You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
/// Compute the indicator of efficiency to win, which means the minimum number you should memorize on the situation both of players always choose one of the best move list.
381
-
///
382
-
/// [level] is search depth. <br>
383
-
/// If [onlyBestScoreLink] is true, only compute the best score links on the current moves.
384
-
///
385
-
/// This is a Dart level function, and unique feature to libedax4dart.
386
-
///
387
-
/// btw, you can also see the graph of tree generated by this function, with using `BestPathNumWithLink.root.exportGraphvizDotFile()`. <br>
388
-
/// e.g. on a light book, when current moves is 'f5f6', you can get a graph like this([graphviz online](https://bit.ly/3edxWGO)).
389
-
///
390
-
/// This function is not only experimental but also __advanced__. <br>
391
-
/// __You must understand [the book structure of edax](https://choi.lavox.net/edax/book) and following important notice list__.
392
-
///
393
-
/// * Because internal each node lookup is stopped when book has no links, the more lower the accuracy of your book is, the more lower this feature accuracy is.
394
-
/// * In addition, if your book has some depth links (e.g. 24, 30, 40, ...), shallow depth link (≈ path is few) can be taken with a reasonable probability.
395
-
/// * __In a word, if the accuracy of your book is low, you shoudn't use this function__. For you reference, if your book has N GB and the depth is more than 30, this feature can probably be inidicator.
396
-
/// * This function only lookup the best score links. So, this indicator can't consider easy win links which is _not_ best score. In reversi game, the situation can sometimes be found in my experience as a player.
397
-
/// * __In a word, as you know, this indicator isn't perfect. This is just a indicator__.
398
-
/// * The depth of this feature depends on [level] or your book.
399
-
/// * The moves which meet up with anoter moves is counted respectively.
400
-
/// * btw, symmetric moves is counted 1 because of edax book structure.
401
-
/// * O(k^N). slow.
402
-
/// * TODO(developer): consider following implementation.
0 commit comments