Mr_Adams wrote:I may have missed the response, but I mentioned earlier that the game stats now tell more information than the log in fog/nuc games in the territory colum. you can't tell who lost terrs by a cash in nuc games by the log, but it is taken into account in the statistics.
Came to the thread to say this. Is anything being done about it?
I was also going to say the same thing. The information gap only lasts until the player's next turn, since then the log tells you the territory count, but still, it might be valuable.
On the other hand:
When statistics were introduced, they were designed to only contain information that could be deduced from the log, in order not to change any of the existing game types. But nuclear spoils didn't exist then. So the requirement that the statistics don't contain any extra information isn't as important in this case -- the statistics don't "change" how the game with nuclear spoils works; there was a choice what the statistics would contain when the nuclear spoils were introduced, and it wasn't constrained by backward compatibility. Given what the log says when you nuke, there were basically three options for the statistics:
a) always display the territory count (the current behaviour)
b) never display the territory count in fog+nuke games
c) display the territory count in fog+nuke games whenever it can be easily deduced from the log
Both options b) and c) have disadvantages -- b) is a bit of an overkill, and c) would involve more programming and would lead to the territory count column being an ever-changing mixture of numbers and question marks. Another solution would have been to put more information in the log. The straightforward way of doing this, saying for each card whose region it nuked, like for a bombardment, would actually have disclosed even more information; to disclose only the information that's currently being disclosed in the statistics, the log would have to say something like ", annihilating A's troops in two regions and B's troops in one region", which again would involve more programming and would seem a bit artificial -- tayloring the log to correspond to the statistics rather than putting what one would naturally put there.
So all options had some disadvantage, and it might be argued that violating the principle that the statistics shouldn't disclose more than the log (the main reason for which had been not to change existing game types) was the least of them.
I'd be interested to know what others think about this. If you want something "done about it", please specify what exactly should be done. The only option that wouldn't change the game now (in the same sense in which introduction of the statistics didn't change the game at the time) would be the one putting something like ", annihilating A's troops in two regions and B's troops in one region" into the log -- every other option (in particular every option that would change the statistics so they don't reveal more than the log) would now, somewhat ironically, violate the principle underlying the principle "statistics shouldn't go beyond the log" -- namely, "the game shouldn't change"