I’ve played a few games against HaChu 0.18 on my new laptop. The change of CPU (O/S unchanged – Fedora 20 – I threw Windows 8 away after 5 minutes of checking that the hardware was working) causes HaChu to select different opening moves. In particular, as gote, it moves the Lion out first move, and attempts to win my Go-Between. It also came up with an interesting new variation. More on that when I resume my blog posts on that opening line.
Contrary to what you might expect, giving HaChu (and yourself) more time, doesn’t significantly improve the program’s ability.
Well, it depends how you look at it. It does improve the search ahead a little. But it doesn’t improve the evaluation function (all this is to be expected).
So the effect of extra time is to eliminate some horizon effects (so it doesn’t sacrifice so many pawns to “avoid” a bigger loss). And to give you more thinking time whilst it is thinking.
H.G.Muller has adapted the bot adaptor so it now runs on Linux. Therefore (with Hidetchi’s agreement), I am running HaChu as a bot on JUPITER. I tested it last night. A configuration parameter was wrong (now fixed), so it thought it had 60 minutes every 40 moves, instead of for the whole game, and so lost on time. But the game is still instructive:
When it runs out of time after move 95, material is still equal (only the edge pawns have been exchanged). But my positional advantage is huge, by my evaluation Function. HaChu thinks otherwise – xboard was showing +1.76 for the bot – roughly a Silver General, I think. Incidentally, I had used only 5 minutes of time. Mostly I thought in its time. This becomes much harder to do when it is playing faster.
A few weeks ago, I upgraded my MacBook Air from Fedora 14 to Fedora 20. Immediately HaChu 0.17 started playing the opening differently against me. The set of moves it chooses from doesn’t appear to have altered significantly, but the order in which it plays them is significantly different.
The difference is most obvious when I take gote. Instead of opening with a Lion move, it opens with P – 5h. Then when I reply with P – 5e, it still doesn’t bring the Lion out. It either plays the other pawn, or more often it moves a Dragon Horse – either straight back, or more often into the centre (as Black, it often moves the Dragon Horse all the way to the edge in the opening – something it never used to do.) . Sometimes it doesn’t move it’s Lion at all.
All-in-all, once I got used to this style, I find it easier to win against it (with either colour).
I suppose one can attribute the difference to changes in gcc and/or glibc. Or maybe GNOME 3 has greater overhead (I’ve never noticed it – BTW I love GNOME 3 – that was the reason for upgrading).
Black will normally start the game with Ln – 6h (or Ln – 7h, but moving the Lion to the side on which the opposing King is situated is symbolic of an intent to attack). White’s most principled counter-attacking move is Ln – 7e (the principle being to get your lion close to the opposing King. Of course, that is hardly significant at this stage). Black will then often open the diagonal with P – 8h, in order to support the Lion to the high position on the opposing King’s side. If White follows suit with P – 5e, and Black plays Ln – 6g, we reach the standard High-Lion position:
The High Lion on 6g dominates the centre. White in turn would like to get a High Lion too. One way is to try to drive the Black Lion from it’s position in the centre. Another is to eventually move his own Lion to 9f, after first pushing the Go Between from 9e – 9f -9g. Meanwhile Black might be aiming to attack on the fourth or fifth files. Many other strategies are possible.
But Black has another High Lion position possible. After the first two moves, Black can play P – 5f instead of P – 8f. Then he plays Ln – 7g, to directly oppose the White Lion:
Here, rather than anticipating a mutual attack, Black can try to restrict White’s development. Personally, with my style, I tend to find this more difficult to play with as White.
If I were to try to avoid this by playing Ln – 6e as my first move, then Black can just play P – 8h and Ln – 6g as before, again with an opposing High Lion.
So as White, I usually play P – 5e as my first move (I think this is better than P – 8e, because of the position of the White Phoenix), intending to see on which file Black places his High Lion before bringing my own Lion out.
But this allows an early Lion foray for Black. He might play Ln – 8f as his second move, giving us this situation:
Now Black can take the Go Between. Is this OK for White? That will be the subject of a future post.
As far as I can remember, in our family we always played card games and draughts. But my uncle Peter played chess, and one Christmas (I think – I was about 7 years old), I was given a chess set and uncle Peter taught me to play.
Later, after reading how James Bond taught Drax a lesson on cheating with a 7-clubs redoubled hand, I went to the library to teach myself how to play Contract Bridge. In the sixth form at school, English teacher Paul Davies started up a bridge club. I got keen, and it lead to my organising a pupils versus staff match (which I’m glad to say, we won). I also started going to Thetford bridge club.
Similarly, after read the chapter “A game of Mahjong” in Agatha Christie’s classic “The murder of Roger Ackroyd”, I bought a set from a shop in Thetford, and learned to play.
I never played very much chess, and was never a member of a club, until one year at university (I was, of course, playing bridge throughout my university career), I was in the same student house as a fairly strong chess player. This lead to joining the university chess club (as well as Hull chess club), and I ended up as team captain, whilst simultaneously bridge vice-captain (although I was never the strongest player on either team – far from it in the case of chess).
After leaving university, I never played either game much again. Some years later (it must have been 1989, I think), I was introduced to Mark Ivey in a pub in Preston. For some reason chess was mentioned, and we went round the corner to his flat to play a game. He wasn’t very good at all, and I beat him easily. He suggested he teach me the rules of Go, and I readily agreed. He gave me a 9-stone handicap, and beat me easily, but by the third game I had got down to 7 stones (I think I won the second game with 8 stones). “You’re getting the hang of this”, he said. “You wait”, I said drunkenly, “in a couple of months I’ll be beating you without a handicap”. “I bet you a gallon of my home-brew that you won’t” he replied, and handed me “In the beginning” to help me learn.
The next morning, I soberly reflected that I had better get reading if I was to win that rash bet. At the end of 2 months, I was indeed able to win (with the black stones, it took another week, I think, to win as white). Meanwhile Mark had got in touch with the British Go Society, we formed an officially affiliated Preston Go club, started going to tournaments, and I was hooked.
My friend in Reading, Tony Atkins, always stronger at Go than I, taught me to play Shogi whilst I was staying with him to attend a meeting of the BGS council. Unlike at Go, I was soon beating him regularly. This was either in 1995 or 1996.
In 1996 I finally got promoted to shodan. That August, I went to Abano Terme, in Italy, for the European go congress, flying to Venice (my first ever journey in an aeroplane, at the age of 36). Whilst there, Tony introduced me to Arend van Oosten, a 4-dan Shogi player, and several times European champion. He told me that the European championships were going to be held a few weeks later in Belgium. So I went, entering as 3-kyu (a guess – I ended up losing to 2 kyus and beating 4 kyus, so it was probably a good guess. Later, after joining the British Shogi federation, I was promoted to 1-kyu. ). There I met Thomas Majewski, who told me that Chu and Tori were the best variants. When I got home, I ordered sets of each, plus the Middle Shogi Manual, from George Hodges. And then I taught myself to play.