|
From: | Albert Silver |
Subject: | [Bug-gnubg] CSS and HTML export |
Date: | Sat, 2 Aug 2003 16:43:02 -0300 |
Hi, I'm attaching here the external CSS I created so that you can easily use it, and possibly incorporate it as the native CSS of GNU. Apart from the color changes it also includes justifying the text in the Annotation area. On a same note, I'm including here a sample commented game in which I changed the cube presentation just a little adding a few key boldfaces which I think improve the readability of the data. I'd also like to propose a small option in the export allowing a user to define the margin for "close cube decisions". I understand that the program itself determine close decisions as within 0.250 when calculating its statistics, and while I would like to see decisions that were with 0.050 equity in the HTML export, I don't like to see constant 0.200 decisions that neither I nor the player would really consider. If the player makes the blunder fine, but I'd rather not see a ton of useless cube decisions there. Again, the images should be in a subdirectory called \html-images. Albert
Game number 1Moshe Tissona (0 pts) vs. Jon Royset (0 pts) (Match to 25)[First Game] [Next Game] [Last Game] Match Information
Moshe Tissona's rating: 0 (Exp 0) Move number 1: Moshe Tissona to play 61 Pip counts: Moshe Tissona 167, Jon Royset 167 • Moshe Tissona moves 13/7 8/7
Annotation
Casper van der Tak:
Introduction Tissona makes the bar, Royset answers with a small split, and Tissona with a major split. Royset hits, and a blot-hitting contest ensues in which Tissona may have missed a good opportunity for an unusual hit. The hitting stops for a moment when Tissona makes an advanced anchor, but then Royset hits two of Tissona's checkers. Tissona answers with a useful 44, played in an unusual but strong, tactically motivated manner. He sets up 2s from the bar as a root number for Royset. Royset promptly rolls a two from the bar, but fails to give himself a playable two from the bar should he be hit again. Tissona may have missed a cube opportunity, but hits Royset again. Royset again rolls a two from the bar and his position collapses. Tissona cashes. Move number 2: Jon Royset to play 51 Pip counts: Moshe Tissona 160, Jon Royset 167 • Jon Royset moves 24/23 13/8
Move number 3: Moshe Tissona to play 42 Pip counts: Moshe Tissona 160, Jon Royset 161 • Moshe Tissona moves 24/20 13/11
Annotation
Casper van der Tak: 24/20 13/11 is my play
also, but the very quiet alternative of 13/7 is not bad at all. Making the 4-point
leaves too many shots and is not as good as either of these alternatives.
Move number 4: Jon Royset to play 51 Pip counts: Moshe Tissona 154, Jon Royset 161 • Jon Royset moves 13/8 6/5*
Move number 5: Moshe Tissona to play 41 Pip counts: Moshe Tissona 159, Jon Royset 155 • Moshe Tissona moves bar/20*
Move number 6: Jon Royset to play 65 Pip counts: Moshe Tissona 154, Jon Royset 175 • Jon Royset moves bar/14*
Move number 7: Moshe Tissona to play 42 Pip counts: Moshe Tissona 168, Jon Royset 164 • Moshe Tissona moves bar/21 13/11*
Move number 8: Jon Royset to play 51 Pip counts: Moshe Tissona 162, Jon Royset 175 • Jon Royset moves bar/20 6/5*
Move number 9: Moshe Tissona to play 31 Pip counts: Moshe Tissona 167, Jon Royset 169 • Moshe Tissona moves bar/22 21/20*
Annotation
Casper van der Tak: After some early sparring,
with mostly automatic decisions, comes the second real decision of the game.
Which 5-point should one fight for? Normally you'd like to play B/22 21/20*
rather than B/22 6/5*. Both plays fight for key points and are therefore positionally
attractive, but B/22 21/20* gains more pips in the race and does not leave hit
and cover numbers. This normally would more than compensate for bringing a checker
from the stack on the 6-point into play.
However, after Tissona's B/22 21/20* all numbers except 66 and 21 play OK for Royset, either making the 20-point anchor or hitting Tissona's blot on the 11-point. Hitting 6/5* has the advantage of taking away Royset's good entering 6s, and 66, 61, 62, 63 and 64 become so awful that Tissona would have a very strong cube after either of these replies. This is a tactical justification for B/22 6/5*. The normal tactical drawback of 6/5*, the hit-and-covers is leaves, is mitigated by Royset's poor distribution up front which makes it difficult for him to fully benefit from these. A strategic justification for 6/5* is that if both players make their 20-anchors a lot of the value of the made bar-point would have been negated, whereas if both make their 5-point Tissona would have a big advantage thanks to the better prime. I think Tissona made a slight error here and would have played B/22 6/5* myself. I also predict that the normal B/22 21/20* play will be best if you do either of the following: a) Move the checker on the 11-point to the 10-point or midpoint b) Give Tissona a made 11-point by moving a checker from the 6 to the 11-point c) Take away Tiossona's bar-point Move number 10: Jon Royset to play 61 Pip counts: Moshe Tissona 163, Jon Royset 189 • Jon Royset moves bar/24 20/14*
Move number 11: Moshe Tissona to play 54 Pip counts: Moshe Tissona 177, Jon Royset 182 • Moshe Tissona moves bar/20 22/18
Move number 12: Jon Royset to play 65 Pip counts: Moshe Tissona 168, Jon Royset 182 • Jon Royset moves 13/7* 6/1*
Annotation
Casper van der Tak: I like 14/9 13/7* better
than the hit on the ace which throws away the last spare on the 6-point. The
double hit has the advantage of reducing the number of return hits, but the
positional price paid is high.
Move number 13: Moshe Tissona to play 44 Pip counts: Moshe Tissona 176, Jon Royset 171 • Moshe Tissona moves bar/21(2) 6/2*(2)
Annotation
Casper van der Tak: Making the 2-point on
Royset's head shows excellent judgment. The play has several advantages:
- Reducing the number of Royset's rolls that can be used to hide or cover the blots; - Bringing some of the spare checkers on the 6-point into play - Setting up bad entering numbers. If you have the chance to put your opponent on the bar, it is a good habit to checker how his numbers corresponding to your home board points play. Here: how do his 6s and 2s play after Tissona makes the 2-point? 6s will be playable, but Royset will only have terrible entering 2s. It is an unusual play (making the 9 or 4 would be more common) but very strong. Move number 14: Jon Royset to play 52 Pip counts: Moshe Tissona 160, Jon Royset 173 • Jon Royset moves bar/20 8/6
Annotation
Casper van der Tak: I have a slight preference
for B/20 24/22. Although risky, this tries to fight for the advanced anchor
that Royset badly needs, and also leaves Royset with some reasonable entering
2s should Tissona hit again. Tissona has only 9 checkers up front, so the risk
of getting blitzed is not too great.
Tissona should double either play. Move number 15: Moshe Tissona to play 64 Pip counts: Moshe Tissona 160, Jon Royset 166 • Moshe Tissona moves 21/11* Alert: missed double ( -0.117)! [bad]
Annotation
Casper van der Tak: Tissona should double
because Royset's position is so disjointed and because Royset will continue
to have difficulties playing his numbers. Several hitting and point sequences
are very strong and could lose Tissona's market by a lot.
Hitting 21/11* is clear, sending another checker back and gaining in the race. And who knows, Royset may roll another 2. Move number 16: Jon Royset to play 21 Pip counts: Moshe Tissona 150, Jon Royset 177 • Jon Royset moves bar/24 6/4* Alert: unlucky roll! ( -0.350)
Annotation
Casper van der Tak: And another entering 2! I'd play B/22 again, but it should not matter, since after either play it is both a double and a drop. Bots: According to GNU 0.14 it is a clear Double/Pass whereas Snowie 4 says it is borderline (+0.002) take. Move number 17: Moshe Tissona doubles to 2 Pip counts: Moshe Tissona 154, Jon Royset 174 • Jon Royset rejects
Moshe Tissona wins 1 point
[First Game] [Next Game] [Last Game] Output generated Sat Aug 02 14:53:03 2003 by GNU Backgammon 0.14-devel 1.1227 030727 (HTML Export version 1.124) |
gnubg.css
Description: Text Data
[Prev in Thread] | Current Thread | [Next in Thread] |