July 10, 2020
42% | |||||
26% | |||||
14% | |||||
8% | |||||
2% | |||||
2% | |||||
0% | |||||
0% | |||||
0% | |||||
0% | |||||
| Total votes: 215 | |||||
|
Rosemarie44 2052 votes Joined: March 2016 |
Friday 3:14 AM
Another Jack to dealer today! Kept the pegger and tossed a really low valued T-K to dealer's crib. Every cut increases hand but a cut of a 4. Rosemarie44 says: As I see it, (tell me if I am wrong) we need to push ahead to CPZ 95-99. We want to be in that range (first) as dealer to have the advantage of a 50% win. Looks like offense.
If this strategy is correct as Ras mentions the other day, as offense we are only concerned with the pegs we score and not what the opponent scores. Eolus619 says: I understand what you have said. I just took the time to go to Liam brown link and look @ the max values of hand verse crib based on discard. It shows the potential pitfalls of being aggressive with discards ..not saying you are wrong in any way....became interested after your comment Rosemarie44 says: you will determine strategy again after you see the starter card. Sometimes you will choose safe pegs - optimal strategy. Sometimes you will disregard the pitfall from aggressive discards, and do anyway. Just go for peggs. |
|
mrob2199 1436 votes Joined: February 2009 |
Friday 3:15 AM
Almost a guaranteed 4 here with a very well disguised pegging hand which in many case might garner 3 to 4 pegs-plus the added benefit of the K-10 to opponents crib? Looks like a plan to me |
|
james500 3924 votes Joined: June 2013 |
Friday 3:41 AM
Three card sixteen, (T,5,A), with a possible fake flush (T,2,A), if Dealer doesn't respond appropriately.
Generally, a lack of synergy between a King and 7, by which I mean they can't be part of the same run or 15. They're both diamonds today though, so maybe a flush. Only the cut of a 6 is unhelpful to me. 10 lead. |
|
thelawnet 262 votes Joined: January 2020 |
Friday 3:50 AM
This one seems too easy? Trash dealer's crib with the horrible KT, keep the 5 of course, and keep all the good pegging cards.
Keeping 5TK gives us 0.7-0.8 more points on average, which makes sense, but arguments for this as an offensive hand fall down in that it will peg fewer points , so any offensive advantage is tiny , while defensively this fails in that any toss other than KT will concede more points, so it is not worth considering whether this is a defensive or offensive position. After the cut we can lead the 2. |
|
dec 6359 votes Joined: April 2008 |
Friday 4:03 AM
I kept peggers and will try to balk the crib here. We have four to start and will lead the two and will try to play into a run here at least. dec |
|
JQT 4143 votes Joined: October 2008 |
Friday 4:24 AM
I'm almost tempted to Toss (A 7) today; I mean, how bad could it be? Two Points is Two Points, am I right? And we desperately wish to get to Hole 95-ish and/or beyond, ASAP!
We took a real drubbing yesterday: After all, who thought Mr. Computer was so against leading the Deuce, that it was to become DISCARDED! (To be clear, I think the program was slightly more concerned with preventing the Dealer from pegging those originally Four, and then post-Jack Cut Two Holes, and it gave less emphasis on our own ability to WIN by pegging just One Hole; and thus to me at least, its numbers seemed to be based upon winning with First Hand Show as much as by pegging. And if programs do share one weakness, these often do show up and manifest themselves during such an intricate endgame. Anyway, that was yesterday.) Today, I am still inclined to Toss (T K) in what is a desired-offense situation, because sometimes a defensive discard is indeed the best OFFENSE. After the Jack Cut, I'll press on and lead the 7 Card, and take points in almost any case; I've even pegged over Two Dozen Holes from such crazy hands by "leading from the edges" and then "enticing the RUN(s)." We could quite easily 'make up for' those "lost" Two Points" during the pegging! JQT says: For example, pegging might proceed: 7 (7) 3 (10) 5 (15-2) 4 (19=3) A! (20) 6 (26) 2 'go' (28-8) and Dealer gets One for Last Card. Such pegging coups often require just the right cards, and often a slip-up by our Opponent, but I have SEVERAL examples of these in my game archives! Mark6 says: I too was surprised that all 15 Posters kept the 2 yet the Computer discarded it . And that the Computer chose a hand rejected by 99% of the 222 Voters . Does the wisdom of the crowds prevail here ? thelawnet says: The computer was just .... wrong. I am still thinking about it but at some point I will try and iterate through the possible tosses using a custom program and see if I can come up up with more sensible numbers |
|
Gougie00 5731 votes Joined: March 2008 |
Friday 6:00 AM
I've got better position. I will toss K-10 and go for the stifle.
Lead the 2 |
|
Eolus619 1343 votes Joined: June 2020 |
Friday 6:07 AM
Well....I had to decide “O” or “D” based on board position. So would it be K-10 or A-2? Either toss seemed to max my hand @ 6. So...my decision was to balk the crib and try not to spoil my lead. Plus, meager as it might be, it seemed to me that every cut helped my hand if I tossed the K-10. I now must be prepared to deal from probably being short in the CPZ. I didn’t even consider a 2-7 toss which the cut helped out considerably....Thanks to all for the pegging tutorials..whew do I have a LOT to learn about that aspect of the game. And...BTW...Also surprised.....when I looked at yesterday’s vote when total was @ 184...only one other plus Hal had voted for Hal’s discard. Rosemarie44 says: Hi: I think either JQT or Ras in particular could explain how Hal (cribot) as he calls it, comes up with the analysis we see at the end of the day. I forgot Couredelion who runs the program. thelawnet says: The author is quite active on twitter not sure if he answers questions
https://twitter.com/halscrib |
|
mfetchCT425 1400 votes Joined: February 2009 |
Friday 6:12 AM
I’m in the minority today. I kept the 4 points and will be as aggressive as I can during the pegging to try to get into position on 4th street. I will lead the K here and play/peg aggressively. Definitely see the logic with the 10-K toss mentioned by the gang above and maybe I’ve learned something today. |
|
JRCeagle78 1054 votes Joined: June 2016 |
Friday 6:17 AM
I have to press for my position. The high potential of pegging for this hand should get me where I want to go. Hopefully my attempt to blank the crib (T-K) will not be thwarted by the starter card.
I'll lead the 7, hoping to save the A-2 for more pegs later during the play. JRCeagle78 says: I just checked out yesterday's computer choice and I am just befuddled. Can anyone help me make sense of its determination ?? thelawnet says: I'm putting it down as imperfect algorithms.
Here's an interesting article about how one might do better.
https://www.quantamagazine.org/is-alphago-really-such-a-big-deal-20160329/
Essentially Halscrib is not likely playing 'like a human', because that wouldn't have been possible when it was written, but it would be possible to take the recorded results of games between skilled players and develop a bot that plays like a skilled human, and then improve from that.
A naive AI will play very very badly indeed, and simply training an AI against itself will typically result in results that are barely better than playing at random. While an algorithmic approach might do relatively well, but if we employ anti-computer techniques, i.e. we know that the computer follows a fixed set of rules, then we can defeat the computer.
In other words, what we can say is:
1) where every skilled human says that water is wet, and the computer insists it's dry, then the very best conclusion is 'computer is wrong', unless we have strong evidence that the computer is infallible, which we do not.
otoh 2) clearly a computer can easily evaluate every possible outcome of hands & crib with 100% repeatability and infallibility, so it has a substantial advantage over a computer, but
3) specific esoteric pegging scenarios are not at all a matter of taking the full set of millions of possible cards and trying to crank through them with brute force, but much more a matter of human intuition, which a computer doesn't understand at all, unless it is made artificially intelligent, which halscrib is not. hence
4) the logical approach would be to run through this hand a few hundred times against unsuspecting human players and see how they played the hand, and therefore which approach works best, and thus
5) completely ignore the results from the bot.... thelawnet says: "so it has a substantial advantage over a human", rather, " for hand evaluation".
So for example a basic bot can relatively easily be programmed that will beat naive humans, simply by choosing the hands that have highest expected value for hand + crib.
But that doesn't mean that such a program does not fail horribly in specific edge cases, and that a moderately skilled human might in fact perform far better than the bot, because the human finds it extremely easy to intuitively hold certain cards in the end game based on our 'neural network', whereas most compute cribbage has no 'neural network' whatsoever and will completely fail to learn anything at all from its experience, whereas we may learn to beat it where it is weak.
that doesn't mean that the computer is bad, but if we consider for example the example of music compression, then it might be that we can store digital music at a certain filesize and that humans will fail to tell the difference between the compressed music and the original. That doesn't however mean that if we found a 'torture test', say a particular percussion sound, that the algorithm for that sound is identical to the original. In other words 'very good' is not 'perfect', and not being perfect does not mean something is not very good! |
|
RubyTuesday 915 votes Joined: January 2019 |
Friday 6:38 AM
I kept the points, looked at the discard tables and saw that discards of A 2, A 7 and 2 7 would cost about the same so kept the 7 with my 5. |
|
wasa 3019 votes Joined: November 2014 |
Friday 7:20 AM
I will admit to second guessing myself. My goal right now is to get into position, therefore I want maximum points to get to the mid 90s. Is keeping the 2 points and 4 lowest cards better than defence and tossing the T-K? Mr. Brown has tossing the A-7 (or A-2 or 2-7) giving us 4 points and a better expected hand by 0.7-0.8. In hindsight, I think I agree with mfetchCT. |
|
SallyAnn3 908 votes Joined: March 2020 |
Friday 7:23 AM
Putting a 5 face into my crib and hoping! Thought of the a-2-5-7, but worse trash for my crib. SallyAnn3 says: ok wrong crib lol. Choice would be the A-2-5-7 then. |
|
glmccuskey 4102 votes Joined: April 2011 |
Friday 9:07 AM
With this hold a five gets me 10, a.six gets me 7 and a face card gets me 6-8 points. Plus a seven or eight gets me 6. I’ll lead the king and take any pegs offered. Hillchem says: I went for this offensive discard as well. It is not a great pegger, but with no small cards we might get a point or three. |
|
rob 1806 votes Joined: December 2013 |
Friday 10:18 AM
I kept the 4 points because I thought I'd have a better shot at reaching 95. But reading the comments above, I suspect K-10 was the better choice. |
|
cribbagepogo 3251 votes Joined: October 2007 |
Friday 11:41 AM
Like the K/10 but who am I. In the 80's a game set Crib King/Gin King had various opponents of different skill levels. Best I have seen since the. Ask the General Read. |
|
Mark6 702 votes Joined: June 2020 |
Friday 11:48 AM
Kept the 4 points too but think now I should have kept the Ace for a 31-2 instead of the deuce |
|
horus93 1282 votes Joined: December 2017 |
Friday 11:59 AM
Interesting puzzle. It'd be nice to get to 96+ but these cards aren't very good. Dealer is only 6 points past par and a 0 crib would be a game changer.
But cribs are very uncertain things. I'd rather keep the points. horus93 says: And I'd play total offense to the cut and lead the five. |
|
dgergens 938 votes Joined: January 2018 |
Friday 12:48 PM
My reasoning process is usually a bit simpler that the other major players here, and doesn't usually take in as many variable/considerations as the tournament players consider. So I'm always pleased when I arrive a the same conclusion. Although I still haven't memorized par holes, today I did use information learned on this site to decide. Using averages, in 2 deals from now, the dealer should be dealing again from 102, and should also be 26 points down the road at 113, scoring first. So seems to me anything I can do to get further down the road will reduce my need for an average score to win. And while I did give up 2 fer sure points, my pegging hand is the best possible and gave dealer a lopsided boat anchor in his crib. |
|
Coeurdelion 5595 votes Joined: October 2007 |
Friday 2:24 PM
I think there are 4 possible keeps for this hand - 2-5-10-K (A-7), A-5-10-K (2-7), 5-7-10-K (A-2) and A-2-5-7 (10-K). I don't like the A-7 discard though as it can score 12pts or 24pts if it combines with other 7s:
2-5-10-K: 4pts - 4¾pts (Schell: 4.89) = -¾pt A-5-10-K: 4pts - 5pts (Schell: 4.97) = -1pt 5-7-10-K: 4pts - 5pts (Schell: 5.07) = -1pt A-2-5-7: 2pts - 4pts (Schell; 3.88) = -2pts Potential: 2-5-10-K: Improves with 222, 3333, 555, 8888 + 14xXs = 28 cuts = 28/46 = 60.9% up to 8/10pts with 3333, 555, 101010, KKK = 13 cuts. A-5-10-K: Improves with AAA, 4444, 555, 9999 + 14xXs = 28 cuts = 28/46 = 60.9% up to 8/10pts with 4444, 555, 101010, KKK = 13 cuts. 5-7-10-K: Improves with 3333, 555, 6666, 777, 8888 + 14xXs = 32 cuts = 32/46 = 69.6% up to 7/8/10pts with 555, 6666, 101010, KKK = 13 cuts. A-2-5-7: Improves with AAA, 222, 3333, 555, 6666, 777, 8888, 9999 + 14xXs = 42 cuts = 42/46 = 91.3% up to 5/6/7/8pts with AAA, 222, 3333, 555, 6666, 777, 8888 = 24 cuts. Position: We're on target to reach 4th street positional hole in 9pts, less than average. Dealer is 6pts past 3rd street positional hole so I think we should slow them down and play Defense. Pegging: A-2-5-7 should peg best playing Offense or Defense I think. Summary: 2-5-10-K has the best starting value by ¼/1¼pts but A-2-5-7 has the most cuts for improvement and has 24 cuts for 5-8pts. It also should peg best so I'll throw the excellently balking 10-K. |
|
HalscribCLX 5318 votes Joined: February 2008 |
Friday 2:26 PM
At 87-76* playing an nOffense strategy for the pegging the dynamic expected averages and Win/Loss %s are:
______________Our Offense__Hand_Pegs__Crib___Total___W3 %____W4 % A-2-5-7___5.17+2.07+(-3.79)=3.45____48.7____57.4 5-7-10-K__6.00+1.50+(-4.88)=2.62____49.6____56.0 2-5-10-K__5.91+1.35+(-4.64)=2.62____48.3____56.5 A-5-10-K__5.91+1.46+(-4.82)=2.55____47.2____54.0 A-2-5-10__4.48+1.91+(-4.16)=2.23____41.9____49.8 A-2-5-K___4.48+1.87+(-4.27)=2.08____41.6____49.1 Offense_____L3 %____L4 % A-2-5-7______26.0____32.3 5-7-10-K_____28.9____33.2 2-5-10-K_____26.8____32.9 A-5-10-K_____28.4____34.9 A-2-5-10_____27.9____38.5 A-2-5-K______28.5____39.0 A-2-5-7 is best for expected averages by 0.83pt and is slightly best for Win %s and slightly lowest for Loss %s. So I'll select 10-K to discard. After the J cut I'll lead the 7 and play Offense: Lead_______Our Pegging Pts. 7_______________2.08 5_______________2.01 2_______________1.88 A_______________1.66 |
|
Ras2829 5155 votes Joined: November 2008 |
Friday 4:27 PM
Yes, it's off., off., off. at all decision points (placement of pegs, retention/discard, and after seeing the starter card). If choolsing offense wiht this hand, the 7 is the correct lead kee ping A-2 intact as might pick up a trey or be able to pair one or small cards. Also it 7 is paired have a chance to grab two pegs with the Ace for 15-2. The increased peging potential of A-2-5-7 and the low crib score of 10-K more than make up giving up two points in hand score. That is ofen the case in hands of four points or less in which cads of increased peggign vlaue can be retained. The hand scorer for !-2-5-7 increases with every cut other than a four post; so the potential hand value is closer than most would expect when compared to 5-7-10-K or 2-5-10-K. Ras2829 says: KNowing that I have a mere four points after the seeing starter card, peg aressively attempting to reach 4th Street CPZ (95-99). |
- Today's Results
- -
- Yesterday's Results
- -
- Past Results
- -
- Suggest A Hand
- About This Site
- Contact Webmaster by email: dailycribbagehand at gmail.com
- Visit our friends:
- eCribbage -
- Cribbage Pro -
- Huddle Around Games -
- Cribbage Discard Pro -
- Cribbage Corner