Opening Rolls - Extensive rollouts by Snowie 4.1

Rene Cerutti

unread,

Apr 15, 2004, 2:18:05 PM4/15/04

to

OPENING ROLLS

Best Second best
Third Best Fourth Best

1 - 2: 13/11, 6/5 (0.024) 24/23, 13/11 (-0.017)
13/10 (-0.057) 24/21 (-0.058)

1 - 3: 8/5, 6/5 (0.235) 24/23, 13/10 (-0.232)
13/10, 6/5 (-0.233) 13/9 (-0.248)

1 - 4: 24/23, 13/9 (0.009) 13/9, 6/5 (-0.014)
13/8 (-0.052) 24/20, 24/23 (-0.060)

1 - 5: 24/23, 13/8 (0.012) 13/8, 6/5 (-0.013)
24/18 (0.035) 13/7 (-0.81)

1 - 6: 13/7, 8/7 (0.144) 24/23, 13/7 (-0.184)
24/18, 24/23 (-0.191) 13/7, 6/5, (-0.214)

2 - 1:

2 - 3: 24/21, 13/11 (0.013) 13/10, 13/11 (-0.005)
24/22, 13/10 (-0.012) 13/10, 6/4 (-0.042)

2 - 4: 8/4, 6/4 (0.169) 13/9, 13/11 (-0.153)
24/22, 13/9 (-0162) 24/20, 13/11 (-0163)

2 - 5: 24/22, 13/8 (0.001) 13/8, 13/11 (-0.004)
13/8, 6/4 (-0.032) 13/6 (-0.095)

2 - 6: 24/18, 13/11 (0.032) 24/16 (-0.021)
13/5 (-0.046) 24/18, 24/22 (-0.079)

3 - 1:

3 - 2:

3 - 4: 13/9, 13/10 (0.028) 24/21, 13/9 (-0.017)
24/20, 13/10 (-0.019) 24/20, 24/21 (-0.052)

3 - 5: 8/3, 6/3 (0.088) 24/21, 13/8 (-0.049)
24/16 (-0.077) 13/8, 13/10 (-0.086)

3 - 6: 24/18, 13/10 (0.020) 24/15 (-0.012)
24/21, 13/7 (-0.053) 13/4 (-0.054)

4 - 1:

4 - 2:

4 - 3:

4 - 5: 24/20, 13/8 (0.030) 13/8, 13/9 (-0.012)
24/15 (-0.021) 13/4 (-0.063)

4 - 6: 24/18, 13/9 (0.012) 24/14 (-0.001)
8/2, 6/2 (-0.019) 24/20, 13/7 (-0.063)

5 - 1:

5 - 2:

5 - 3:

5 - 4:

5 - 6: 24/13 (0.083) 24/18, 13/8 (-0.042)
13/7, 13/8 (-0.140) 13/2 (-0.218)

6 - 1:

6 - 2:

6 - 3:

6 - 4:

6 - 5:

Rollout settings: Snowie 4.1

Truncated rollout, depth 9

250 games (equiv. 78237 games),

played 3-ply

seed 1, with race database.

Rene Cerutti

unread,

Apr 15, 2004, 6:55:04 PM4/15/04

to

Bob Koca

unread,

Apr 20, 2004, 6:09:59 AM4/20/04

to

"Rene Cerutti" <cerut...@cox.net> wrote in message news:<wpEfc.9379$dZ1.8653@fed1read04>...

> OPENING ROLLS
>
>
> Best Second best
> Third Best Fourth Best
>
>

> ...

>
> 1 - 3: 8/5, 6/5 (0.235) 24/23, 13/10 (-0.232)
> 13/10, 6/5 (-0.233) 13/9 (-0.248)
>

> ...

> 1 - 5: 24/23, 13/8 (0.012) 13/8, 6/5 (-0.013)
> 24/18 (0.035) 13/7 (-0.81)

>...

Thanks Rene.
Surely though the results for 13/7 and the 1-3 plays besides 8/5
6/5 are misprints.

,Bob Koca

Zengirl

unread,

Apr 21, 2004, 4:11:15 AM4/21/04

to

> Thanks Rene.
> Surely though the results for 13/7 and the 1-3 plays besides 8/5
>6/5 are misprints.
>

>Bob KocaOoh.

Ooh, You beat my boyfriend (HFB) on Dailygammon. I watched that game.

Zengirl.

Michael Sullivan

unread,

Apr 21, 2004, 9:19:39 AM4/21/04

to

Bob Koca <bob_...@hotmail.com> wrote:

I'm not sure about the 13/7, but it's really hard for me to believe that
13/10, 6/5 is that much worse than 13/11, 6/5. I can certainly believe
it's worse, but not that much worse. Especially not that much worse
than 13/9, 6/5. What would be the reason that a builder on 9 and 11
with the slot are within .04, yet one on 10 is .2 worse than either?

That makes no sense.

Or do those minuses mean how much equity is given up to the correct
play, rather than absolute equity?

If that's the case, I could these being correct.

Michael

Gregg Cattanach

unread,

Apr 21, 2004, 10:08:17 AM4/21/04

to

"Michael Sullivan" <m...@panix.com> wrote in message
news:1gcl5nu.14ia17a19kx26xN%m...@panix.com...

My best guess is that part of the time he posted the differential between
the two plays instead of the absolute equity.

Here's the absolute cubeful equities I show (at 3-ply) for 31:
8/5 6/5 0.223
13/10 6/5 -0.006
24/23 13/10 -0.013
13/9 -0.024

the 13/7 should likely be -0.081 (I show -0.078).

Gregg C.

RedTop

unread,

Apr 22, 2004, 7:37:38 AM4/22/04

to

In preparation for their upcoming book, Nack Ballard and Paul Weaver
have rolled out some of these plays as many as 20,000 times, on
multiple bots, with multiple settings. While I thank Rene for the
effort, I'm not sure that 250 rollouts is really "extensive."

Rene Cerutti

unread,

Apr 22, 2004, 11:27:48 AM4/22/04

to

Or do those minuses mean how much equity is given up to the correct
> > play, rather than absolute equity?

YES.

Rene Cerutti

unread,

Apr 22, 2004, 11:29:50 AM4/22/04

to

I'm not sure that 250 rollouts is really "extensive."

About 4 days non stop on pentium 4 3GHz