bsamseth / Goldfish
92%
master: 92%

LAST BUILD BRANCH: action-build
DEFAULT BRANCH: master
Repo Added
Files 36
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst

LAST BUILD ON BRANCH search-refactor
branch: search-refactor
CHANGE BRANCH
x

325

Pull #27

travis-ci

web-flow 
Add games for v1.9.0

Preliminary testing indicates a great improvement of ~100 ELO
compared to previous versions. All test matches were bullet,
1 min no increment. The games were played _without_ arbitration,
so all games where played out to their end. This was to verify that
engines could actually perform a checkmate when given the chance.

Will add more games for future revisions,but this seems quite
clearly to be a positive change. Strangely, on of the version
v1.7.0, v1.8.2 or v1.9.0 crashed after 100 or so rounds. Could be
a random thing, but should update chester package to handle this
with more information in the future.

Hard to tell exactly which part caused the increase, since this PR
contains quite a few bugfixes pluss changes. In any case, nice to see
that the now less buggy version plays better!

Head to head statistics:

 1) Goldfish v1.9.0 2353.1 :    430 (+201,=159,-70),  65.2 %

    vs.                    :  games (   +,   =,  -),   (%) :    Diff
    Goldfish v1.7.0        :    215 ( 101,  84, 30),  66.5 :  +101.6
    Goldfish v1.8.2        :    215 ( 100,  75, 40),  64.0 :  +119.1

 2) Goldfish v1.7.0 2251.5 :   1699 (+383,=989,-327),  51.6 %

    vs.                    :  games (   +,   =,   -),   (%) :    Diff
    Goldfish v1.9.0        :    215 (  30,  84, 101),  33.5 :  -101.6
    Goldfish v1.7.1        :    250 (  44, 168,  38),  51.2 :    +7.7
    Goldfish v1.8.0        :    150 (  32,  84,  34),  49.3 :   +14.5
    Goldfish v1.8.1        :    500 ( 114, 305,  81),  53.3 :   +17.4
    Goldfish v1.8.2        :    214 (  48, 139,  27),  54.9 :   +17.5
    Goldfish v1.6.0        :    160 (  47,  85,  28),  55.9 :   +32.8
    Goldfish v1.5.1        :    160 (  43, 102,  15),  58.8 :   +82.7
    Goldfish v1.5          :     10 (   4,   5,   1),  65.0 :   +92.9
    Goldfish v1.4          :     10 (   4,   6,   0),  70.0 :   +98.0
    Goldfish v1.3          :     10 (   5,   5,   0),  75.0 :  +120.8
    Goldfish v1.2          :     1... (continued)
Pull Request #27: Search Refactor & Verification

1582 of 1725 relevant lines covered (91.71%)

108757406.96 hits per line

Relevant lines Covered
1725 RELEVANT LINES 1582 COVERED LINES
108757406.96 HITS PER LINE
Source Files on search-refactor
Coverage ∆ File Lines Relevant Covered Missed Hits/Line

Recent builds

Builds Branch Commit Type Ran Committer Via Coverage
325 search-refactor Add games for v1.9.0 Preliminary testing indicates a great improvement of ~100 ELO compared to previous versions. All test matches were bullet, 1 min no increment. The games were played _without_ arbitration, so all games where played out to thei... Pull #27 web-flow travis-ci pending completion  
324 search-refactor Add games for v1.9.0 Preliminary testing indicates a great improvement of ~100 ELO compared to previous versions. All test matches were bullet, 1 min no increment. The games were played _without_ arbitration, so all games where played out to thei... push bsamseth travis-ci pending completion  
322 search-refactor Remove unused import push bsamseth travis-ci pending completion  
323 search-refactor Remove unused import Pull #27 web-flow travis-ci pending completion  
321 search-refactor Fix bug with move_count implementation. Forgot to reset on copy construction. Pull #27 web-flow travis-ci pending completion  
320 search-refactor Fix bug with move_count implementation. Forgot to reset on copy construction. push bsamseth travis-ci pending completion  
319 search-refactor Bump version number to v1.9.0 Pull #27 web-flow travis-ci pending completion  
318 search-refactor Bump version number to v1.9.0 push bsamseth travis-ci pending completion  
316 search-refactor Add test for playing matches. No assertions made other than demanding that no errors occur during play, such that a result is found. Not a strict test, but if for some reason the game clock logic breaks, this might catch it. push bsamseth travis-ci pending completion  
317 search-refactor Add test for playing matches. No assertions made other than demanding that no errors occur during play, such that a result is found. Not a strict test, but if for some reason the game clock logic breaks, this might catch it. Pull #27 web-flow travis-ci pending completion  
See All Builds (366)