And here is the distribution of scores for the Dealer's Hand.

EV[Dealer's Hand] = 7.98.

The EV of the dealer's hand is actually a little less than the EV of pone's hand. This surprised me at first but I think a simple explanation is that the dealer is often just dumping some of their points into the crib, decreasing at times the value of the hand. I have been tracking dealer hand and crib info independently but this probably isn't the best way to track this because I'd expect the Dealer Hand and Crib to be negatively correlated.

Points	Prob	Prob <=
0	0.007069747	0.007069747
1	0.000514794	0.007584542
2	0.062170947	0.069755488
3	0.018106712	0.087862201
4	0.113797045	0.201659246
5	0.038720328	0.240379573
6	0.155812713	0.396192287
7	0.064339336	0.460531622
8	0.172061748	0.63259337
9	0.065406676	0.698000047
10	0.088580273	0.78658032
11	0.015070997	0.801651316
12	0.099757817	0.901409133
13	0.007736141	0.909145273
14	0.032742316	0.941887589
15	0.005187266	0.947074855
16	0.033206965	0.98028182
17	0.006902495	0.987184316
18	0.001801701	0.988986017
19	0	0.988986017
20	0.005743221	0.994729238
21	0.001746931	0.99647617
22	0.00039002	0.99686619
23	0.000209294	0.997075484
24	0.002856391	0.999931875
25	0	0.999931875
26	0	0.999931875
27	0	0.999931875
28	6.52298E-05	0.999997105
29	2.89516E-06	1

Sigh the info went on different pages so quoting this again so they both can be on the same page.

Cool. Definitely looking forward to crib scores. Any possibility of analzing pegging scores?

Yes I definitely want to do this as well but the problem is I’m not a great pegger so this will be harder to simulate accurately.

Crib scores are close to done and the averages I’m getting are almost all within 0.07 points of the averages in the cribbageforum tables so the distributions should hopefully be pretty accurate.

Very interesting, thanks.

A delightful thread! I thought I was the only person my age who plays this game. (Kioshk probably plays)


I cant remember exactly how old, maybe 8. Learned from my father. I know this game is more popular in New England & Canada than anywhere else in the USA. My father's family is from New England and it's always played at family gatherings. Have a lot of positive memories associated with the game, my father and I used this game a lot for bonding (which mainly was taunting each other during pegging Was never really close with Dad, this game is my most positive memory with him as kid.


Other memories:

1)When we finally taught Mom she beat me 6 straight times and I cried
2)I had a 28 hand once only, never have seen a 29 hand. Has anyone else?
3)Taught all my college roommates the game, we played this and Hearts constantly
4)Had my first kiss with a girl over a game of cribbage

Welcome to the thread!

This leads me to ask: How old is everyone?

I am 36 and def on the young side of the curve in my experience. I wish more people played, it's such a great game. Will never stop trying to get people into it.

52.

It would be particularly cool to see a comparison of dealer vs pone pegging averages.

If ecribbage keeps/allows game histories, it should be possible to get good numbers from actual play.

How did you calculate the discards? Evenly distributed or did you apply some strategy bias?

52 as well.

I saw 3 28 hands in a couple of week period as a teen. None before or since. My dad had a draw at 29 last week. Never had 28 myself.

My poor Dad. He plays a bajillion games. My late mom probably 80% less. My Mom got one about 5 years ago and didn't know what she had at first.

We have determined that my Dad will get a 29 but his opponent will go out before he counts at which time he will promptly have a heart attack. He does have a couple of hole-in-ones in golf.

I've had a couple 28s live, no 29 though. Got one on CribbagePro a while back however. Had to take a screenshot.

I'm 34. Learned really young with my family but we always play partners (usually 2v2v2). I'm fairly new to 1v1 cribbage really.



Repeat the following:

1. Deal the dealer a 6-card hand. Consider every possible combination of discards and compute the discard that maximizes EV[Hand + Crib]. EV[Hand] is done by considering every possible cut card (46 options), and EV[Crib] is done using the tables found here: http://www.cribbageforum.com/SchellDiscard.htm

2. Deal pone a hand, and do the same thing except maximize EV[Hand - Crib].

3. Now I have a 4-card Dealer Hand, Pone Hand, and crib. Consider each of the 40 possible cut cards, and record how many cut cards produced each score.

When I get time to process the output, I'll have a histogram for each discard: (e.g. if pone discards 78 to the crib, what is the probability that the crib scores a certain point total).



Thought about this a bit more, and I wonder if I can somehow estimate the pegging value of a hand by simulating a bunch of other hands for the opponent and checking each way of playing my cards and take the best one. It's only a rough idea, but there aren't so many options and brute force might be able to give a starting place.

Of course if actual data was recorded somewhere then that'd be best, but if we are going to get distribution data then to simulate it I'd actually need to know the order in which I should play the cards.


------------------

Probably could kill the program anytime. I've simulated over 80 million hands since last night and even have over 12,000 simulations of Pone discarding 55 to the crib (which is the least common discard by far). The issue is the tables are hard to read and really need to be represented by bar graphs (there's 91 of them for dealer discards and 91 of them for pone discards). I'm probably going to have to write a script to produce those so I might not be able to post it until tomorrow.

One interesting thing that I didn't know before, if a 5 (or two cards that sum to 5) go into the crib, the crib is guaranteed to score at least 2 points. The 5 may not score the points, but if it doesn't, then the other cards have to score at least two.

Okay I have all the charts for the crib information. It loads kinda slowly though as it recomputes it when you load it.

Charts were generated automatically, and unfortunately it used different y-axis scales which makes it a little annoying to directly compare some of them. I'll maybe fix this in the future but Google's API for generating these things was a huge pain in the ass and I'm sick of dealing with it for now.

Pone Discard Spreadsheet

Dealer Discard Spreadsheet

Finally bought the Cribbage Pro Online version app ($2.99). It's great to be facing actual humans again. Been playing a ton, but not running well.

Holla if anyone wants to play: richgangi

Ran silly hot against my grown sons the past couple of weeks. Literally 20-3 including up 3-0 in a best of 7 last night. Then I gloated. Then Karma. Got Crushed 4 in a row including skunked twice.

Hahaha gotta love it. The cards do not care.

Played a crazy league game a while ago where I got skunked 121-73 then the very next game doubled him 121-58. Good times!

New season of my local league has commenced....

On third st, ~a dozen behind as pone: 10 10 J Q 7 8 no suits. WWYD?

I think you need to take a chance for a big score. Toss the 78. I may go differently depending on actual scores, but you are in trouble regardless.

Dealer was around the 80 mark iirc....

Opponent has a good chance of winning on the next deal, whereas you might not even win on the third. I think this is a case for "damn the torpedos! Full speed ahead!" You may hasten the opponent's victory, but you really need to hit the gas hard.

Thanks for your thoughts.

That is precisely what I did, and why.

From John Sucklings Wiki:
Awesome.

Ah, man, and here I am with a job like a total shmuck.

End game scenario: You're dealer from 8 out, pone is 5 out. Cut's a 4. You have Q223 and pone leads an A. WWYD?