Changes

[[FILE:Hurricane Betsy.jpg|border|right|thumb|240px| Hurricane Betsy <ref>[https://en.wikipedia.org/wiki/Hurricane_Betsy Hurricane Betsy] in the [https://en.wikipedia.org/wiki/Gulf_of_Mexico Gulf of Mexico] in September of 1965 as taken by the [https://en.wikipedia.org/wiki/Television_Infrared_Observation_Satellite TIROS-8] weather satellite. Source: [https://en.wikipedia.org/wiki/National_Oceanic_and_Atmospheric_Administration NOAA] [http://www.photolib.noaa.gov/ Photo Library], September 4, 1965</ref> ]]

'''Betsy''',<br/>a [[WinBoard]] compatible chess engine by [[Landon W. Rabern]], written in [[C]] with a little [[Assembly]], released in September 2000 <ref>[https://web.archive.org/web/20131109133342/http://wbec-ridderkerk.nl/html/details1/Betsy.html Betsy] from [[WBEC|WBEC Ridderkerk]] ([https://en.wikipedia.org/wiki/Internet_Archive Internet Archive])</ref>.

According to a former [[Arena]] site, Betsy was the first published chess engine able to play [[Chess960]] <ref>[https://chess-tigers.de/index_news.php?id=308&rubrik=4&PHPSESSID=d71dfe17e7e8aae16adce6f8fb284410 Der Chess960-Express ist nicht mehr aufzuhalten], [https://chess-tigers.de/cttc_main.php?rubrik=7 Chess Tigers Training Center], July 20, 2005 (German)</ref>, and was therefore Arena partner engine.  =Neural Networks=Quote by [[Landon Rabern]] <ref>[https://landon.github.io/#about landon rabern - about] (2019)</ref> As a child, I was obsessed with machine intelligence. I coded a strong chess AI (codenamed Betsy) and experimented with using [[Neural Networks|neural networks]] in Betsy, both for the [[Evaluation|static evaluation]] at [[Leaf Node|leaf nodes]] and within the [[Search Tree|tree]] for [[Pruning|pruning]]. The networks learned from self-play to get about as good as my hand-tuned functions (discounting the slowdown incurred by [[Pawn Advantage, Win Percentage, and Elo|sigmoid evaluation]]). I concluded that to do better, I would need to use raw game state data instead of the set of features I preselected as network inputs; unfortunately, this was 2000 and I did not have nearly enough processing power to do so. =C# Port=As of 2014, Landon Rabern started to port Betsy to [[C sharp|C#]], available as [[:Category:Open Source|open source engine]] under the [[Massachusetts Institute of Technology#License|MIT License]] at [https://en.wikipedia.org/wiki/GitHub GitHub]. All the major components of a chess engine are there, just not a tuned [[Evaluation|evaluation]] <ref>[https://github.com/landon/Chess GitHub - landon/Chess: Beginnings of a port of Betsy to C#]</ref>. The board class embeds an [[8x8 Board|8x8 board]] and the [[Bitboard Board-Definition|Bitboard board-definition]], and implements [[Rotated Bitboards|rotated bitboards]].The [[Search|search]] is implemented using a derivation chain of classes, the abstract base ''Brain'', ''TranspositionTableBrain'' ([[Transposition Table]], [[Iterative Deepening]]), and the concrete ''BasicAlphaBetaBrain'' ([[Alpha-Beta]]), ''NullMoveBrain'' ([[Null Move Pruning]]) and ''MTDfBrain'' ([[MTD(f)]]) classes.

* [https://www.stmintz.com/ccc/index.php?id=125117 Re: C or C++ for Chess Programming?] by [[Landon Rabern]], [[CCC]], August 18, 2000* [https://www.stmintz.com/ccc/index.php?id=128512 Betsy 5.0 is now winboard compatable] by [[Landon W. Rabern]], [[CCC]], September 06, 2000* [https://www.stmintz.com/ccc/index.php?id=178491 Betsy 5.26] by [[Landon W. Rabern]], [[CCC]], July 06, 2001* [https://www.stmintz.com/ccc/index.php?id=302797 Re: FRC_TheBaron_101 Vs Fritz8 (Castling vs Not Castling rules)] by [[Landon Rabern]], [[CCC]], June 24, 2003

* [https://web.archive.org/web/20131109133342/http://wbec-ridderkerk.nl/html/details1/Betsy.html Betsy] from [[WBEC|WBEC Ridderkerk]] ([https://en.wikipedia.org/wiki/Internet_Archive Internet Archive])