A Rails app for maintaining your chess game database and gleaning insights with powerful statistical analysis.
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
app
bin
config
db
lib/tasks
public
spec
.gitignore
.rspec
Gemfile
Gemfile.lock
README.md
Rakefile
config.ru
package.json

README.md

Chess Sense

WORK IN PROGRESS. A Rails app for maintaining your chess game database and gleaning insights with powerful statistical analysis.

Installation

  1. Consult Gemfile for Ruby and bundler versions, install those.
  2. Depends on PostgreSQL v9.6+, install it.
  3. Create a database.yml from the database_example.yml
  4. Bundle
  5. Set up with rails db:create db:migrate db:seed
  6. See if specs are green with rspec

Deployment

Production runs two processes - Puma webserver and Sidekiq worker, so make sure to (re)boot those on deployments.

Currently the app is hosted on Heroku @ TODO.

ChessSense querying

The query interface is located in the Sense model and is modeled somewhat after SQL.

  • You have TELL ME statements that let you pick an insight you want answered about the given dataset.
  • You have WHERE statements that help you select the subset of games you are interested (who played, what opening was used etc.)
  • You have LOOK AT statements that let you narrow down the dataset to some count or percent

Limitations

1. Querying games played by others
Currently you only can query your database games either by all of your nicknames or by one specific nickname (of an opponent.)
I could allow defining a list of nicknames of your arch-nemesis in the profile form and then looking up games matching those...

2. It is tough defining what a "piece trade" is
Currently for "queen_captures" insight the system assumes that a "trade" is a sequence of moves where players each lose a piece of equal value (Bishops equaling Knights) within 3 moves.

This allows to tell that these were trades (without programming game-situation awareness into queries):

  • a single move queen trade: 8. Qxd8 Kxd8
  • a recapture with a check intermezzo: 8. Qxd8 Bxh2+ (checking the castled white king with a sac) 9. Kh1 (white king steps out of check) Kxd8

Feature roadmap

TELL ME

  • [ ] "size" | How many games are there? # simple count

  • [ ] "results", "Epigene" | What are (player)'s win-draw-loss results?

  • [ ] "openings" | What openings are played?

  • [ ] "queen_captures" | What Queen captures and losses characterize each game from (player)'s perspective?

  • [ ] "positions", "10" | What positions occur the most after the (n)th move? # n at least 10

  • [ ] "captures", "Epigene" | What are (player)'s piece capture outcomes? (Does (player) capture Rooks/Queen with knight, bishops, what is lost?)

WHERE

  • [ ] "played_on", "2018-01-01", "2018-01-31" | played on (start_date) - (end_date)

  • [ ] "played_by", "Epigene", "any" | played by (player) as (white, black, any)

  • [ ] "outcome", "Epigene", "drew" | (player) (won, lost, drew)

  • [ ] "opening", "A00" | (opening (eco code)) is played

  • [ ] "opening_regex", "TODO" | (moveset regex) is played # Haard...

  • [ ] "queen_capture", "Epigene", "initiate" | (player) (initiate, take_back, any) a Queen trade

  • [ ] "capture", "Epigene", "queen", "dark_bishop" | (player) captures (piece) with (piece) # Haaard..

LOOK AT

  • [ ] "all" | all games # easy default

  • [ ] "last", "100" | latest (n) games

  • [ ] "last_percent", "33" | latest (n) percent of games

GENERAL

  • [ ] Users can save query scopes (like played_by(ereatest_genemy))