Staying alert

"I was reading The year in review 1999-2000 where you were holding 2-4-7-8 against Hal Lamon, needing to peg one point (after a 6 cut) to go out. You led the 2, instead of the usual 4, since it scores on more replies. How do you condition yourself to be able to look at a hand like 2-4-7-8 and, within a polite amount of time, calculate whether the 2 or 4 is the favorite to score?"

- Pat Hayslett (Potsdam, NY)

Finding unusual plays like this requires a combination of knowledge, alertness and judgment. First you have to know the concept (in this case, maximizing your immediate winners, a common principle in endgame pegging). Then you have to be alert enough to spot its possible application in a game situation. Then you must judge whether competing factors (such as covering your opponent's scoring cards) weigh in favor of an alternative play (in this case leading the "normal" 4).

Studying the game will increase your knowledge of cribbage, while analysis and experience will improve your cribbage judgment. That leaves us with alertness. It's not easy to concentrate on every play of every game, especially at the end of a long tournament, and even the best of us occasionally succumb to fatigue or distraction, and make a mistake such as misreading a card, miscounting a hand, or forgetting to change our way of thinking when the endgame rolls around. I haven't figured out how to eliminate careless plays, but I have come to realize that in order to be alert when you really need it, you pretty much need to be alert all the time.

Fast play

"I've watched my father play cribbage recreationally, and things move pretty fast. How long does a typical tournament game take to play? Does it go more slowly because good players are constantly thinking about discard averages, opponent's cards, etc.?"

- Philip M. Langlois (Belchertown, MA)

Top level players do tend to think about things more deeply, but they're also generally experienced enough to be able to count their hands quickly and play many common card combinations from memory. In over-the-board tournaments, games typically last 15 minutes, while in Internet competition they go considerably faster, since players are spared having to shuffle, deal and count the cards.

Finding tournaments

"I would love to play in a cribbage tournament in my area but am not able to find links for information. I live in Stamford, CT, and am willing to travel a bit in the New England/New York area."

- Carol Bonaire

I receive many enquiries like this. The American Cribbage Congress is the game's preeminent sanctioning organization, and offers weekend tournaments and local club play (Grass Roots) all over the US and Canada. They also sanction organized Internet play, which you can enjoy regardless of where you live.

Cutting for deal

Cutting the deck to determine who deals first is a longstanding controversy in tournament cribbage. Some events require a cut before every game, while others have all players deal first in alternate games. Paul Gregson (Antioch, CA) writes:

"I'd be interested on your take on cutting for deal. I just got home from Grass Roots, where I lost all nine cuts for deal, and a few months back I drove several hours to a 14-game tournament where I lost the first 10 cuts. I know that these things even out over time, but you can't count on that happening over the course of a 9-, 14- or 22-game tournament, and it's frustrating to travel a long way and get hosed by the odds like that! I think tournaments should not have cutting for deal. Let everybody deal first half the time, so that we all have a shot at it!"

Paul, I can imagine nothing fairer than granting players the privilege of first deal exactly half the time. Between equal opponents, the first dealer will win 56–57% of the time, so this one chance operation represents a 12–14% swing in winning chances before a single playing decision has been made. And that's not all: first dealer will win a skunk about 8% of the time, compared to 4–5% for first pone. So using the most common valuation of three game points for a skunk and two for a normal win, the advantage of dealing first averages out to about game point. That's 3 full game points over the course of a nine-game Grass Roots event, and roughly 7 over the course of a 22-game Tournament Trail qualifying round. We've all missed qualifying in tournaments by a lot less than that!

In addition to setting up extraordinary luck swings, the cut for deal is particularly vulnerable to abuse by unscrupulous players. Controlling the position of a few key cards (such as As and Ks), is not very difficult for a competent mechanic. I usually insist on shuffling the cards last prior to cutting, and I try to give the deck a firm grip to make sure there is no pre-existing crimp. I do this not because I suspect every opponent, but because I don't wish to single anyone out. Alternating first deal eliminates the need for all this, in addition to slightly speeding up the game and allowing skillful play to more consistently make a difference in individual games.

The proponents of cutting for deal try to downplay the statistical impact of losing the cut on your winning chances. They also point to tradition. But I'm not persuaded by either argument. If the statistical impact is slight, then why fight to retain it? And if we slavishly upheld traditional rules in cribbage, then we'd still be dealing out just five cards per player, pegging to 31 just once per deal, and ending the game at 61 points. I might point out that no one seems to object to the practice in Tournament Trail playoff matches of allowing the loser of the previous game to deal first in the next game (without cutting). Ultimately, the proponents of cutting for deal have no real argument beyond nostalgia for a custom that has little place in serious competition.

Despite all this, I'm not in favor of requiring tournament directors to use alternate deals. There's already too little diversity in ACC tournament formats, and I think TDs should have more flexibility rather than less. I'd rather see the Board of Directors urge TDs not to cut for deal, and when it's still to be used, to at least mandate a more tamperproof kind of cut, such as the OUCH cut devised by Hal Mueller:

1. Player A cuts a card from the deck, but doesn't reveal it
2. Player B guesses whether the rank of the cut card is over or under 7. If the guess is correct Player B deals first. If it's wrong, Player A deals first
3. If the cut card is a 7, the deck is reshuffled and the process repeated with Player B cutting and Player A guessing. Repeat until someone cuts something other than a 7

Scoring combinations

"I'm a senior who helps out in a local elementary school where the teacher uses cribbage as a vehicle to teach math. Is there a chart that would show, for example, what you would have to have in your hand to get to 21 points, 24 points, etc.? I have already explained that you can't get 19 and I have showed them how you can get 29."

- Paul Anderson (Marietta, GA)

Joe Wergin's book Win at Cribbage includes a listing of all the possible high scoring hands. Unfortunately it's out-of-print, but you should be able to locate a used copy through one of the Internet book vendors, or else borrow a copy through interlibrary loan.

Superranks

"I enjoyed the September 2003 Ask the experts. I'm glad most of the panel agreed with me. I would like your opinion on this hand: score 114-114* and I'm dealt 2-2-3-7-8-8. What would you keep?"

- Roland Hall (Napa, CA)

Over the board with no J or flushes to worry about, I'd probably count superranks, which I define as winning ranks + close ranks, the latter being starter ranks that get you close to the game hole. In this particular case I'd consider close to mean the stinkhole or better. 3-7-8-8 wins on a 3, 4, 5, 6, 7, 8 or 9 and gets close on anything, for a total of 20 superranks. 2-7-8-8 is inferior, winning on a 2 or 5 but not a 3 or 4. 2-2-3-8 wins on an A, 2, 3, 4, 5, 8, 10, J, Q or K and gets close on a 7, missing altogether on a 6 or 9, thus making 21 superranks. Based on this I'd keep 2-2-3-8, which I think will win somewhat more often than 3-7-8-8.

The theory here is that I consider myself only 50% or so to peg one point with a hand like 3-7-8-8 in this position, where dealer will presumably be playing desperation defense. So holding six points after the cut is worth half as much equity as holding the full seven. This is a simplification, of course, but it's a reasonably practical one, and I find superranks much easier to calculate over the board than the full endgame count (described here), which is nevertheless more accurate for complex hands, such as those involving flushes or His Nobs.

Cribbage magazines

"What a fantastic site! Besides yours, are there any magazines on the game?"

- Nick McBride (Kingston upon Thames, England)

Thanks for the kind words. The only print magazine devoted exclusively to cribbage is Cribbage World, the monthly house organ of the American Cribbage Congress. It consists mainly of tournament reports, schedules and standings, with little in the way of playing tips. The ACC does maintain a lot of material on its Web site though, including tips articles and regular features on ACC Internet cribbage. Years ago, there was another print magazine devoted to cribbage. It was called Cribbage Today, and was edited by Dan Barlow, who compiled highlights from its brief publishing run into a book called Fun with Cribbage.

Skunk value

We received a couple of questions regarding skunks:

"I had an enquiry recently about skunking: my correspondent wanted an 'official' ruling as to whether a skunk was worth two games. I looked this up in the ACC rule book and found to my surprise that although they define the terms 'skunk' and 'double skunk', they do not lay down anywhere whether there is a special award for these."

- John McLeod (The Card Games Web site)

"How do you score a skunk in cribbage? Hoyle says one game, but some others say two."

- Judy King (Lebanon, NH)

There is no set valuation for skunks in sanctioned play. In Grass Roots and the qualifying round of most ACC sanctioned tournaments, normal wins are worth two game points while skunks are worth three. In playoff matches, skunks are generally disregarded entirely. On the Internet things vary depending on the site: MSN Games counts normal wins as two and skunks as three, whereas PlaySite counts skunks as two normal wins. Double skunks are never recognized in serious competition.

In gambling games, the most common arrangement is for skunks to count double, but you should get an explicit agreement on this beforehand if you're playing an unfamiliar opponent.

By the way, Edmund Hoyle died in 1769, and never played six-card cribbage. Don't put too much credence in latter-day authors and publishers who wrap his name around products of dubious legitimacy.

See you all next time. Until then, keep on pegging!

- January 2004