Uberdude wrote:I've finally got round to making a video about how to use Lizzie to review games.
Many thanks.
This is an important topic, but one which is daunting to many players, both in terms of computer technique and in terms of improving one's go. You are providing a real service to us all.
I think I got rather carried away with maybe too much Go content rather than focused on Lizzie instruction and ended up going for over an hour, but then discovered Quicktime only lets you upload 15 minutes even if your Youtube account allows longer as mine does. So here are 2 videos for the first 30 minutes. What do you think?
I agree with the sentiment that it would be good to break up the presentation into a number of short videos. Some software that I am learning has a series of tutorials, each of which addresses one point and is 3 min. long. IMO, that's too short, but 5 min. might be good. 15 min. is long enough for someone who is not generally familiar with the material to get lost.
Here's my thoughts so far:
- Sometimes I waffle too much, e.g. should just say "you can use Lizzie to review games", no need to say your games, other games, pro games, joseki etc.
No, that's OK. There is a mnemonic for giving a talk: Say A Few Words. That is, State your idea, Accentuate your idea, For instance your idea, Wind up with your idea. Offering examples is fine. OC, if you don't think that the idea is worth emphasizing, then don't give examples. But I think that you should trust your instincts. You felt like giving examples, and for what you were trying to say, you were probably right.
- I could put a table of contents with time links to skip to the sections.
Excellent! Different people will have different questions at different times. Letting them pick and choose is a good idea.
- Did I address enough the how kyu players can use it well? Maybe I should review a 10k game rather than 4d game as example? Do weaker players end up thinking "Well yeah great, you can use Lizzie because you are 4 dan so can interpret its output, but I'm lost"? I want to help make these tools widely accessible and useful.
That's a big question, and an important one. I think that you need feedback from kyu players about that. I know that a number of top bot self play games have left me feeling bewildered. My response was like that when I was a child learning bridge.
Wow! What happens next? Not trying to understand, but simply grokking it. But not everybody can take that approach or feels that it is useful. People need a guide through the labyrinth.
One thought that I have had is for players to use a version of LZ that is not too much stronger than they are, with the knowledge that sometimes it will suggest bad moves. That way they may be able to stretch their knowledge and understanding without getting overwhelmed. And, with the next higher level version available, they can move on when they are ready. Gradus ad parnassum.
Another thought is to have a video or two with Q&A with a kyu player or two. That format is a pretty good one.
----
This may just be me, but I do not find those long variations helpful, especially when they are presented instantaneously. I know that the deeper you go, the more questionable they become. Sometimes they give a big picture, but often my response is Huh? If I were preparing a presentation using variations, I would have LZ explore each move. You don't have to go very deep to discover that it will recommend a different play than you see in the instant variation that it shows. What's the point of showing several moves that go down the "wrong" path from there?
)