This post explores the need to adjust default parameters of chess engines to play correspondence chess. As part of playing CC the need to understand and control chess engines is necessary.
Familiarizing yourself with available parameters for engines you use is important. Some engines default options are to play against humans without their full strength while others are set to play other engines.
Although each engine, hardware, programs protocol and graphic interface combinations are specific, there are general observations that can be made regarding how to set CC configurations.
In order to understand what your engine is doing you should study the engine’s options. Two basic examples are engine’s depth and contempt. Many parameters exist and specify other options.
Depth parameters determine if variations’ analysis is shallow and wide or deep and narrow. Usually deep searches provide more strategic accuracy. Shallow searches help in unclear positions being exhaustive.
Contempt parameters set the engine’s way of playing by avoiding drawn positions or adopting more aggressive play. Engines can define contempt as other options differently, so make sure you check.
By choosing different depth and contempt for example, you can adapt your engines to different game stages or tasks to provide you with input as you consider different analysis situations.
By understanding the engine’s options, CC play can combine different engine configurations and engines output to conduct chess analysis controlling for the factors available to parametrize the engines.
The use of multiple engines that you have configured serves as a robust way to analyze a game or position as you assign each engine to study it emphasizing a deep or shallow search for example.
Before starting to use a new configuration in actual play it is important to test it against a position to make sure you see first hand how the engine behaves and if there are any issues.
Ultimately, this tailoring and bespoke settings ensure you are making the best of the engines and helps develop a better interpretation of its output and recommendations to use it in competitive CC.