why is this move a blunder?

[admin]

Let's ask the experts at rgb..




The score (after 0 games) is: GreedyG 0, tetraHydro 0 (match to 3 points)
Match Information:

Date: 1/3/2020
Move number 17: tetraHydro to play 12

GNU Backgammon Position ID: Y7sGAwTMbhMDQA
Match ID : UQllAAAAAAAE

Pip counts: O 108, X 125

* tetraHydro moves bar/22

Cube analysis
2-ply cubeless equity -0.423 (Money: -0.428)
0.403 0.112 0.005 - 0.597 0.342 0.008
Cubeful equities:
1. No double -0.269
2. Double, pass +1.000 ( +1.269)
3. Double, take -0.389 ( -0.120)
Proper cube action: No redouble, take (8.6%)
Alert: very bad move ( -0.154)

Rolled 12 (-0.162):
1. Rollout bar/23 10/9 Eq.: -0.356
0.369 0.113 0.012 - 0.631 0.330 0.016 CL -0.501 CF -0.356
[0.002 0.003 0.002 - 0.002 0.002 0.002 CL 0.004 CF 0.005]
Full cubeful rollout with var.redn.
1297 games, Mersenne Twister dice gen. with seed 657938119 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 2-ply cubeful prune [world class]
2. Rollout bar/23 6/5 Eq.: -0.381 ( -0.025)
0.369 0.114 0.017 - 0.631 0.362 0.019 CL -0.521 CF -0.381
[0.001 0.003 0.002 - 0.001 0.003 0.002 CL 0.004 CF 0.005]
Full cubeful rollout with var.redn.
1296 games, Mersenne Twister dice gen. with seed 657938119 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 2-ply cubeful prune [world class]
* 3. Rollout bar/22 Eq.: -0.510 ( -0.154)
0.337 0.104 0.015 - 0.663 0.413 0.018 CL -0.639 CF -0.510
[0.002 0.002 0.002 - 0.002 0.003 0.002 CL 0.004 CF 0.005]
Full cubeful rollout with var.redn.
1296 games, Mersenne Twister dice gen. with seed 657938119 and quasi-random dice
Play: 0-ply cubeful prune [expert]
Cube: 2-ply cubeful prune [world class]
4. Cubeful 2-ply bar/23 5/4* Eq.: -0.594 ( -0.238)
0.328 0.099 0.006 - 0.672 0.478 0.039
5. Cubeful 2-ply bar/23 8/7 Eq.: -0.680 ( -0.324)
0.291 0.080 0.003 - 0.709 0.457 0.016


Output generated Fri Jan 03 23:54:25 2020
by GNU Backgammon 1.02.000-mingw 20130728 (Text Export version 1.108)

[admin]

excellent responses from Tim Chow and Paul on RGB
Thanks guys!


Paul
4:20 AM (13 hours ago)
- show quoted text -
Immediate tactics predominate here. A good proxy for the quality of the
play is how many good numbers it gives the opponent.
Your play gives the opponent good 3's, 4's, 5's, 6's.
The bot play gives the opponent good 4's, 5's, 6's.
Same with the bot's second choice.

If you like, you can do a more precise good-number count, taking into account
both dice. But you can see that your play is indeed much worse than the
alternatives, just by counting one-dice hitting possibilities.


Paul
4:24 AM (12 hours ago)
- show quoted text -
Also, re your plan of escaping with a 6 after the opponent makes the bar point:
Almost any play will pay off under some circumstances.
The question is: How likely is the event that makes your choice good.
It's a minute parlay -- your opponent would either have to miss a shot as well
as make the bar point. Or you would be hit and your 6 would be accompanied by
an entering number. It's not as if any 6 on your next roll vindicates your play.

Paul


Tim Chow
9:48 AM (7 hours ago)
On Saturday, January 4, 2020 at 2:29:39 AM UTC-5, greedygammon wrote:
> What am i missing..?
> Why is staying on the 23 point better?

I think Paul is right that the primary factor is that bar/23 10/9
duplicates your opponent's 5's to hit.

You're on the defensive here and are trying to avoid getting blitzed.

You're well behind in the race so it is *not* your strategic goal to
escape your back checker, as you seem to be thinking. Of your three
possible game plans "attack, race, or prime," racing is the least
promising in this position.

As a general principle, when you're trying to avoid having your straggler
getting blitzed, it's better to stay back out of harm's way than to step up.
It also typically gives you better chances of hitting a shot.

---
Tim Chow