A few questions about how this works about python-chess-annotator

[lazydroid]

https://en.wikipedia.org/wiki/Numeric_Annotation_Glyphs

• $2 poor move or mistake (traditional "?")
• $4 very poor move or blunder (traditional "??")
• $6 questionable or dubious move (traditional "?!")

etc.

[vesper8]

Excellent thank you! I guess these are the NAGs I saw many references to. Would be helpful to drop that wikipedia link in the Readme IMO.. not everyone is familiar with this terminology

[lazydroid]

This repo is not maintained since 2019 (it says this clearly on README page), so the chances are slim. I'll make a fork and fix a few things here and there, but cannot promise if the author will be willing to put them back into the codebase.

[rpdelaney]

Thanks for jumping in with the explanation. See also #15. However, python-chess has changed a lot since I originally wrote this script, so maybe the behavior has changed there. Edit: Also, #14

[rpdelaney]

The first part I'm confused about is.. what does do exactly? It seems like it goes through the whole game and makes note of missed opportunities.. on both sides?

It compares the engine evaluation of the played move with the engine evaluation of the engine's bestmove. If the played move is sufficiently worse, it adds an annotation with the engine's primary variation. Which NAG is used is determined by how bad the move was.

Is that right? What if I only want to see missed opportunities from one side's perspective.. such as from white's perspective.. is there an option for that?

This script does not have that feature, no. I've leveled up considerably in python since I wrote this (it was actually a project to teach myself python in the first place, so it's the very first serious python application I ever did) and I regard the current state of this code as frankly more effort to maintain than it's worth, let alone extend. Every now and then I kick around the thought of doing a python-chess-annotator-ng from scratch with more of the python skills I've developed since, but other things always get in the way.

[rpdelaney]

Lastly, is it possible to pass it an additional option that would return the score after every single move?

This is also not a feature, although it wouldn't be a tremendous amount of work to extend the script to do that. You would probably want to add a command-line switch to change the value of NEEDS_ANNOTATION_THRESHOLD to 0 (or a user-defined value).

Edit: Bear in mind that a value of 0 would mean you would also get annotations when you played the engine's bestmove, which might not make a lot of sense.