come from the data delivery (entry M-NNNN One Elo measurement: a build anchored against foreign engines with a listed rating, on a measuring machine of its own. Numbered in the order delivered; the register run it measured is named with it. The series continues the entries E-0001 to E-0012 and starts at M-0013.); the build measured is the one from run T-0195 of 20 October 2026. The reservations below are those delivered by the day of publication; the full and current account is on the playing-strength page.
Coherent 0.2.21 played 4,000 games against 10 foreign engines with a published rating, cinder 0.6.1, halogen 16.0.0, hobbes 3.0, justbot 0.5.0, pzchessbot 7.1, raphael 4.2.0, reckless 0.9.0, renegade 1.3.0, stormphrax 8.0.0 and viridithas 20.0.0, 400 against each at 120+1. The result is 3701 ± 51 (anchored to CCRL Computer Chess Rating Lists, a public ranking of chess engines from games under standard conditions. blitz ratings), ± 5 of it statistical. From 22 October 2026 that is the figure at the head of this site; it replaces the anchoring of 0.2.10 at 3670 ± 52.
The runs
Project measurement, not in the register. 10 horizontal intervals, one per opponent, on a scale of anchored playing strength: cinder 0.6.1 at 3710 with a half-width of 17, halogen 16.0.0 at 3673 with a half-width of 14, hobbes 3.0 at 3822 with a half-width of 18, justbot 0.5.0 at 3690 with a half-width of 17, pzchessbot 7.1 at 3688 with a half-width of 16, raphael 4.2.0 at 3686 with a half-width of 17, reckless 0.9.0 at 3694 with a half-width of 16, renegade 1.3.0 at 3680 with a half-width of 15, stormphrax 8.0.0 at 3689 with a half-width of 16, viridithas 20.0.0 at 3683 with a half-width of 17. Below them the published figure for 0.2.21, 3701 with a band of 51, shaded behind every row. The plus-minus values are half-widths of nominal 95 % intervals; the published band also carries the transfer term.
| Opponent | Listed | Games | Score rate | Difference | From it |
|---|---|---|---|---|---|
| cinder 0.6.1 | 3762 ± 12 | 400 | 42.60 % | −51.60 ± 11.80 | 3710 ± 17 |
| halogen 16.0.0 | 3721 ± 8 | 400 | 43.10 % | −48.10 ± 11.50 | 3673 ± 14 |
| hobbes 3.0 | 3744 ± 12 | 400 | 61.00 % | +77.70 ± 13.50 | 3822 ± 18 |
| justbot 0.5.0 | 3705 ± 14 | 400 | 47.90 % | −14.80 ± 9.90 | 3690 ± 17 |
| pzchessbot 7.1 | 3713 ± 12 | 400 | 46.40 % | −25.20 ± 10.90 | 3688 ± 16 |
| raphael 4.2.0 | 3731 ± 12 | 400 | 43.60 % | −44.50 ± 11.50 | 3686 ± 17 |
| reckless 0.9.0 | 3766 ± 10 | 400 | 39.80 % | −72.20 ± 12.90 | 3694 ± 16 |
| renegade 1.3.0 | 3694 ± 12 | 400 | 48.00 % | −13.90 ± 9.80 | 3680 ± 15 |
| stormphrax 8.0.0 | 3745 ± 11 | 400 | 42.00 % | −56.10 ± 12.00 | 3689 ± 16 |
| viridithas 20.0.0 | 3748 ± 11 | 400 | 40.80 % | −65.00 ± 13.00 | 3683 ± 17 |
All half-widths at the 95 % level, as the delivery states them; the column From it is the anchor plus the measured difference. The listed ratings are from the CCRL Blitz Rating List, single-core rows only, as of 2026-10-20.
How the figure is formed
Recalculated from the anchors: the 10 anchored values, averaged with weights inverse to their variance, give 3698.3 with a statistical half-width of ± 5.2; the delivery rounds it to ± 5. With the measuring side's transfer term the band is ± 51.3; the delivery names 3701 with ± 5 statistical and ± 51 transfer apart, combined as in earlier deliveries 3701 ± 51, and this site publishes that figure. The anchored values contradict one another (chi-square 209.09 at 9 degrees of freedom, bound 16.92; half-widths read as 95 % half-widths).
Against 0.2.10 no anchor was played in both campaigns; the 31 points between the two headline figures are therefore not a measured gain.
What belongs to this figure
- Measured outside the formal registration procedure (project measurement on the project's Arena machine, no oracle finding, Law I).
- The privileges of the measuring agents on the testbench are not yet settled.
- Cancelled opponents do not count: none.
- Own book (8moves_v3) and no endgame tablebases instead of the conditions of the list.
- Statistically 95 % from games and list errors of the anchors; the transfer uncertainty to the conditions of the list (book, endgame tablebases, hardware) is not measured and stands as a separate band.
Without a provenance record in the delivery: cinder 0.6.1, halogen 16.0.0, hobbes 3.0, justbot 0.5.0, pzchessbot 7.1, raphael 4.2.0, reckless 0.9.0, renegade 1.3.0, stormphrax 8.0.0 and viridithas 20.0.0.
Conditions
120+1, book 8moves_v3 with 34,700 positions, each colour pair with an opening of its own, 64 MB hash, one thread, 15 concurrent games, no adjudication, a fixed number of games without SPRT Sequential probability ratio test. Games are played until the accumulated evidence reaches one of two bounds; then the test stops.; 86 losses on time. The measurement is dated 21 October 2026, 22:30 UTC.
Generated from the data delivery. The recalculation is the site's; every other figure is the measuring side's.