Stats and more stats

Some people, like me, like to see stats.

So here we go:

In the opening book, there’s 277 568 921 positions. 14,02% (38901665) of them have already been computed by an Azor.
In this 38901665 calculated positions, 34,1% are WIN or LOST positions.

To compute a position, Azor during his search , look at about 1.2 millions positions. Less if it find a win path or a lost.
On a 3.2 GHz processor , Azor’s speed search it’s about 400 000 position per second. More than 50 different computer had running an Azor by 20 different people.

In average in a suicide game , each player have to choose between 11,5 different moves.

About the database, there’s about 10 tables, a total of 630 106 946 rows in those table.
The current size of the database: 41.2 Gb of data and the index (to improve speed) size is 66.9 Gb . So a total of 108.1 Gb.
In average 476 queries on executed per second on the database.
Some procedure / function have been programmed directly on the database, about 900 line of code.
The database is running on a AMD 6400+ (3.GHz dual-core) with 4 Gb of ram.

Most of the code needed for the project are in Java. Azor, main-server, Gore server, SuicideViewer, all the back-end, etc are made in java.
More than 10 000 line of code are made, separated in 214 classes and 45 packages. 7 external libraries are also use.

About the website, made in PHP, include about 2000 line of code. During the last month, the website have been visited by 663 unique visitor. The bandwidth needed his about 4.5 Gb by month. By far, it’s the opening book section that is the most popular, more than 86% of the page view.

It’s cost about 350$ USD to run the project, and that doesn’t included the electricity cost to run servers.

If you want to know some others stats, you can request it via comments.


3 réflexions sur “Stats and more stats

  1. Out of the 65.9% calculated positions that are neither WIN nor LOST, how many are « terminal » in the sense that they don’t inherit their score from a child?

    I can clarify, if my question is not clear

  2. Also, out of the 34.1% of calculated positions that are WIN-LOST how many are children of positions that are LOST-WIN and how many are not (i.e. how many are at the top)?

  3. In case , you didn’t got my answer via chat.

    Last night, there was 13579709 win-lost calculated positions in the database. 10083568 of them have children (so about 74,25%)


Entrez vos coordonnées ci-dessous ou cliquez sur une icône pour vous connecter:


Vous commentez à l'aide de votre compte Déconnexion /  Changer )

Photo Google

Vous commentez à l'aide de votre compte Google. Déconnexion /  Changer )

Image Twitter

Vous commentez à l'aide de votre compte Twitter. Déconnexion /  Changer )

Photo Facebook

Vous commentez à l'aide de votre compte Facebook. Déconnexion /  Changer )

Connexion à %s