Hrqls: Oh my, that's dedication. Sounds like the job for a program if you intend to keep this list updated.

Marfitalu:

Hrqls:

Hrqls: Remember that miracle with those 10 grand pianos that fell off a skyscraper and crashed to the ground all around me yet I didn't even get hit by a splinter? I think I'll take a stroll through the lion's den at the Zoo. I've always wanted to see them close up and stroke one.

Hrqls: In that game you wouldn't care about gammons. It's such an overwhelming loser that you'd drop it like a smelly dead cat and be grateful for your opponent's mistake.

Your opponent should have kept the prime and hit you as soon as possible in the hope of being hit as you come off the bar and, in turn, hitting your blot on 4, or otherwise picking up a companion for your trapped man. That would increase his gammon chances considerably. Even without hitting you, it is, as grenv said, too good to double.

grenv: Aye, I'll never play Anti.

Hrqls: Here's a favour:


frolind: ________4,  16,  _7,  34,  _3,  42,  17.67
Hrqls: _________33,  12,  18,  19,  _8,  17,  17.83
basplund: _______8,  15,  _5,  _6,  38,  17,  20.00
Arpa: __________7,  89,  22,  _1,  _1,  _4,  20.67
Zach: _________21,  60,  20,  32,  24,  36,  32.17
whikki: _______48,  33,  31,  49,  41,  25,  37.83
lovelysharon: __69,  52,  21,  29,  53,  20,  40.67
incognito: _____42,  19,  _9,  36,  87,  90,  47.17
TC: ___________87,  29,  48,  _4,  17,  98,  47.17

the list of the bkrs


frolind: _______2308,  1920,  1969,  1896,  2016,  2111,  2036.67
Arpa: _________2248,  1760,  1842,  2048,  2102,  2192,  2032.00
basplund: _____2248,  1929,  2012,  1958,  1815,  2106,  2011.33
Hrqls: ________2186,  1937,  1868,  1930,  1909,  2148,  1996.33
incognito: ____2177,  1917,  1963,  1890,  1727,  2077,  1958.50
Zach: ________2202,  1801,  1854,  1909,  1862,  2117,  1957.50
lovelysharon: _2158,  1817,  1846,  1911,  1783,  2140,  1942.50
whikki: ______2174,  1858,  1729,  1847,  1811,  2136,  1925.83
TC: __________2149,  1840,  1517,  1979,  1877,  2073,  1905.83

Hrqls: The workaround is to use underscores. (Dots don't work as they aren't a full character width).

Ideally you should be able to use   to get additional spaces but there's a bug in the message code which (unpredictably) ruins the layout of anything long so   is only useful when inserting a couple of additional spaces.

Hrqls: It may not be that simple programmingwise. It depends on whether the game result is stored with the game. If not then I imagine the code would have to run through each game totting up the points until it got to the match in question.

But certainly I'd also like to have it show the match score at the start of a game when I'm reviewing it. I'd also like to see the correct value of the cube if I click to any move before the last accepted double. At the moment that's frozen at the final value too. And I'd still like to see the dice roll displayed!

Hrqls: LOLOL.

What a pity. If you're talking about the match I think you are, then I already made that list when I analysed the position acouple of days ago. But it was only a temporary need so I didn't save the file.

Hrqls: Until Fencer adds it to the match table (a trivial task, eh, Fencer, eh? ) you have to open each game in turn and check the cube and the position. , and

grenv: Lol. I think that Chess player was on to something there!

Non-game reasons: I've had an opponent stuck on the bar and resign in protest against the lack of auto-pass. That's a shame in a sense 'cos I like to play and win rather than be gifted with one, but I understand and accept the feelings behind that decision.

Hrqls: "but i wont resign before i am completely sure :)"

Absolutely. I had one of those miracle wins where I won even though my opponent had borne off all but his last man, so I'm totally convinced!

I was playing someone at a different site and they resigned a position in which I was bearing off but had only two spares, one on the acepoint and one on the 5-point, with the 6-point open. He had 2 men on the bar and me leaving a blot for one, tow or more moves was pretty likely. But the guy resigned. I couldn't believe it and asked him why. He said "I don't mean any disrespect to you but I do resign when it looks hopeless. After almost two years you get a feel for the game..thanks". Somehow he's "got a feel" for resigning a 12% chance!

Hrqls: Lol. My eyes are closing. I thought the game had ended with a resignation at that point.

If I were white I would expect my opponent to gratefully drop the cube if it were offered. The hit is only going to happen 30% of the time and the chances of keeping the blot thereafter are not high enough to justify taking. You were very lucky.

Thus says me. Let's see whether GnuBg is going to agree or tell me off!

....

Okay. After White's move 34, and just before Black's 1-1, the percentages are W 91.1%, Wg 55.3%, Wbg 0.1. After the 1-1 and Black has closed his home table, they are W 93.9%, Wg 75.6, Wbg 4.4%

Dice decisions.
Double: Are you crazy???? Gammon, man, go for the sizzle!!!
Take: Are you crazy???? The frying pan's out and hot! And you want a doubled cube???

Hrqls: Lol. Aye, but when? If he'd doubled on his first move you'd have been a fool to drop! And what situation are you talking about? There are 94 "situations" in that game. ;-)

Hrqls: And the cube... What do you want to know?

Hrqls: I'm happy to analyse a position or match every now and then if it helps someone learn something.

What do you mean by a cube analysis given that this is a single game? In order to analyse it I need to know how long the match is and what each player's score is at the start of the game. Also to know at what moves any cube decisions are supposed to have happened.

grenv: That's a very good point and suggests that I made a mistake in setting up the analysis. I'll do it again....

But no. At a score of 2-away, 2-away, whether in a 2-point match or the 21-pointer that it was, the number and conclusion are the same: -0.238 (3-ply) and a very bad take.

Hmmmm......

Doh! It's in what Hrqls said in the original query, "i am not sure if i have a better chance winning the next 2 games than i would have had to win the doubled game (which i declined)". The answer is a resounding Yes to trying to win the next two games rather than that single one.

I normally use the equity figures when analysing but GnuBg can alternatively show the equivalent in Match Winning Chance and it's much better for this query.

Cube analysis, 3-ply, MWC

1. Double, pass _____ 68.75%
2. Double, take _____ 73.20%  (+4.45%)
3. No double ________ 68.33%  (-0.42%)

Proper cube action: Double, pass

This time you can see that taking the cube gives away 4.45% more than playing the next two games.

BIG BAD WOLF: "If I think I have a good chance to lose - deny the double"

Aye, that's a reasonable guideline to use at the start but you'll be at the mercy of those who know they can scare you by doubling. There are plenty of situations that look like you're going to lose when you should actually take. You'll know you're getting the hang of the cube when you confidently take what you would once have dropped.

Paradoxically a common feature in games with those who don't know how to judge the cube is for them to take a cube that should be dropped like a hot rivet. And when they get the 6-6, or whatever, and go on to win... Doh!

grenv, Hqrls I won't claim to be an expert in this area so here's what I understand myself.

W 72.8%, Wg 29.3%, Wbg 2.6%, L 27.2% Lg 6.4%, Lbg 0.5%

Wins (all), Wins by gammon, Wins by backgammon, Losses ditto
The percentages are for the person cubing.
(That's why I said "nicely judged" to you grenv because Hrqls' chaces were 27.2%)

1. Double, pass _______ +1.000
2. Double, take _______ +1.217 (+0.217)
3. No double __________ +0.964 (-0.036)

The equity figures are how many points you'd gain or lose on average. Drop the cube and you'll lose every match 1.000. Fail to double now and you'll only win .964, ie. delaying for 1 roll will cost 3.6% of the possible points (because of losing the game or the next one or from subsequent cube decisions). If the opponent takes the cube then the gain will be .217 above expectation. That's a huge increase, and a huge additional loss for the loser given that they could have dropped and only lost the single point.

The 75%/25% double and take rule works because it's the break-even point for the taker. if they play 4 nmatches and take, then losing 75% at 2 will cost 6 points but they'll win 2 back with the other 25% for a net loss of 4. That's the same as if they had dropped the cube in all four games.

That rule doesn't take into account the gammon and backgammon wins which, in this example, are considerable. Two men on the wrong side of a 3-prime with good builders. That's the extra factor that makes it such a blunder. The maths now becomes
(72.8 - 29.3 - 2.6 = 40.9) x 2 win points +
(29.3 - 2.6% = 26.7) x 4 gammon points +
2.6 x 8 backgammmon points
versus
27.2 x 2 lost points.

As far as knowing exactly how the equity figure is worked out I can't say. In a single game it's as straightforward as the maths for the 75%/25% example shown but for longer matches I believe the equity value also has the future matches factored in through the use of the match equity table. That's where I get uncertain because to it seems logical that equity for a match can only go as high as 1. Perhaps Alan can explain this aspect.

Hrqls: Heh heh. I considered a "tiny bit" too. As always, it's much easier to find justifications for an answer that has been proved. Such was the case with the description of the position. ;-)

Aye, such a nice set of builders was a big threat.

Hrqls: Looking at the position you've got two men back versus Gamek's single man which ready to escape. If it doesn't manage to escape, (eg a 2-1) then you've got to hit it and cover the blot on your 4-point. You've got no home development and only the initial builder's points while Gamek has her bar point and three sources of builder. Both your back men are blocked on 6s and 5s. And the pipcount deficit is 20 points plus the roll. Not a lot of joy in that scenario. ;-)

grenv: Nicely judged.


Cube analysis, cubeful equities:

3-ply
W 73.2%, Wg 18.5%, Wbg 0.6%, L 26.8%, Lg 5.0%, Lbg 0.1%

1. Double, pass _______ +1.000
2. Double, take _______ +1.238   (+0.238)
3. No double __________ +0.978   (-0.022)
Proper cube action: Double, pass


Rollout, 1296 games
W 72.8%, Wg 29.3%, Wbg 2.6%, L 27.2% Lg 6.4%, Lbg 0.5%

1. Double, pass _______ +1.000
2. Double, take _______ +1.217   (+0.217)
3. No double __________ +0.964   (-0.036)
Proper cube action: Double, pass

The figure in bold is how much would have been given away by taking the cube. A value of .200+ makes it a major blunder.

BBW Did you report robyou/roguetoo, or whatever his name is, and BaDBoY to Fencer by any chance?

grenv: You did, you did! I think it's because I appeased the Dice Gods with another trio of my youngest.

grenv: I've a gammon loss that I'm trying to fend off. I desperately need to know how to do that thing with the chicken.

Czuch: Did you name names or address a class of people?

grenv: Aye, ignorance is something that we should scorn and look down upon. Let the fools and their money be parted. We who are clever and smart an on the ball can profit by them and good luck to any of us who does so in a systematic way. Don't suffer the fools gladly - gladly make the fools suffer!

Have I expressed your attitude correctly there?

Pedro: Good on you. Sock it to him!

Hrqls: Sure. The 5-point (5 and 20) and barpoint (7 and 18) are the best advanced anchors to have.

Groucho: It's a tough question. The purpose of the clock is to encourage fast play. A game with a hour total and no bonus is going to be over in two hours maximum. But I've had games that have been quicker yet had 1 day clocks. That was against opponents who had my speed style, were online at the same time and didn't have hundreds of games to play.

Strangely enough the Very Fast Fischer's Stairs that I play has been slower moving than the ordinary Stairs!

Hrqls: Yep. Most players don't make their last move, go to bed, get up and log in again... so they get sucker-punched.

Groucho: Sure, but if you're using it in as deliberate a manner as he seems to be doing then he will be checking in frequently to ensure that he always gets to make his move. He's relying on his opponents not being as on the bal about it as he is. In the relatively few matches that that's been the case, he's had an honest game on his hands.

It's the overwhelming number of matches that he's won by timeout that makes it looklike cheating. Any normal person would feel wrong winning so much that way and would stop using such a clock because it's so obviously unfair to others. Such wins are hollow victories to real people. You don't even need to check individual games as I did. Simply look at the finished games lists and see how many games are done with in a sprinkling of moves.
Battleboats
Checkers
Anti-Checkers

Hrqls: All his Battleboats, Dark Battleboats, Linetris and Checkers games have been won on an 11 or 13 hours-per-move clock. Many of his recent (since October) Anti-checkers have been won that way too. A huge number of his Line4 wins have been on time. This goes back to last year as well. He used the 1-day clock until Fencer introduced the Fischer's clocks. A complete bas-, er, unfair player, if you ask me, though it's not against the rules to use those clocks. Completely meaningless stats, though. He makes a mockery of everyone including Fencer.

BeeBeeDubbayoo: Where does it say that? Could it mean the 2 each that is also the limit for deletion?

grenv: Lol. Silly me. I went straight to the games and didn't even notice tha last moves column in the games list.

Privacy allows people to hide their tactics. Hardly a major priority for most people but perhaps relevant in Battleboats if you don't want your search plans revealed. Maybe that's the case in other games but in most games it makes nary a hoot.

I find it annoying that my opponent can make a game private and I won't know until I try an access it without having logged on first (which I do sometimes).

grenv: Highest rated? Jah. Best player? Pah!

Fencer: Could you make it so that provisional games cannot be played in private, please? There's no way for us to check for cheats otherwise.

grenv: I'm off to bed so just a quickie..

46656 / 36 = 1296 and 1296 / 36 = 36. Such a number ensures that each dice roll is represented fairly for the first three moves.

I've done rollouts to 1296 and to 12960. Similar principle; 10 goes each for each pair of opening rolls in the latter case.

The 1296s were just so many wasted CPU cycles. Even at 12960, though, the expected error value exceeded the difference between the top moves in many cases. Getting the error values small enough to reliably decide between two moves would require rollouts of 500,000 and more.

I agree with 3-1, and 6-1's up there too. Whether the results can reliably differentiate between the least popular rolls I'm not sure. I'll have a look at what I found in a few hours...

grenv: Which rollouts are you talking about here? How many iterations and what's the error margin, etc, etc?

volant: Lolol. Nice one!

grenv: I think 2 rounds may be optimistic, too. 18 players so far and several weeks to recruit more. Only the board readers know about it so far as it's not on the tourneys front yet. And if volant fancies doing a broadcast to a bunch of players from the Backgammon rankings .... Also, we should go by the average of the slowest rather than everyone, so perhaps 4.88 days per move.

We could have us a 10-year tournament!

Frolind's 21pt tourney from October is racing ahead.

volant: Very nice of you to create this tourney.

Blimey, it doesn't start for about another 50 days! Any bets on how long it's going to take?

alanback: "BG is never played in 2-game matches for the very reason that the predicted outcome in such a case is 1 win apiece

This puzzles me. Between equal players the win rate will the same for each over the long term. For unequal players the stronger is going to win more matches. "Possible", yes, "predicted", not necessarily.

grenv: "Are you sure you want to leave a blot there?"

Czuch: Yep. I quite agree. Only one hour ago an opponent of mine sent me a message informing me that he had just accepted my double by mistake! Colour-coded buttons and a nice checkbox. Dead easy!

Czuch, does WebTV handle javascript, do you know?

txaggie: "He could easily spend all his time doing that, and not get to the other features and new game ideas that people are asking for."

I think that would be a seriously difficult challenge. lolol. But you're right, do this one thing and it's the start of a slippery slope that ....

BIG BAD WOLF: Yep, I'd go for the checkmark. I don't like a whole new dialogue/page for confirmations.

grenv: LOL. No, no, that's not what I meant!

grenv: WebTV .... Nothing .... WebTV .... Nothing .... ?

Aye, you're right. Best to do without.

Hrqls: MS Rules! Waahaayy. Actually, I prefer image and colour coding - like Borland programs, or GnuBg. In GnuBg I don't read "Decline", I hit the red cross. I don't read the text "Accept", I hit the green tick. That's much simpler for a brain to learn.

Czuch Ahah. I know where you're coming from. Makes sense to me. I have a similar problem with a dodgy keyboard in as much as I "know" I've typed correctly but my shifts haven't and my programs fail because of syntax errors which shouldn't be there.

Fencer, Czuch: I know what I want to do with the cube but I still have to read twice before clicking and do the "Am I sure this is the right button?" routine. What harm in compromising and shading those buttons in green and red?