Stackruns

Coherent 0.1.113: 3410 ± 52 over 8 anchors

Published Updated chess-reporter Project: Coherent Chess anchoringmeasurement

Created by AI from the text of this post and read by a synthetic voice; not a word-for-word version, and it may contain errors.

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-0131 of 27 September 2026. The reservations below are those delivered by the day of publication; the full and current account is on the playing-strength page.

Coherent 0.1.113 played 3,600 games against 8 foreign engines with a published rating, Velvet 3.3.0, Xiphos 0.6, Princhess 0.22.0, Frozenight 6.0.0, Black Marlin 7.0, akimbo 0.7.0, Winter 4.0 and Minke 4.0.0, 450 against each at 120+1. The result is 3410 ± 52 (anchored to CCRL Computer Chess Rating Lists, a public ranking of chess engines from games under standard conditions. blitz ratings), ± 8 of it statistical. From 28 September 2026 that is the figure at the head of this site; it replaces the anchoring of 0.1.112 at 3394 ± 52.

The runs

Coherent 0.1.113: the anchors and the published figure — Project measurement, not in the register Project measurement, not in the register. Eight horizontal intervals, one per opponent, on a scale of anchored playing strength: Velvet 3.3.0 at 3428 with a half-width of 25, Xiphos 0.6 at 3428 with a half-width of 20, Princhess 0.22.0 at 3457 with a half-width of 28, Frozenight 6.0.0 at 3405 with a half-width of 18, Black Marlin 7.0 at 3399 with a half-width of 20, akimbo 0.7.0 at 3368 with a half-width of 23, Winter 4.0 at 3384 with a half-width of 23, Minke 4.0.0 at 3419 with a half-width of 23. Below them the published figure for 0.1.113, 3410 with a band of 52, shaded behind every row. The plus-minus values are half-widths of nominal 95 % intervals; the published band also carries the transfer term. 3340 3360 3380 3400 3420 3440 3460 3480 3500 Velvet 3.3.0 3428 ± 25 Xiphos 0.6 3428 ± 20 Princhess 0.22.0 3457 ± 28 Frozenight 6.0.0 3405 ± 18 Black Marlin 7.0 3399 ± 20 akimbo 0.7.0 3368 ± 23 Winter 4.0 3384 ± 23 Minke 4.0.0 3419 ± 23 0.1.113 published 3410 ± 52 playing strength anchored to CCRL blitz ratings
Each row is one opponent: the value anchored through it, with its 95 % half-width. The bottom row is the published figure for 0.1.113, 3410 ± 52; the shaded strip carries its band, transfer term included, across all rows. Project measurement, not in the register.
OpponentListedGamesScore rateDifferenceFrom it
Velvet 3.3.03329 ± 1445063.89 %+99.11 ± 20.323428 ± 25
Xiphos 0.63353 ± 745060.67 %+75.28 ± 19.093428 ± 20
Princhess 0.22.03357 ± 1645064.00 %+99.95 ± 22.483457 ± 28
Frozenight 6.0.03362 ± 945056.11 %+42.68 ± 15.483405 ± 18
Black Marlin 7.03468 ± 1045040.22 %−68.83 ± 17.773399 ± 20
akimbo 0.7.03470 ± 1445035.78 %−101.63 ± 17.823368 ± 23
Winter 4.03480 ± 1045036.56 %−95.78 ± 20.243384 ± 23
Minke 4.0.03485 ± 1645040.56 %−66.42 ± 16.383419 ± 23

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 as of 2026-09-12, 2026-09-21.

How the figure is formed

Recalculated from the anchors: the 8 anchored values, averaged with weights inverse to their variance, give 3408.6 with a statistical half-width of ± 7.7; the delivery rounds it to ± 8. With the measuring side's transfer term the band is ± 51.6; the delivery names 3410 ± 52, and this site publishes the delivery's figure.

Against 0.1.112 six anchors were played in both campaigns, Velvet 3.3.0, Xiphos 0.6, Princhess 0.22.0, Frozenight 6.0.0, Black Marlin 7.0 and akimbo 0.7.0. Per anchor the difference of the two measured differences is +31.08 ± 29.77, +33.39 ± 28.13, +15.73 ± 29.94, −8.65 ± 23.39, +16.21 ± 25.83 and +16.18 ± 26.76; combined, +15 ± 11. The per-anchor values are compatible with one another (chi-square 6.70 at 5 degrees of freedom, bound 11.07). That is the measured change: the opponent's listed rating and the transfer term cancel in it. The 16 points between the two headline figures are not a measured gain; they carry both bands in full.

Coherent 0.1.113 against 0.1.112: the change per shared anchor — Project measurement, not in the register Project measurement, not in the register. Six horizontal intervals, one per anchor played in both campaigns, on a scale of Elo difference against 0.1.112: Velvet 3.3.0 at +31 with a half-width of 30, Xiphos 0.6 at +33 with a half-width of 28, Princhess 0.22.0 at +16 with a half-width of 30, Frozenight 6.0.0 at −9 with a half-width of 23, Black Marlin 7.0 at +16 with a half-width of 26, akimbo 0.7.0 at +16 with a half-width of 27. Below them the combined value, +15 ± 11. A dashed vertical line marks zero, no change. −40 −20 0 +20 +40 +60 +80 Velvet 3.3.0 +31 ± 30 Xiphos 0.6 +33 ± 28 Princhess 0.22.0 +16 ± 30 Frozenight 6.0.0 −9 ± 23 Black Marlin 7.0 +16 ± 26 akimbo 0.7.0 +16 ± 27 combined +15 ± 11 difference of the two measured differences against 0.1.112, in Elo
Per anchor played in both campaigns, the difference of the two measured differences against 0.1.112; the bottom row combines them to +15 ± 11. The dashed line is zero: no change. The opponent's listed rating and the transfer term cancel in each row. Project measurement, not in the register.

What belongs to this figure

  1. Measured outside the formal registration procedure. Every result line of this measurement path is recorded twice, independently, with its checksum.
  2. The calibration of the total band is not confirmed from the complete raw data.
  3. The evidence for the scale showing that this book maps one-to-one to the list at 2+1 comes from a round robin with only three openings per pairing across ten colour-swapped rounds. The grouping is not accounted for in the supplied calculation (intraclass correlation around 0.30, factor 1.93), so the evidence is roughly half as strong as reported. This is a reservation regarding the model assumption, not regarding the error calculation: Coherent's own anchoring runs used a distinct opening for each colour pair (225 of 225 per anchor), leaving the statistical half-width of 7.7 unaffected.
  4. Statistical half-width plus transfer uncertainty. A round robin among the five anchors showed that a single published rating under foreign conditions deviates; averaged across eight anchors, ± 51 of this remains at the 95 percent level, which together with the ± 7.7 from games and list errors yields ± 52.

Without a provenance record in the delivery: Velvet 3.3.0, Xiphos 0.6, Princhess 0.22.0, Frozenight 6.0.0, Black Marlin 7.0, akimbo 0.7.0, Winter 4.0 and Minke 4.0.0.

Conditions

120+1, book 8moves_v3 with 34,700 positions, each colour pair with an opening of its own, 64 MB hash, one thread, 30 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.; 0 losses on time. The measurement is dated 28 September 2026, 09:40 UTC.

Generated from the data delivery. The recalculation is the site's; every other figure is the measuring side's.