Life In 19x19
http://www.lifein19x19.com/

Professional advice?
http://www.lifein19x19.com/viewtopic.php?f=10&t=8674		 Page 1 of 4
Author:  daal [ Mon Jul 08, 2013 3:23 am ]
Post subject:  Professional advice?
Wouldn't it be great if you could play a game and have a pro looking over your shoulder and after every one of your moves, the pro would suggest a better one? Well, you can! The only catch is that you have to undo your move and play the move the pro suggested. Want to give it a try?

Here's how: Open up a professional game with the SGF viewer of your choice, jump to move 4 and then decide where you would like to move next. Then see what the professional did. If the professional chose a move different than yours, think about why their move was better ( it was). Then do the same with the next move, etc.

Although you might find that the professional tends to be rather tight-lipped, chances are that once you see the their move, the fault of your own reasoning will become self-evident. I've tried this now a few times, and it's astonishing how often, without saying a word, the professional tells me to adhere to the fundamentals and not try something whacky.

I was going to title this "Free Professional Teacher," but I decided I didn't want to tick everybody off. :)
Author:  snorri [ Mon Jul 08, 2013 3:47 am ]
Post subject:  Re: Professional advice?
Guessing next move (or guessing the next three moves) is always a good exercise.

But the way in which the pro move is "better" is often subtle. Sometimes your move is just as good, but would just lead to a different game. Sometimes your move is better! It's actually quite common. What happens is that the pro will play an odd move because he doesn't think the normal move is quite enough. Then other pros will review the game, analyze it death, and then come to the conclusion that the player should have just played the normal move (i.e., your move) and stayed patient.

Unfortunately, your terse pro will not refute your bad moves. Sadly, that's still up to you... :)
Author:  paK0 [ Mon Jul 08, 2013 5:21 am ]
Post subject:  Re: Professional advice?
Mh, thats actually a nice idea, gotta try it sometimes.
Author:  tchan001 [ Mon Jul 08, 2013 5:21 am ]
Post subject:  Re: Professional advice?
Maybe you should suggest commented games because in some game moves the pros would have preferred to have played some move other than what they had played in the game.
Author:  daal [ Mon Jul 08, 2013 6:32 am ]
Post subject:  Re: Professional advice?
tchan001 wrote:
Maybe you should suggest commented games because in some game moves the pros would have preferred to have played some move other than what they had played in the game.

When pros second guess other pros (albiet with hindsight) they might end up talking more to each other than to you. :)

Besides that, I think that there's something to be said for looking just at stones, and not at words. In my experience, just seeing a stone played that's different than the one I might have played, is enough to let me hear a voice in my mind's ear of someone stronger than myself. I don't mean this in a supernatural way, but rather that seeing the stone often helps me to recall something that I'd been taught before.

Or maybe I should start drinking less coffee. :)
Author:  Polama [ Mon Jul 08, 2013 6:35 am ]
Post subject:  Re: Professional advice?
Try to find games with slower times to minimize the bad moves the pro's play. Hard to guess when they'll play a timesuji in a blitz game ;)

I also wouldn't worry too much about choosing a better move than the pro. If what they play really doesn't work, the opponent should refute it for you. If neither finds the refutation, it's probably safe for amateurs to play. And of course, if you find an interesting position you think was misplayed, that'd be an excellent topic for discussion here.

As for cases where multiple moves are playable, try to come up with a few candidate moves every turn. It's a good habit to get into anyways, and it lets you differentiate between cases where you waffled between your move and the pro move, and where the pro move was totally unexpected.
Author:  Subotai [ Mon Jul 08, 2013 7:26 am ]
Post subject:  Re: Professional advice?
If I have always wondered about professionals reviewing a game after a big match. If I had just made silly reading mistake that caused me to lose the game I wouldn't want to review the game right after because I would be frustrated with myself. Or if you are just too tired after the game I wouldn't want to review it either. Are pro players forced to review the game afterwards? Perhaps only in major tournaments or title matches?
Author:  Bill Spight [ Mon Jul 08, 2013 7:47 am ]
Post subject:  Re: Professional advice?
You can also use GoGOD's GoScorer software, which lets you guess more than once and keeps track of how well you have done. :)
Author:  daal [ Mon Jul 08, 2013 10:16 am ]
Post subject:  Re: Professional advice?
Bill Spight wrote:
You can also use GoGOD's GoScorer software, which lets you guess more than once and keeps track of how well you have done. :)


I've used GoScorer quite a bit too, but what I've tended to focus on was guessing the moves the pro made. What I'm suggesting is that the worthwhile part of the exercise might be to think about the moves as corrections to one's own faulty line of thought.
Author:  snorri [ Mon Jul 08, 2013 10:36 am ]
Post subject:  Re: Professional advice?
Bill Spight wrote:
You can also use GoGOD's GoScorer software, which lets you guess more than once and keeps track of how well you have done. :)


That was may favorite tool when I was starting and still is for the guessing game (unless I'm using a physical board and trying ignore much of a printed kifu :)). I can compare some programs:

GoScorer:

  • Windows Only
  • Commercial (comes with GoGoD CD)
  • Keeps track of percent of guesses correct
  • Provides three kinds of hints: quarter board hint, side-to-side hint, and line hint

MultiGo:

  • Windows Only
  • Free from www.ruijiang.com
  • Can tutor one side or both sides
  • No hints (except using move tree coordinates to cheat) or success statistics

SmartGo:

  • Windows, iPad*, iPhone*
  • Commercial, get from www.smartgo.com or Apple store for Apple versions.
  • Cumulative statistics over multiple games (can be reset as desired)
  • Hints and statistics on wrong move, right move with wrong timing, right area but wrong move

Drago:

  • Windows only
  • Free from www.godrago.net
  • Works on game collections
  • Keeps number of replays and best guess score for each game
  • Can display a "distance to correct move" as a hint

The reason GoScorer is my favorite out of these is that the hint options are just about right. I don't care about cumulative statistics that much because I don't think move guessing is necessarily a great measure of progress.

Unfortunately, I don't know of good Linux programs. It would be cool if eidogo added a guessing feature.

* I am not sure if the iPad and iPhone versions have comparable features for guessing as the Windows version. (Edit: they appear to have it. See smart go site.)
Author:  Splatted [ Mon Jul 08, 2013 10:43 am ]
Post subject:  Re: Professional advice?
I really like this idea. I wonder if you could take it to the next level and use a database instead of a single game? I don't have kombilo so I'm not sure how practical that is but it sounds like a lot of fun.

P.s. Gokifu has a guessing function with a hotter warmer hint system.

http://www.gokifu.com
Author:  Boidhre [ Mon Jul 08, 2013 10:51 am ]
Post subject:  Re: Professional advice?
snorri wrote:
Bill Spight wrote:
You can also use GoGOD's GoScorer software, which lets you guess more than once and keeps track of how well you have done. :)


That was may favorite tool when I was starting and still is for the guessing game (unless I'm using a physical board and trying ignore much of a printed kifu :)). I can compare some programs:

GoScorer:

  • Windows Only
  • Commercial (comes with GoGoD CD)
  • Keeps track of percent of guesses correct
  • Provides three kinds of hints: quarter board hint, side-to-side hint, and line hint

MultiGo:

  • Windows Only
  • Free from www.ruijiang.com
  • Can tutor one side or both sides
  • No hints (except using move tree coordinates to cheat) or success statistics

SmartGo:

  • Windows, iPad*, iPhone*
  • Commercial, get from www.smartgo.com or Apple store for Apple versions.
  • Cumulative statistics over multiple games (can be reset as desired)
  • Hints and statistics on wrong move, right move with wrong timing, right area but wrong move

Drago:

  • Windows only
  • Free from www.godrago.net
  • Works on game collections
  • Keeps number of replays and best guess score for each game
  • Can display a "distance to correct move" as a hint

The reason GoScorer is my favorite out of these is that the hint options are just about right. I don't care about cumulative statistics that much because I don't think move guessing is necessarily a great measure of progress.

Unfortunately, I don't know of good Linux programs. It would be cool if eidogo added a guessing feature.

* I am not sure if the iPad and iPhone versions have comparable features for guessing as the Windows version. (Edit: they appear to have it. See smart go site.)


Kombilo also has a guessing function. It's Python so in theory it runs on anything but getting it to work on OSX is unpleasant.
Author:  daal [ Mon Jul 08, 2013 10:54 am ]
Post subject:  Re: Professional advice?
snorri wrote:
I can compare some programs:


You might also add Kombilo.

Platform: Windows, Linux, Mac OS X
Price: Free - but it comes without a database http://www.u-go.net/kombilo/
Guess mode: "If the guess is wrong, it displays a red rectangle; the rectangle is roughly centered at the position of the next move, and the closer your guess was, the smaller is that rectangle. Furthermore the number of correct guesses and the number of all guesses, as well as the success percentage are given."

Darn, ninja'd
Author:  Kirby [ Mon Jul 08, 2013 12:05 pm ]
Post subject:  Re: Professional advice?
I like this game, but it is more fun after the opening, where there are many valid choices (hence, I suppose, the 4 move clause in the OP).

Even in the middle game, though, this game has a weakness: moves can be tested, but not plans.

There may be multiple good moves from a position, but only if your plan and follow up is good can the result work out. Different pros may respond differently in the same situation.

Still, this is the best way that I know for studying games, and allows for you to test your wits against a pro.
Author:  ez4u [ Mon Jul 08, 2013 4:12 pm ]
Post subject:  Re: Professional advice?
Subotai wrote:
If I have always wondered about professionals reviewing a game after a big match. If I had just made silly reading mistake that caused me to lose the game I wouldn't want to review the game right after because I would be frustrated with myself. Or if you are just too tired after the game I wouldn't want to review it either. Are pro players forced to review the game afterwards? Perhaps only in major tournaments or title matches?

I have forgotten exactly who, but one of the Japanese pros that I have met told me the reason so many obviously lost games are played out to the bitter end in the NHK Cup and the Ryusei Sen is to avoid/minimize the need to analyze their mistakes on television! :blackeye:
Author:  Unusedname [ Tue Jul 09, 2013 12:04 am ]
Post subject:  Re: Professional advice?
I like the idea. But I sort of disagree with it. (Although that sounds dumb)

It assumes that your motives are parallel with the pro's motives and if they aren't they should be.
The pro might be playing for thickness when there is a move for territory that is just as good. same with simplicity/complexity

Obviously I would prefer the pro over my shoulder knowing my idea but showing me a better move along the same vein, but this is a great cheaper alternative.

Although this small problem can be fixed by finding pros who have similar playstyles (To you or to each other) That way you aren't trying to mimic 100 different pros and their ideas and end up with conflicting advice.
Author:  Boidhre [ Tue Jul 09, 2013 12:59 am ]
Post subject:  Re: Professional advice?
Unusedname wrote:
I like the idea. But I sort of disagree with it. (Although that sounds dumb)

It assumes that your motives are parallel with the pro's motives and if they aren't they should be.
The pro might be playing for thickness when there is a move for territory that is just as good. same with simplicity/complexity

Obviously I would prefer the pro over my shoulder knowing my idea but showing me a better move along the same vein, but this is a great cheaper alternative.

Although this small problem can be fixed by finding pros who have similar playstyles (To you or to each other) That way you aren't trying to mimic 100 different pros and their ideas and end up with conflicting advice.


One thing I notice at my level (weak sdk) is that I'll guess a move and the move itself is fine but it's timing is wrong by about 20 moves, actually the upper left is bigger right now or whatever. This is useful, it's teaching me that despite seeing a good move somewhere it might not be the right move right now. I'm not sure how much of an effect it'll have on my play but it's probably good that it's being "pointed out to me" over and over.
Author:  John Fairbairn [ Tue Jul 09, 2013 1:02 am ]
Post subject:  Re: Professional advice?
Quote:
I like the idea. But I sort of disagree with it. (Although that sounds dumb)


I wrote the daddy of them all, the original GoScorer best part of two decades ago, so presumably I liked the idea. But after the first flush I have never used it again in all that time. So I suppose I too "sort of disagreed with it", and must equally sound stupid. But as I think my objection to it was that it involved work, avoiding the dreaded four-letter word may not be as stupid as it sounds.

In similar vein, T Mark often told me that GoScorer was one of the most popular items on the GoGoD CD, and I saw that for myself once when we ran a competition for best score at one of the London Opens. But I tend to assume that this too was a case of first flush then zzzzzz.

Quote:
It assumes that your motives are parallel with the pro's motives and if they aren't they should be.
The pro might be playing for thickness when there is a move for territory that is just as good. same with simplicity/complexity

Obviously I would prefer the pro over my shoulder knowing my idea but showing me a better move along the same vein, but this is a great cheaper alternative.

Although this small problem can be fixed by finding pros who have similar playstyles (To you or to each other) That way you aren't trying to mimic 100 different pros and their ideas and end up with conflicting advice.


I think that this encapsulates some of the real problems (other than the need to work). When I wrote GoSCorer I was inspired by a series called How Good Is Your Chess? from a magazine of my youth. That series had two advantages. One was that a pro or strong chess player would give comments on the moves and some guidance about the way play should develop. The other was that second-best and third-best moves were given a mark, too. My hints idea was one way to compensate for lack of these features. As to playing styles, I agree with that point to the extent that I included on the CD a file that listed thumbnail sketches of the styles of as many pros as I could find (quite a lot). This was meant to be used in tandem with GoScorer. I gave permission for that file to be placed on Sensei's Library and I assume it is still there. I have not updated the CD version much. I can't speak for the SL version's updates, but there are more than enough pros there so that you can choose one whose style you will know and/or can follow.

From memory, at the London Open, even the best player (a 7-dan) scored only 60%, and that included the trite endgame moves, so it is not realistic to expect high scores. The exercise only really has value if you try to work out why your moves were different, with (ugh) work as the operative word.
Author:  amnal [ Tue Jul 09, 2013 5:07 am ]
Post subject:  Re: Professional advice?
Since there are no android apps discussed above, I'll put forward my own program noGo. It's currently in beta (lifein19x19 thread here), but it has a simple guess mode not dissimilar to kombilo's including a percentage correct score.

As for the more general discussion, like others I like using guess modes, but I've found it mostly fun as a vaguely-work-like exercise on a train or whatever rather than as a deep study tool - that is, I spend time thinking about the moves, but not significantly beyond working out what they are and moving on. I think this is valuable, but perhaps optimally should be approached more like a full game review.
Author:  Clossius [ Tue Jul 09, 2013 5:00 pm ]
Post subject:  Re: Professional advice?
This is a good idea, but I would have to agree that it is not %100 accurate due to reasons other have posted. BUT! I think that if you applied a database to this as was previously suggested it would make for an interesting program.

Here's the catch though, one Go game does not look like another! So it would not help your middle game or end game as effectively as one might like it to.

But what about the opening? Wouldn't this make for a great opening training program? With this program we can learn different openings that professionals play so we don't have to waste precious clock time.

Downside, you might not understand why all the moves were played the way they were played. Upside, it's free! Don't complain about a training program without comments when you aren't paying for anything!

It is a great idea though.
Page 1 of 4 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/