I would never call that extra understanding just “slightly” less random. For example going from “random” to “understands what a pass alive group is” is HUGE to me. Understanding the very basic capture rule is already a huge step from random. And that does not even require any notion of eyes at all, just “count liberties, play adjacent to enemy group having less liberties in total” for example.
It is relatively simple to code to forbid suiciding a pass alive group (if you do allow it to suicide when in seki, at least, as detecting seki is harder). Just add the rule that if you have any pass-alive group (which can be checked mechanically with Benson’s algorithm) , then a move that turns a pass-alive stone into a no longer pass alive stone is forbidden. Then just play any (other) random move, until only forbidden moves remain, then pass. This bot would fill it’s own territory, but if it has a group with two very well defined eyes, it will not fill them. It will however fill it’s own eyes and kill itself many times, when the group was alive but not yet pass-alive. That is yet another much more powerful level of understanding, as it is not just a “simple fast check” as benson algorithm, but rather requires tsumego-solving (even if only very simple tsumego like the very basic elementary shapes a beginner learns on day 1 and 2 of experimenting).
It would be an interesting experiment to see what the win% between random and “random but does not suicide pass-alive groups” is (is it 90%? 99%? 99,9999%?). It will completely crush a random bot, just look at the previous game where both players have pass-alive groups near the end, but white for example suicides a large group by self-filling.
it doesn’t matter how we call it, its not “Language Learners’ Library” topic, name of such bot is not what I’m talking about.
there is just nothing to research in 100.0% random bot, so its better to research something different, that is “near” original idea.
I think a truly random bot appeals to one’s aesthetic sense as a nominal anchor for a rating system. It gives a sense of objectivity to what is otherwise a rather arbitrary choice in a rating system.
The single worst “bot” is of course the bot that just resigns on move 1 or 2 depending on when it’s first move is. It’s basically incomparable to any real player though, so it may as well be called kyu. Similarly one could have a bot that always resigns on move N and they would all be slightly better than each other.
Then one has to think about what the worst bot that actually plays the game is? They shouldn’t resign “early” or maybe at all?
What if there isn’t a worst “random” bot though. A bot playing randomly will have to choose moves according to a distribution, but I wonder could one pick a distribution for example which was bad enough to on average lose more than win, to a bot playing uniformly randomly.
Or pick tiers of distributions which tend to beat each other on average more often. I suppose in the silliest case one could have bots which only play in subsections of the board and then pass when there’s not more legal moves on those points (assume suicide is forbidden and use something like tromp-Taylor rules to score).
I think this is it though, bots playing moves randomly until game is over and bots “really playing the game” are probably very different and might just be incomparable in a real rating system.
If there is a perfect “win”, is there a perfect “lose” play? Will it be the other end of the spectrum? There should be just limited amount of possible games, so it must be finite on both ends.
Not a known one, but connect 4 is “solved” and it is theoretically possible that chess and go could one day likewise be “solved” given enough computing power.
kyu. Similarly one could have a bot that always resigns on move N and they would all be slightly better than each other.