Talk:Safety play

I wouldn't consider the first example as a safety play - just good play. To me a safety play is only relevant in a single suit. (Whether it is needed may depend on other suits however...)
 * Well, that's not the definition; "safety play" is indeed a quite wide term, and it includes, but is not limited to, safe play of card combinations in one suit. I think the example is OK. Duja 19:33, 30 March 2006 (UTC)
 * Hmm, I suppose you are right, that makes my comment about precautionary plays a bit wrong - they are distinct from safety plays. However, they should be mentioned - perhaps an example of leading up to honours catering for stiff honour onside would be ok? Cambion 15:34, 5 April 2006 (UTC)
 * You're certainly welcome to add it. Duja 16:20, 5 April 2006 (UTC)

I would consider a precautionary play to be cashing the A from KQ92 opposite A6543 in order to pick up the 4-0 break in front of KQ92. The imperfect saftey play seems fine though. Any thoughts? Cambion 18:00, 30 March 2006 (UTC)
 * That could be mentioned too. We don't have a list of plays per suit combinations, though, (but it's doomed to be a copy of Encyclopedia of Bridge). Duja 19:33, 30 March 2006 (UTC)

At the risk of belaboring the obvious, I think that this topic is going to cause trouble as long as the term "safety play" goes undefined. Until we reach consensus on its meaning, we're going to have difficulty agreeing on examples.

One problem is that authors seem to have the same difficulty that we've encountered: it's a difficult term to define. Most broadly, I suppose, it can mean generally playing safe to minimize the possibility of going set, usually in rubber bridge or in IMP team games. My sense is that such a definition is so broad that it's useless for our purposes.

On the other hand, defining the term to mean to the safest way to play a particular suit (while that's a valid topic) seems too narrow. I think that the definition has to take account of how the play in a particular suit relates to the deal as a whole. Certainly there's a place for discussing suit combinations, in the manner of Eric Crowhurst in the Encyclopedia. But it seems to me that a discussion of the considerations involved in playing AK8xx opposite Jxxx, cogent as they are in general, is marginally relevant in an article on safety plays unless in the context of the full deal.

I'm partial to Marshall Miles' definition, from an old book (How to Win at Duplicate Bridge), which remains fresh 50 years later (Hvala Bogu). He says, in the chapter titled "Is Duplicate Really Different?" that "By 'safety plays,' I mean playing in such a way as to lose a trick with average breaks in order to avoid losing additional tricks with bad breaks . . . Taking the safety play at duplicate is like betting even money on a ten-to-one long shot" – (from xlmvp: but you expect that making the apparently irrational bet is the best way of ensuring a good result on the full hand; also, the "additional tricks" that Marshall refers to are the ones that would set the contract – this ties in to the notion of the context of the full deal.)

Now, safety play, so defined, has much more applicability at rubber or IMP teams than at pairs or BAM, although I think the topic is intrinsically more interesting in the context of pairs (evaluating the contract and how the defense has gone, for example).

One classic example of safety play is the AKJxxxx, in an otherwise entryless dummy, opposite xx. If declarer is in an exceptionally good contract, or if the defense has already guaranteed declarer a good board, the recommendation is a low card from both hands – preserving communications as well as the option of finessing the J on the second round. But the play makes no sense unless the entire hand exhibits one or more of the conditions – the entry situation, the quality of the contract, or the nature of the defense.

As to the eventual article, I think it makes sense to mention some typical un-safety plays, such as cashing the A in AKJxx opposite xxx in a one-entry dummy, before finessing the J.

And in the article as it's constituted today, I vote with someone that the first hand, the pure elimination, doesn't belong in an article on safety play. There's nothing in it about losing a possibly unnecessary trick. (And the text doesn't match the trick 9 diagram, BTW, although that's beside the point.)

On the second, 4 spade hand, the preceding sentence in the text says "we will see that the hand shall be played differently depending on the form of scoring." I don't think we see that – I agree with the subsequent comment that "the choice is not clearcut at all." Is it a reasonably strong game, so that we can assume virtually all pairs will be in 4S? And the early play might be unusual – it gave South a trick in hearts that he probably would not have made if left to play the suit himself. So in a tournament where the quality of the opposition is uneven, so that 4S might be a relatively good contract, I'd be inclined to take the safety play, but in a strong game I think I'd play the AK. In sum, I agree that the deep finesse is a true safety play, but I don't think I'd bring in the question of whether to make the play at pairs. There are too many imponderables for the hand to make a crisp point about safety play at pairs vs. total points.

The third hand, 3NT, looks good for this article. Xlmvp 16:43, 7 July 2006 (UTC)


 * Well, what else to say but "I mostly agree to all points". The article currently offers somewhat simplistic view of the term. The first example, as you pointed out, is rather moot. It perhaps belongs more to Endplay article than here; it can be regarded as "safety play" only in a fairly wide sense of the word.
 * However, in my opinion, we should keep the scope of the article focused. How can I put it... this article should be about Safety play, and not dwelve too much into intricacies related with type of scoring, strength of the field etc.
 * That being said, we could have two more articles, covering some of those aspects, called e.g.:
 * Suit combinations, containing the list of percentage plays per each combination and number of tricks required, as in OECB, and.
 * Duplicate bridge tactics, which would discuss relationship between bidding and play tactics with type of scoring, field strength, strength of the opponents. It would necessarily be somewhat essayistic, so
 * Do you have a better proposal for an example of safety play on any type of scoring? (Although, it can be well argued that if it can't lose anything, it ain't "safety") Duja 12:47, 10 July 2006 (UTC)


 * P.S. "Hvala Bogu" ??!! (Yes, it's my mother tongue, but I don't get how a Californian with German-sounding name would plausibly use it?) Duja 12:52, 10 July 2006 (UTC)


 * >>Do you have a better proposal for an example of safety play<< I think so. I've edited the page, substituting for the first two hands in the prior version.


 * As to "Hvala Bogu" -- One of my very favorite authors is Rex Stout, who peppered several of his novels with Montenegrin characters, and of course his key character was Montenegrin. Stout used Hvala Bogu in Over My Dead Body, and I've been itching for an opportunity to use it for 20 years. I stretched a point and used it in reference to Marshall's landmark book. Much of Stout's The Black Mountain takes place "within sight of Tsernagora." (I know my name sounds Germanic, but I am of Swedish ancestry.) Xlmvp 17:20, 10 July 2006 (UTC)


 * D'oh! I was about to write an article about Belladonna coup and now you stole it :-). Seriously, I'd rather have a separate article about it, but then, we're again without a "perfect safety play" example... Duja 07:08, 11 July 2006 (UTC)


 * Duja: perhaps you'd like to remove the Giorgio (thanks for correcting my spelling) coup from the Safety Play article and substitute this – which is a "perfect" safety play.

--

Here is another Kelsey hand, this time from a team of four game:

South plays 6NT against the lead of the 10. Dummy's K wins, and the A and K are played. East discards a small spade on the second club trick.

This particular hand is one of a relatively small group in which perfect safety is available after the third trick. Single-dummy, using a safety play in diamonds, it is now possible to guarantee the contract against any remaining distribution and play of the E-W cards.

(Duja: I've never heard the term "perfect safety play" or "imperfect safety play" in English. And I can't find it with Google. Is it a standard term in French, or Italian, or Macedonian? If so, I guess it belongs in the article, but I would think that it belongs in its original language.)

South was expecting to win five clubs, one or two diamonds, four hearts and two spades, but the 4-1 split in clubs complicates things. Still, a safety play in diamonds will bring in twelve tricks.

It's just coincidence, but the proper diamond play on this hand is the same as the percentage play with this diamond holding, considering the suit in isolation. The best play for three diamond tricks is to play the A, and then lead toward either the Q or the J. This play brings in three diamond tricks 73% of the time.

Using that play on this deal brings in twelve tricks 100% of the time. Cash the A, and lead small toward the Q. Then:


 * If diamonds are 3-2, South will always win three clubs, three diamonds, four hearts and two spades.
 * If diamonds are 4-1 and West has the singleton, East cannot play the K on the second lead without setting up both the Q and the J. If East plays low and West shows out, concede a club and take four clubs, two diamonds, four hearts and two spades.
 * If diamonds are 4-1 and East has the singleton, West can capture the Q with the K. But in that case West is known to hold four cards in each minor, and on the run of the major suits will be squeezed out of his guard in either diamonds or clubs. In with the K, West can attack dummy's entry in diamonds or in clubs, but not both.

Reference: Kelsey, H.W. Test Your Match Play. Faber, 1977.

Xlmvp 03:16, 13 July 2006 (UTC)


 * (Duja: I've never heard the term "perfect safety play" or "imperfect safety play" in English. And I can't find it with Google. Is it a standard term in French, or Italian, or Macedonian? If so, I guess it belongs in the article, but I would think that it belongs in its original language.)
 * No, I believe those terms were actually invented ad hoc, but I kind of liked them. Although it's maybe slightly original research, or, rather, original terms, they IMO nicely described a point. For example, I also kind of invented the terms in Bridge convention, but I felt some taxonomy was called for. Duja 12:58, 13 July 2006 (UTC)

The start of the article is wrong.
I make reference to http://www.northerncoloradobridge.com/archives/playerscorner/SafetyPlays.htm in which we find the section: "Safety plays that give up nothing." The standard one is missing 4 cards including the jack to cash the top honour in the hand that has two, catering for the possibility of Jxxx in either hand. Yet the article claims that safety plays '[maximize] the chances for fulfilling the contract (or achieving a certain score) by ignoring a chance for a higher score.' This is just not true and ignores other common safety plays like AQxxx opposite xxx where one cashes the ace first (or ducks one) catering for the stiff K offsides. These safety plays cost nothing at all and should be used even at matchpoint scoring.190.102.142.82 (talk) 22:08, 25 October 2012 (UTC)


 * No. You're misinterpreting the term "safety play." (And by the way, so is Mr. Petrie in the reference you cite, at least in the section "Safety plays that give up nothing.") It is in the very nature of a safety play that it gives up something -- else it is not a safety play as the term has for decades been used in the bridge literature.


 * To say that the K first from a holding of A10xx opposite KQ98x is a safety play is to conflate "safety play" with "correct play." The term "safety" pertains not to the played suit. It pertains to the safety of the entire deal -- to insure against bad breaks or other adverse holdings that threaten a contract at rubber or IMPs, or that threaten the result of an unusually good contract at pairs. The safety play often involves playing against the odds in a particular suit, so as to make sure of a particular outcome on the full hand.


 * As such, the "start of the article" -- far from being wrong -- is apodictic. TurnerHodges (talk) 17:11, 18 November 2012 (UTC)


 * Quite by accident, I ran across the following hand not long after posting the prior reply:

It is from the semifinals of the 1991 Vanderbilt; NS vul. Kit Woolsey declared on the auction 1S-2C-2S-3C-3S-P-4S. West led JD. East won and continued diamonds. As Edgar Kaplan describes the continuation in The Bridge World: "([East] had a blind guess; West might easily have had more clubs and fewer diamonds.) Declarer ruffed the second diamond, played a spade to dummy's ace, then a spade back to his jack -- the finesse was now a safety play, since he could pitch two clubs from dummy on his hearts: plus 620."

Kaplan's use of the term "safety play" to describe the spade finesse was exactly right (as was nearly everything else he ever wrote). The percentage play in spades, given that there are no restricted choice aspects, is to play for the drop, slightly more likely to score five spade tricks than the finesse. But by going against the odds with the finesse, declarer guaranteed the contract. West could not profitably attack clubs, and declarer could later get rid of two of dummy's clubs and ruff a club at the end. If declarer played for the drop and East turned up with the guarded QS, East could ruff in before declarer could discard two clubs from dummy, and a club shift would give the defense a diamond, a spade and two clubs.

Declarer's play could also accurately be termed an "avoidance play," but that would obscure the irony that Kaplan wanted to point out in his description of the play at the other three tables.

My point is that the term "safety play" is apt here because declarer declined the percentage play in spades in order to guarantee the contract. TurnerHodges (talk) 20:28, 18 November 2012 (UTC)