Single Comment

Hi Kevin,

I am not that sure that your perceived failure of an nextchess idea is totally justified.

Your argument about the grandfather-grandson game is definitely correct. More over as you have said before (in different words I cannot recall exactly), it is difficult to make a game as perfectly done as the "Mad Queen" variant ended up being :)! This is why I had argued for (may I shall now name it) the many nextchess hypothesis. Together they may prove to become even more popular then it's predecessors, but separately indeed one game will never be as dominant. That makes sense mostly because of the different paths next chess can take. Let's think about 10x10 games which seem the most convenient. I'd vote for Grand Chess, Omega Chess, Eurasian Chess and Shako although I have some quarrels with the first 2 (and technically omega chess in not exactly an 10x10 game). Each has it's beauty although I won't go into details here. Further even, Gross Chess offers a melding of them. A very well made game. Another path is Chess with different armies. Betza's one is definitely the most popular one but certainly Spartan chess deserves it fair share of respect. For multi move or at least interesting 8x8 ideas (again technically not 8x8), I really want to mention 8 stones chess. Musketeer chess has it's merits also, although I'm not that happy with the execution. Maybe just an 9x8 board would be an improvement as to me the game seems way to crowded. The pieces are quite creatively done. Although no griffin?!.. Come on :(! There are also drop games in the shogi family. Pocked shogi copper seems quite promising, and here maybe is fair to mention chu shogi with it's special Lyon piece (and oh boy,oh boy that tenjiku stuff). Xianqi offer special  things with it's confined king and weaker armies and the more rarely used hopper pieces. I think this fills everybody's hands with next chess opportunities. And in closing I'd like to add my (I'd like to say humble but I'm actually quite proud of them) apothecary games. I have designed them (in the conditions where Grand, Gross, etc. already exist) to add some cream to the cake by adding weird pieces like the joker, special promotion moves, stronger knights, bruhaha squares with different opening powers, and even alternative endgame conditions. I'd like here to express my concern than nobody noticed my proposed modification to the 2 apothecaries :(! (check the last 3 comments- the last one is actually me turning a bit too hippie ) : https://www.chessvariants.com/index/listcomments.php?itemid=171817414be164e8&order=DESC .

And now let's get to serious business. But first le me explain something in order to go further. Bear with me please.

First I will refrain from using words like chaotic, random and deterministic as they do hold specific meaning for the more mathematically aware, the pop culture tends to toy with their normal usage. They way I see things at least, knowledge acquiring has two extremes. The first is the tumultuous trial and error with learning from big data, creativity and maybe even a touch of madness found most in  peak science (the last one and a half decade even more) or in the natural evolution of life. The second is the steady, safe, consistent, orderly and reductionist way of the engineers. Meaning principles are fine, but how do we automate things to get things going. And I insist on automate. Chess is very much towards the second type. I'm sure many would disagree but remember what type of game chess actually is mathematically speaking.

So why this sudden philosophy burst?

Mostly because I do not think we should look for nextchess in the past  tradition but in the holdings of the future. As I have insisted earlier on the word automate. I want to argue for the fact that chess was always meant for computers, just that at the time of it's and it's predecessor birth there were no such things (in a practical 20th century way at least). Still of course there were people who found pleasure in such findings starting roughly in early middle ages (although the Western historical timeline does not have fully explanatory power I'll stick to this approximation). But then chess and it's relatives became increasingly mechanistic in a accelerated manner in the 20th century. So, the point here is that chess is a celebration of thoroughness and that is the context where we should start our search. But, what is chess in a general way? Hard to say exactly but I'll venture a description with a flavor of definition. Chess has the attributes of being a game of usually perfect information, usually without random elements (and for know I will insist on these usual cases), where pieces with different attributes interact on the discrete and finite board (again usually but the exceptions to these are rare). A more narrower definition could impose the regular winning condition, among other things. This description leaves enough room for what if scenarios so that we will put our lab coats, make "fun" variants, and design AI's that play them well.That to see what happens and in the process bring a new level over the type 1 way of acquiring knowledge. These are the nextchess(s). And dear colleagues the show must go on!

A small addition regarding the AI vs human challenges of the last few decades comes next. Looking in hindsight, it seems that the initial ones were designed to see how good chess engines of the time were. There were no better challenges for AIs than the best humans. Now AIs have each other for that (check TCEC). In time human vs AI became just a exercise of blunder checking for the human candidate as seen in the recent MVL-Komodo challenge I've mentioned a few days ago (https://www.youtube.com/watch?v=exgVvSauvhQ). So today is just engines. In time, when people will become androids, or who knows what, we will probably have individuals (for lack of a better term I'll use the popular transhumans) that will choose such a "career". But for know we should accept that and invent chess variants accordingly.

Lastly I'd tackle the popularity issue. People worry a bit about machine on machine challenge not being popular, but TCEC does decently and that is not even a game designed with the new ways in mind. And the next human generation will obviously have a totally different outlook on life (some of your are probably parents and know better). So, never say never.

Thanks for reading so far, I appreciate it.

And I closing please if you'd like to continue the discussion please touch on the following issues:

1. Is the many nextchess hypothesis reasonable?

2. Should we try to promote successful (be it at our small scale) variants?

3. Is the chess is for Automation/analyses hypotheses for what chess means and is best for pinpointing viable?

4. Should we invest time in AI contests?

5. (I have not actually tackled that yet) How should we encourage spectators for chess variants in TCEC like events (understanding that the first step is that spectators should have a good knowledge not just on the rules but also on the strategy and tactics of the game).

- please find the time to take a look on my latest apothecary changes :)!