Re: [Bug-gnubg] Tutor mode

[Jim Segrave]

On Mon 12 Aug 2002 (11:35 +0100), Ian Shaw wrote:
> Windows build 020804
> Tutor mode = Analysis settings =  Expert
> 
>     GNU Backgammon  Position ID: 29wBAByz3RkCAA
>                     Match ID   : QQmuAAAAAAAA
>     +24-23-22-21-20-19------18-17-16-15-14-13-+  O: GnuBg (Cube: 2)
>     | O  O  O     O  O |   |                  |  0 points
>     | O  O  O     O  O |   |                  |  
>     |             O  O |   |                  |  
>     |                  |   |                  |  
>     |                  |   |                  |  
>     |                  |BAR|                  |v 5 point match
>     |                  |   |                  |  
>     |                  |   |                  |  
>     |    O        X  X |   |                  |  
>     | X  O  X  X  X  X |   |    X             |  Rolled 43
>     | X  O  X  X  X  X |   |    X           X |  0 points
>     +-1--2--3--4--5--6-------7--8--9-10-11-12-+  X: ian
> 
>     1. Cubeful 0-ply    8/5 8/4                      Eq.:  +0.910
>        0.874 0.186 0.003 - 0.126 0.006 0.000
>         0-ply cubeful [expert]
>     2. Cubeful 0-ply    12/5                         Eq.:  +0.869 ( -0.041)
>        0.861 0.178 0.002 - 0.139 0.004 0.000
>         0-ply cubeful [expert]
> 
> 12/5 was marked doubtful so I asked for a hint. I did a World-Class++
> evaluation which reversed the rankings.
> 
> Double clicking on 12/5 to play the move still told me my move was doubtful
> and the analysis window only contained the 0-ply analysis.
> 
> I can see the logic behind this; I've asked it to Tutor on analysis setings
> and that is 0-ply. However, it doesn't play nicely. I can imagine it causing
> some confusion. 
> I think it would be better if the more thorough evaluation were pasted into
> the analysis, too, although I realise it isn't strictly "correct".

That might be hard to do. Tutor mode, always calls the Analyze Move
code after setting which evaluation settings to use (the analysis or the
evalutaion settings). It doesn't look to see if the move has been
analyzed already. If it did, it would still have to decide if the
existing analysis is 'deeper' than the usual settings and should be
used instead. I could see this getting very complicated. I've seen
lots of these reversals when using 0 or 1 ply evaluation and
world-class[++] analysis. Maybe I'm preverse, but I sort of like the
fact that it happens, it helps remind you that the consequences of a
move may not fully show up for a while.
 
> _________________
> 
> It would be nice if your move was highlighted in red in the hint
> window.

That's a good idea - it should always happen if you ask for a hint
after you've made but not commited your move.

> _________________
> 
> At present I've found the best way to use Tutor mode is to use the analysis
> settings on Expert level. This gives a very fast response on my PII 266 MHz
> PC. I have the evaluation settings on World Class++. This means that from
> the Hint window I can easily select a number of moves and click Eval to get
> a deeper evaluation. (I'm finding it interesting how often GnuBg changes its
> mind BTW.)
> 
> I'm wondering about whether it is worth having snmall buttons on the hint
> window specifically to do deeper evalualtions E.g. [0] [1] [2] [3], or [<]
> and [>].

Maybe only from the pre-defined settings - so if you are using one,
then you can do a [<] [>] to another preset. Greyed out if you have a
user-defined setting.

> _________________
> 
> 
> If the tutor dialog pops up and you press Rethink the board gets redrawn in
> the original position.

Which seems right - rethink means you want to re-consider the move
from scratch.

> If the tutor dialog pops up and you press Hint followed by OK, the board
> shows the position after the move. You have to CRTL+Z to get back to the
> start position.
> I think this inconsistency might be confusing. 

I'm not sure I agree there. Rethink, for chequer play in particular,
is for people who are trying to find the best move without help. Hint
is for those cases where you just can't see a better move. I think
they address different types of learning.


> _________________
> 
> The command set tutor eval on|off is counter-intuitive.
> set tutor eval on uses the evaluation settings
> set tutor eval off uses the analysis settings
> 
> It should work the other way round.

Agree - I'll probably change this the next time I do anything, unless
someone else does it first. It's a sub-optimal choice that happened by
accident when I first took advantage of Joern's changes to allow
selecting the evaluation criteria.

-- 
Jim Segrave           address@hidden