Re: [Bug-gnubg] Tutor Mode/Hint question

[Joern Thyssen]

On Thu, Jan 09, 2003 at 07:04:00PM -0500, Hanley Hayes wrote
> Since there has been a lot discussion about 'Hint' lately, I wanted to see 
> if I could get my 'issue' resolved.
> 
> 1)Tutor mode on
> 2)I roll and move my checkers and click the dice to end my move.
> 3) gnubd pops up dialog "You may be about to play a bad move".
> 4) I select "Hint" from the dialog to see just how bad my move is.
> 5) Hint dialog pops up.  I see I'm the xth best move, but I don't really 
> care for the moves that preceed this in Hint dialog.  I really just wanted 
> to see how God-awful my choice was (remember, I'm trying to learn here, not 
> necessarily use the hint to cheat!)
> 6) I select "Ok" and both the Hint dialog AND the "You may be about to play 
> a bad move dialog" are removed.
> 7) I still want to make my same move.  To do so, I must once again, click 
> on my dice, wait for the "You may be about to play a bad move" dialog and, 
> this time select "Play Anyway".
> 
> What I would like to see is in step 6) above, after selecting "OK" I return 
> to the "You may be about to play a bad move" dialog, whereby I could then 
> select "Play Anyway".  This would save me from a) having to click on my 
> dice a second time, and 

The decision whether to show or destroy the tutor dialog depends on
whether you want to "cheat" or not after pressing "hint"!

Any comments on this from the mailing list readers?

> b) waiting from gnubg to reevaluated my move a 
> second time.

It's a known problem that tutor mode re-evaluates the moves a second
time. It's a general problem with the analysis code, so the same problem
exists if you re-analyse a match.

It's on my TODO list to fix this, but unfortunately it's not entirely
trivial, because we only save the highest evaluation for each move, but
the move filters requires us to evaluate all moves on 0-ply, then some
of those on 1-ply, the some of those on 2-ply etc etc. For example, a
sample chequerplay decision has 4 moves evaluated on 2-ply, another 6
moves on 1-ply, and the remaining 37 moves on 0-ply. Assume I want to
re-analyse this move (possibly with another move filter). How should I
do this? The biggest problem is that move filters involves comparing
equities, but you cannot compare 2-ply equities with 1-ply equities as
this is like comparing oranges and apples :-)

I've a number of strategies for this. Two main points are:

(a) avoid re-doing any, possibly expensive, re-evaluations
(b) consistent results, i.e., no difference between re-analysis of a
    move (possibly with a new move filter) and first analysis with a given
    move filter.

Unfortunately (a) is orthogonal to (b), so we have to decide which
approach to take. As a exact-science scientist I prefer that (b) is
fulfilled, however, as an impatient user I prefer that (a) is
fulfilled.

Jørn