Komodo Dragon 2 released.

[Damir]

Do you believe it would be a good idea for you to stop spewing garbage every time you open your mouth ?

[mjlef]

Thanks for the new version Larry! Unfortunately when I try and load the new version I get this message:
The requested URL dragon-2-ce44fe7f.zip was not found on this server.....

Try copying the URL directly to the browser. I have recently notice some email systems put a wrapper around the url (like hotmail), but copying the url into a browser seems to work. If you still have trouble, email us and I will see what I can do.

Mark

[mjlef]

Hi, I have a few questions regarding the MCTS mode that I was wondering if you could answer:

1. Does it support tree-reuse?
In Leela, the search tree might look like
1. e4
-> ...e5
-> ...c5
1. d4
-> ...nf6
-> ...d5

Rather than simply rely on the TT/NN cache, if e4 is played, Leela will just drop d4 (and other) branches of the search tree and preserve the entirety of the 1. e4 search tree. This can occasionally save a lot of compute.

2. Does it support smart pruning?
Leela can calculate whether--at the current nodes per second, with some margin--whether it's remotely probable that the MCTS algorithm will ever switch to a new move within the allocated time budget. If it's not possible (if the current best move has too many visits for any other move to catch up) it will simply play the best move.

This just helps improve TM a bit. It's nothing super useful as an end user, but I was curious.

3. Do you have any plans to support Nibbler-compatible output?
I've asked this before, but I figure that it's worth asking again. Sorry if this is obnoxious

Nibbler is a chess GUI for MCTS engines that displays a lot of useful analysis information. It is very similar to Go analysis GUIs, like Lizzie. Support for Nibbler would be really nice.

Thank you so much for your patience answering these questions, and congratulations to you and the K team on another release!

1. tree reuse: we actually programmed in tree reuse, but found it did not help us, maybe due to the nature of highly selectiv trees in the short searches we use to estimate policy and win probabilities.

2. We have had that idea in the MCTS versions for a long time now, and it does help.

3. I have not looked into nibbler yet so it is not supported. Given time and interest, we can see. Komodo CHess is a pretty tiny company.

Mark

[Cumnor]

I got around the download problem by downloading again the previous dragon then clicking on the dragon2 link and it worked

[Damir]

Thanks for the new version Larry! Unfortunately when I try and load the new version I get this message:
The requested URL dragon-2-ce44fe7f.zip was not found on this server.....

Try copying the URL directly to the browser. I have recently notice some email systems put a wrapper around the url (like hotmail), but copying the url into a browser seems to work. If you still have trouble, email us and I will see what I can do.

Mark

Thanks Mark. It seems was the browser issue. I was using Microsoft Edge. Once I switched to Google Chrome I had no issue and have now loaded Dragon 2... Thanks for the help!

[Robert Flesher]

Buying it!

[cucumber]

I don't know whether tree reuse as defined in Leela would make sense with Dragon MCTS. They are pretty different. With Dragon MCTS, once we play a move the analysis done is not very relevant as now there is some asymmetry between how we handle opponent's moves and engine moves as well as root moves vs. nonroot moves.

That makes sense! Thank you for your patience w/all these questions!

1. tree reuse: we actually programmed in tree reuse, but found it did not help us, maybe due to the nature of highly selectiv trees in the short searches we use to estimate policy and win probabilities.

2. We have had that idea in the MCTS versions for a long time now, and it does help.

3. I have not looked into nibbler yet so it is not supported. Given time and interest, we can see. Komodo CHess is a pretty tiny company.

Mark

1. Huh, that's surprising! But I guess it shouldn't be too surprising given how many assumptions tree reuse relies on and how different KMCTS must be from A0-style MCTS.

2. Sweet!

3. Totally understandable.

4. Thank you for the answers!

[Robert Flesher]

Thanks for the new version Larry! Unfortunately when I try and load the new version I get this message:
The requested URL dragon-2-ce44fe7f.zip was not found on this server.....

Try copying the URL directly to the browser. I have recently notice some email systems put a wrapper around the url (like hotmail), but copying the url into a browser seems to work. If you still have trouble, email us and I will see what I can do.

Mark

I just purchased Dragon 2, it shows dragon 2 Nov/2020 ???

[Robert Flesher]

Thanks for the new version Larry! Unfortunately when I try and load the new version I get this message:
The requested URL dragon-2-ce44fe7f.zip was not found on this server.....

Try copying the URL directly to the browser. I have recently notice some email systems put a wrapper around the url (like hotmail), but copying the url into a browser seems to work. If you still have trouble, email us and I will see what I can do.

Mark

I just purchased Dragon 2, it shows dragon 2 Nov/2020 ???

I'll admit, this is annoying. URL does not work!

[Lion]

Dragon 2 is now the favorite against Stockfish 13 on typical modern hardware.

Should I read that Dragon 2 is stronger than Stockfish 13?
If yes, any estimation how how much?

Rgds