The following is basicly a feature request but parts of it might already be possible with an engine and other tools?! (Is it?)

Besides merely recalling the repertoire lines using CPT's training module I'd also like to play/train my repertoire against an engine with the goal of getting at least an even position against "the beast" up until, say, 5 or 10 moves after leaving the repertoire. So I wouldn't play complete games but only "some more moves" to see if I understand and can handle the position even against the strongest possible opponent. That's basicly the idea.

Some additional features would make this even more interesting and helpful:

• The engine does not always play the complete repertoire lines but deviates now and then, using either a reasonable alternative which is not (yet?!) part of my repertoire or which is even a bad move and I had to find the "punishment". So it does not always play it's main lines but mixes things up to keep me challenged


• Optionally I would play as long as I hold at least an even position - and see how far I can get against the beast.
• After each session the engine would mark my blunders or suboptimal moves - just like the analyze feature of other tools.
• It could even be integrated into CPT's auto-analyze module to add interesting/tricky lines to my repertoire.

Any thoughts?

I'm new to CPT. But I've read the user guide twice and tried most of the functions. It seems to me that you can do this at least two ways.

The first is how I do it. I play online to practice my opening. If the opponent plays a move not in my opening, then I go to CPT and while in the repertoire module, I add the move.

I always have the stockfish engine running (I use Multi PV = 5 to get the best choices for a reply). If the moves are all within say 10 centipawns, I pick the one I like based on how I play my opening even if it's not the best one.

(Note: I will pick an inferior move as long as I don't lose more than 20 centipawns. This is because I can simplify my opening tremendously. For example, my black opening is just a general KID/Pirc combo where I will be up to -.60 but only play two or three different lines against all replies. The reason is that as a Class level player, tactics and blunders decide almost every game. Centipawn improvements per move are for masters. I want a solid position that's easy to learn).

Back to looking at the engine window. If there is one move that is far superior, than that's an indication that the new opponent move was a bad move. The engine window shows the move that will "punish" the opponents bad move. I may continue to play manually against the computer in this way (not looking at the screen until I decide my move).

In moving for the engine, I might not make the engines best move for it just to see if I can figure out the "punishing" move on my own. Then I will look at the engine window to see how I did.

Over time, I have added dozens of positions to my opening. I like this way because I learn new positions one at a time rather than downloading a big set and then training them.

---

The second way is how CPT was intended to be used. There are great functions that can be found in the manual pp 78-82. Try importing your games that you have played online and use the "Analyze Opening" and "Run Min-Max" to get a good evaluation of the game.

Even better, use the "Run against games" and CPT will tell you where you didn't play your opening move and more importantly, where the opponent played a move not in your opening so you can add it and evaluate your reply move.

Playing around with these functions has for me been the best part of learning the power of CPT. For each game I play online, I've added at least one new position to my opening and learned how to play my best moves in reply. Using my first way, it didn't take long to get a good set of opponent moves out to 10 moves of my opening.