Point Count vs. Honor Count

by Marvin L French

Ace, king, queen jack: four, three, two, one. How simple! How easy to remember! This is the Work count, named after Milton C. Work, the mentor and employer of Charles Goren. It was based on the McCampbell count of 1915, publicized by Work in the 20s. He actually did not like point count (PC), writing this in Auction Bridge Complete, 1926:

“Many teachers and writers advocate schemes by which a bidder my determine mathematically whether his hand is strong enough to justify bidding an original No Trump... A more difficult, but much more satisfactory, method is figuring an Ace as one trick, giving to all other single honors and to all combinations of honors arbitrary values which in most cases are fractions of a trick, or one or more tricks and a fraction.”

By 1929 Work, in Contract Bridge for All, gave in to PC's popularity and accepted it for notrump bidding only. For suit bidding he wrote:

“Contract suit bidding is so simple that only the number of 'probable' and 'high card' tricks in the hand need be noticed.”

The great Ely Culbertson agreed with that in his 1936 Contract Bridge Complete, The Gold Book of Bidding and Play, while eschewing point count altogether. Instead he used “Honor Count” (HC) for both notrump and suit bidding. No doubt he could see that PC had two serious drawbacks: (1) 4-3-2-1 does not reflect the relative values of high cards well, especially for suit bidding, and (2) honors in combination are worth more than the same honors lying in different suits.

In his book Contract Bridge in a Nutshell (1952 edition), Goren wrote:

“The point count method for No Trump bidding as we know it today was first introduced by me in Winning Bridge Made Easy (1936) and has for many years been standard with an overwhelming majority of the most successful players in the country. The carryover into suit bidding was accomplished in 1949 and has now been universally accepted."

Notice that he gave no credit to Work (who had given no credit to McCampbell!).

Goren didn't mention that he used HC, not PC, for suit bidding up until 1944, when he gave PC and HC equal value for both notrump and suit bidding, as in his The Standard Book of Bidding (1944-49). “Take your pick, it doesn't matter,” he seemed to say.

Then he saw a way to gain ground on his rival Culbertson by using PC for all bidding. He knew that this was an inferior approach (giving him the benefit of doubt) but realized that most players wanted one simple hand evaluation method, not two, for all bidding. He started with Point Count Bidding in Contract Bridge (1950). The book was so successful that he followed it up with many others. In his highly successful system summary, Contract Bridge in a Nutshell, 1952 edition, he wrote:

“With the introduction of my point count method the honor trick began to fall into disfavor, and today it is all but obsolete. Even those authorities who sponsored the honor trick for twenty years have decided, after witnessing the acceptance of point count, to abandon the old table and to adopt the methods which you will find set forth in this and my other books. Such action became indispensable to survival.”

True, not because of PC's superiority but because of its popularity. “Authorities” had to join up or lose students and readers.

Culbertson was caught out and desperately tried to catch up, coming out with Culbertson Point Count Bidding: Improved and simplified 4-3-2-1 with the new rule of 3&4. But it was too late, Goren had cornered the market.

So what about this 4-3-2-1 count? The Four Aces (Oswald Jacoby, David Burnstine (later Bruce), Howard Schenken, and Michael T. Gottlieb) in their Four Aces System of Contract Bridge, 1935, featured a 3-2-1-½ count. These men were very qualified to write such a book, having as a team won 11 out of 13 major team championships between 1933 and 1935, while none were won by Culbertson's team. Their very complicated book did not sell well and was soon forgotten. Actually, the Four Aces count accurately provides the relative values of honor cards for suit bidding fairly well, but not for notrump bidding. For that they should have retained the 4-3-2-1 count. They made a mistake in having a point count that includes a fraction, and should have doubled the values yielding 6-4-2-1, which players might more readily have accepted. It entails larger sums but the arithmetic is simpler. Incidentally, Goren included the 3-2-1-½ PC in The Standard Book of Bidding, treating it as an equal to the 4-3-2-1 PC, despite the great difference.

A very fine player named George Reith advocated a 6-4-3-2-1 count (giving the 10 a point). The great Howard Schenken had this to say in The Education of a Bridge Player (1973):

“George might be considered the first authority on contract bidding. [His count] is much better than the 4-3-2-1 count in universal use today. However, in those days everyone used the honour trick table and Reith’s count never became popular.”

It is interesting that Schenken used the 4-3-2-1 count in his books, despite the above opinion.

In the September 2001 Bridge World, Doug Bennion defines a more accurate PC, Little Jack Points (LJP), as A = 6½, K = 4½, Q = 2½ and J = 1, which his research confirms as being superior to 4-3-2-1 provided an adjustment is made for honor synergy: add ½ point for each face card that is accompanied by a higher honor in its suit. Danny Kleinman improved this method by doubling the values, producing A = 13, K = 9, Q = 5, J = 2, with whole point adjustments instead of halves. He also added further adjustments to reflect the value of 10s when accompanied by 9s or higher honors, subtracting two points for a 4-3-3-3 hand, reducing values for singleton honors, and devaluing a hand with an unstopped suit.

The result is a count that has a valid strength relationship among the honors and recognizes the increased value of accompanied honors (as HC does!). Bennion researched only the value of high cards when balanced hand faces balanced hand. In other words, this is a notrump count, not a count for suit bidding, which would assign more value to aces and kings vs queens and jacks.

It is very sad to see the Work-Goren PC taught to beginners as the ultimate hand evaluation method, not the temporary rough tool it should be called, and to see it used faithfully by experienced players. “Faithfully” is the right word, because adherence to it is like a religious belief, blind to reason.

Now, what about HC? It has been defined in slightly different ways, but let's use Ely Culbertson's version with slight modifications. After all, he was the one responsible for the name and for making it extremely popular for 20 years, as Goren said. Here are the honor valuations:

The value for AJ10 is thanks to Danny Kleinman, not Culbertson. There are other opinions of authorities to note also. California bridge teacher Spencer Kapp said KJx is 1 and KQ10 is 1½, and not to count more than 2 HC in a single suit. Culbertson said that a king and a queen in different suits is worth 1 HC, and a queen and a jack in different suits is ½ HC.

It may seem strange that AKQ, AKQJ, and AKJ are treated as equals, but that is related (1) to their value in a suit contract, when a third-round winner is less likely especially when defending, and (2) in a notrump contract when a concentration of honors in one suit implies possible weakness elsewhere. (1) is less applicable for a declaring side in a suit contract, when playing tricks are more important than defensive values. (2) is less applicable in a notrump contract when the other suits have strength or when partner has shown a valuable hand for notrump and may have the other suits well-covered. When (1) and (2) are less applicable, the HC for these holdings can be increased slightly.

Here are two hands with different PC values (11 and 13) but identical HC values (2-1/2)

AQJx KJx xx xxxx

AQxx Kxx Qx Qxxx

Despite the different point counts, 11 and 13, these hands are of approximately equal value, as HC says.

If that isn't clear, consider a holding of a lonely king and a lonely queen in different suits. In the hand opposite are a lonely queen and a lonely king in the same suits. Total trick-taking ability is 0 to about 2½, depending on the location of the high honors in those suits. Now put the kings and queens in the same suits, KQx opposite another KQx. Total trick-taking ability is between 2 and 4, depending on the location of the aces. And yet the total PC (10) is the same in either case! HC says the former holding is worth 1½ HC and the latter 2 HC. Not an accurate ratio perhaps, but a lot better than saying the cases are identical.

The principle involved is that honors in combination are worth more than the same honors separated. PC adherents will say it doesn't matter in the long run, evaluation errors will even out. No they won't. PC is seldom if ever superior to HC, so its inferiority is a generally constant companion. Expert players use PC as a starting point in the evaluation of their hand, and make adjustments based on hand shape, location of honours, fit with partner, intermediate cards, opposing bidding, control cards, unguarded honours, and information on partner's suit length and suit strength as it becomes available during the bidding. The starting point is a poor one, unfortunately.

So what rules of bidding are compatible with HC? Those who require five cards for a major suit opening and a strong hand (perhaps game-forcing) for a two-over-one response can stop reading here. They must develop their own rules for HC requirements. What follows is roughly based on Goren’s Standard Book of Bidding (1949). It features four-card major opening bids. Partnerships are free to modify his rules as they wish, establishing HC requirements and forcing/non-forcing attributes for each action.

For a one-level suit opening bid, it depends on one's bidding philosophy. With no suit longer than four cards, but maybe a “poor” five-card suit, it's 2½ HC for light openers (“lights”), 3 HC for others (“heavies”). With a good five-card suit (two of the top four honors, plus at least a 9 with AJ, KJ, or QJ) it's 2½ HC for either philosophy. A four-card major suit should be at least Q10xx or better. A six-card or longer suit, however weak, may be opened with 2½ HC.

Lights can open 1NT with 3-1/2 to 4 HC, heavies should have 3½+ to 4+. That means a jump rebid of 2NT is 4½ for lights and 4½+ for heavies, and a jump to 3NT ½ HC more than that. I recommend 2½+ to 3+ HC for a weak 1NT opening, with a 1NT rebid showing 3½ to 4 HC. Lights can eliminate the + if they wish, but their 1NT rebid then has problems.

A 1NT or one-over-one suit response (1 HC minimum but not a bare ace) should normally show at least Q10xx for a major suit. Not an absolute rule, but weaker majors should be bypassed if another bid (e.g., in notrump) describes the hand better. With 2 HC consider a two-over-one response, but a bare-looking 2 HC with no good suit is insufficient.

Single raises of an opening bid require at least 1 HC, a non-forcing (limit) jump raise 2½ HC, and a forcing jump raise 3 HC. For all these raises, with four-card support count a singleton as an ace and a doubleton as a king. Reduce that by a + with only three-card support, which is supposed to be J10x or better (Three small is okay with a side singleton or a side doubleton with too strong a hand for a 1NT response).

A two-over-one response requires 2+ HC, too strong for a 1NT response over a pass. A long strong suit can count as 1/2 HC extra. It does not promise another bid.

A game-forcing 2NT response requires 3 HC, but a strong five-card suit is worth a +. A 3NT response shows 3-1/2 HC or more, preferably in a 4-3-3-3 hand.

Opener's raise of a one-level suit response, or of a 2 response to 1♠, requires no extra strength. A 1NT rebid is usually preferable with a minimum 4-3-3-3 hand The raise of a 2♣ or 2 response requires an extra ½ HC, but four-card support or two of the top three honors can count as “extra” when opener is stuck for a rebid. After a two-over-one response, a 2NT rebid requires 1 extra HC, and a jump to 3NT an extra 1½ HC. Why the difference between hearts and a minor? Because a light 2 bidder can pass a raise to 3 without worrying about missing a notrump contract, as a minor suit bidder would. A raise to 3♣ or 3 will tempt responder to bid 3NT with 2½ HC, not likely to be successful if opener also has 2½.

Jump raises from one to three of a suit by opener require 4½ HC, but as with responder's raises a singleton can count as an ace and a doubleton as a king. With more, bid game.

Jump rebids in the same suit by either partner are based on playing tricks, not HC exclusively. A jump in the same suit is invitational and a jump in a new suit is forcing to game.

New suit rebids by responder are forcing unless opener has rebid 1NT. In that case a new suit by responder is weak, non-forcing, unless it is a reverse (forcing for one round only).

Of course there are many dozens more partnership agreements that must be agreed. These are only the basic ones.

In his book The Secrets of Winning Bridge Jeff Rubens advises players to focus on just a few hands that partner might be holding, and more particularly on perfect minimum hands compatible with the bidding. This means that in order to reach an informed decision in, for example,deciding whether a hand is worth an invitation to game or slam, you should “visualize” the most balanced distribution with the minimum strength partner might have with the high cards selected such that these fit precisely with your own hand. He advises that “your hand is worth an invitation to game (or slam) if this perfect minimum holding for partner will make it a laydown.”

Also available as a PDF file.