Most of these assume perfect play everywhere. For questions about komi, B should play to maximize win margin - not to ensure a win like a monte carlo bot would. (although the winner in both cases should be the same, the maximized win margin space is a lot more interesting than the much broader space of all possible B wins). And in perfect play, you might as well win by the maximum.
A perfect game is a game in which there is never a point either side could play a different move and get a better result. Generally, game length or complexity is not included - it's only about the final differential. So there could be multiple perfect games from an initial condition.
Also, 4-4 is not considered distinct from a first move at 16-4.
I'd guess somewhere between 5-9 pts for 19x19. It'd also be really interesting to know it for larger board sizes.
And what would the longest / shortest one be?
Also, which perfect game has the most captured stones? That'd be one I'd like to watch.
Which one has the biggest gap in captured stone differential?
Is there another measure of complexity, too, which would have an outlier - the game with the longest ko fight, or most complex semeai?
Or, find the game with the most tenukis - probably an insanely difficult big simultaneous fight - or a really hard long endgame. How about the game with the least tenukis? What would that even be? What about the game with the most captures? If necessary for these you can expand to games which are not perfect, but are within just a few points. Within that much larger space, there are probably some games which are 99% perfect, but also have amazing, extreme characteristics.
In general, selecting among multiple perfect games would be incredibly interesting.
Another question: which instant in a game has the most perfect continuations? (divided by move number - so for example, there might be multiple first move perfect moves - but how about a perfect line which has a choice at move 10 for the next move? It'd be amazing if there were two answers there which both led to the same ideal result.) At move 200 there could be many - multiple ways to play the same endgame. But early on, these multiple points would really be something to see!
No full board repeats, or one which allows draws? Does go really even make sense if you allow ko?
Opposite side hoshis seems pretty bad... And, how would you play larger handicaps?
The default all hoshi play is probably not right - there's a huge difference between 5 and 6 stone handicaps, for example. With free placement, anything is possible.
What is the win margin for B with various 3-stone openings - chinese, san ren sei, shimari + hoshi, shimari + extension.
Somewhere between 9 and 16, probably.
Mostly just filling up the corners - but what about worst placement of 20 stones?
What would the patterns look like on the side? A b2-bomber in the corner? or a line with gaps? Would B ever play a dumpling? What is the worst 5 (N)-stone clump of connected stones, with at least one of them covering tengen?
That is, which would be the best move in a hex-style opening - one which B would not mind playing and then switching sides 50% of the time. Which moves are right on the line between a win and a loss (without komi)?
What is the value of tengen for B? What about 5-4? 5-5, etc. How much worse are these moves than the best one? I'd like to see a map of B point differential, for every opening move - darker green for larger wins, red for losses. Where along the side / corner is the line where the game flips from win / loss for B? in the corner, 3-3 is still probably a win - but is 2-2 on the loss side? how about 3-2?
Obviously I'd like to know this, although it's included in the examples of perfect games.
i.e. other ko rules, different initial setup, board shapes / sizes etc.
if it is, it will most likely have the same ko rule - but board size could easily be 21x21 (or smaller, too)
And what's their play style in general?
What is the history of their go - how did the discover it? from astrology? What order did they discover openings in?
Do they ever use go to resolve political disputes?
What is the history of their Komi?
What are their timing systems like?
For that matter, how did humans discover go?
If we're asking god these questions, we might as well ask for a full history of go's development - and also, bring back famous go-related historical figures, and show them modern games and get their feelings and ideas about them!
Which two amateur games are most similar (through random chance)?
That is, when the human doesn't know who they are playing, and the perfect player can't use information about the person from outside the game. The perfect player would play to maximize win possibility (like a monte carlo bot), not like a human, maximizing win differential.
Obviously, a lot more stones - cause god would know exactly which josekis you knew or where your misconceptions about go were and would constantly play into them.
In both previous questions, God would still be playing online with an innocuous nickname. If you need to play his human guise, there's even more he could do to you - move his hand in exactly the right way to remind you of something a terrible loss, or play at exactly the time you are thinking of the worst possible response line.
And if he is truly acting as God, an even worse version would be if he used his knowledge of your knowledge to lead you down your weakest lines, or get you to play moves you haven't used in a long time. i.e. within every pro's head there are thousands of memorized josekis & variations - God would be able to choose the least accurate ones to lead you towards.
That is, some perfect games may be a lot harder for humans than others. Is there one style that we naturally are better at in a mathematical sense than another? That is, hoshi openings and 3-4 openings may be mathematically equivalent within a few points, but human's win margin vs a perfect player after using that opening may be wildly different.
Which opening to humans in general play closest to perfect?
and which one, while being a valid line, do we play terribly?
With all that center space, there could be some tough fighting.
What is the hardest life and death problem?
A really sad result of talking to God about go would be that there is a way to just play mirror go and win. If mirror go doesn't actually work, how does God defeat it? Does just excluding mirroring (at any stage) fix the issue?