I've tried looking at the faq under analysis and while I haven't spent
much time hacking through its jargon and attempting to recreate its
instructions in gnubg/windows, I believe that it is difficult to
understand. Maybe something with playing the databases would carry me
closer to my destination but if so I can't seem to figure out how to
use them. Oh for a tutorial for the simple-minded!
gracias in advance --d.u.g.
You aren't the first person to find this less than obvious. No, you
shouldn't continue to use the hint function. You want to play matches (or
sessions to simulate money play) and then analyze your moves. Taking
hints during the game isn't the best way to improve your skills.
After you finish a game or match, have GNUBG analyze it (this option is
on a drop down menu).
Then you need to click on the Windows drop down menu and open up both the
annotation and the game record windows. (I think that's what they're
called; I don't have GNUBG on this workstation.) This will give a list of
every move made in the game/match. Click on a specific move and
voila, you have a ranked list of the best moves in that situation.
Your move will be colored red, and the position will be shown on the
board.
These directions are perhaps only moderately more illuminating than the
faq but once you open up the analysis windows you'll figure it out.
Albert Silver has an article, a pretty good one. Has to be on
www.gammonline.com now, will be soon 'freely' available.
Louis Nardy Pillards
"Taking hints during the game isn't the best way to improve your skills." ~Ryan
Long
The best way to improve your skills is going to be a completely subjective
affair. It may very well be that taking hints during the game isn't the best
way for you, but I would suggest trying several different ways of learning
before making that assumption. One way that I didn't see mentioned in any post
is the use of the tutor feature of GNU. This is a great tool that can be highly
beneficial. You set it to the error level you want GNU to detect and it will
flag you when you are about to make a move that exceeds that set level.
Then, it doesn't just give you a hint, it lets you try again. Often I've found
that my second choice play is perfectly acceptable.
But all that is not to say that analyzing your matches after the match is not
good. In fact I completely agree that it's a must if you really want to improve
your play. Rollout the positions where you think you were right and were
flagged as wrong (affectionately called the "appeal process". :) )
Personally, I find the best way to learn to be a combination of all of the
above. But that will be subjective to each person.
Good luck!
~Silverfox
"Ryan Long" <rlo...@emerald.tufts.edu> wrote in message
news:m90y9.282$Y2.1...@news.tufts.edu...
I agree with Silverfox. The advantage to using the tutor feature is that you get
instantaneous feedback, and this is highly beneficial to learning. Certainly you
would want to analyze your matches later on, but I have found that, since I started
using the tutor feature (for about a month now), my ratings when I *don't* use the
tutor feature have advanced from around 1600 to around 1800, for the default 7-point
matches.
Regards, Ron Barry.
IMHO the best way to become a better player, is to learn from your mistakes.
In other words I'd rather have GNU tell me about my mistake after I make it,
kinda like Snowie does, showing you all possible moves with the best moves
on top compared to your move...
GNUbg can do that, but unfortunately not automatically like Snowie does...
The way the hint/tutor-option works I think of it more like a way to cheat,
when you are alerted of a possible "bad move" and by that gets a second
chance to make a better move.
You won't get that IRL! Therefore I'd rather be told about my mistakes after
I make them. This way I can think about why it's wrong and hopefully
remember that when I'm in a similar situation...
/Peace
"Ron & Ann Barry" <r-ab...@telocity.com> skrev i en meddelelse
news:3DC989C5...@telocity.com...
With GNU's tutor you must first select a play and THEN it tells you whether it
thinks it's bad or not. From there you have four options: 1) Play anyway 2)
Rethink (this allows you to try a different play without seeing any help or
hint) 3) End Tutor mode (to play the rest of the time without the tutor) 4) Hint
(which is the same as the hint mode everyone's familiar with)
The options are what make me side with GNU here. You don't HAVE to see the hint
or even change your move after it warns you. While it's true there are no
second chances in IRL, this isn't supposed to be IRL. This is supposed to be
learning time. You must first make a mistake before GNU tells you you've made
one. But most important it allows you to immediately determine the nature of
the error and whether you'd rather see the outcome with the error or see the
outcome without it (by correcting the mistake). These options I've found to be
far superior to SW's telling you ex post facto.
It's all about the options. :)
~Silverfox
"Peace" <pe...@NO.SPAM.hippie.dk> wrote in message
news:3dc996cf$0$50068$edfa...@dread13.news.tele.dk...
> ..The way the hint/tutor-option works I think of it more like a way to cheat,
> when you are alerted of a possible "bad move" and by that gets a second..
sure, this IS cheating if you are using hints left and right then
running "analyze match" and proclaiming to the world that you are
consistantly playing at world class level :-)
I think that using hints is intended to be like playing with a "tutor"
as the name implies :-) who is willing to guide you through the match,
giving "hints" and alerting you to bad play as you make em. Now..
when you are ready to challenge the tutor, turn off tutor mode, play
the whole game, and analyze match to get an honest idea of your true
rating.
regards
tucsonAZ (on FIBS)
As far as training to improve goes, I think the hardest thing, whether
you use Snowie or Gnu, is taking the time to really go over your
errors. It is much more fun to just keep playing then grinding over a
completed match.
But it seems to be one of the only ways to ferret out systemic errors
and identify position/cube decisions you have trouble with
(positions/cube decisions which might not show up that often, but may
cost you the match if played poorly).
Frank Mazza
On Thu, 07 Nov 2002 02:24:31 GMT, "Silverfox" <Axe...@hotmail.com>
wrote:
One other thought on this: I review all my matches with Snowie using the
'MatchQiz' mode which shows you the dice roll on one click, then you get a
chance to mentally guess the right play, then the next click it shows the
play and the move equities. This mode also shows the cube panel at all cube
actions and 'close' cube decisions. As opposed to just jumping from error
to error, this let's you review the correct plays you made as well. I think
there is real value when you make a difficult play correctly, to get that
feedback from the bot showing you that you did it right. Obviously, this
takes a lot more time, but seeing the correct plays as well as the errors is
definitely helpful. Also, this mode keeps you in the 'flow' of the game
better as you don't skip over big chunks of the game, and you also get to
look at the opponent's decisions; often he has a lot more interesting plays
to make.
One last aspect of looking at each position this way is you get to see the
game equity change as the game evolves, and begin to get more of a feel
about for the absolute equities of various types of positions.
Gregg C.
I wish I could skip over "big chunks of the game" as I go from error to
error.
"Gregg Cattanach" <gcattana...@prodigy.net> wrote in message
news:pCyy9.6458$103....@newssvr16.news.prodigy.com...
I've never used that feature in Snowie. I'll try that.
Memorizing match equities at various match scores is all to the good,
but it doesn't help you much if you can't assess with reasonable
accuracy what the equity of various positions are.
I can see how using that match quiz mode would help with that.
Another thing I've found useful is playing against Snowie with panels
off, and after Snowie moves or I move guess the equity and then check.
Its helped my cube efficiency (at least against Snowie) and lowered my
cube error rate.
Now if only I could improve my "expected vs. effective" ratio..:::)
Frank Mazza