I was taught to "bid to make." That is, it would take 26 points to bid a game that was at least a 2 to 1 favorite to make. (I'd bid game with 24-25 only with "something extra" such as a solid five card suit that would take a disproportionate number of tricks.) Under these cirumstances, I would shun finesses (a 50-50 proposition) except as a last resort, because they "average down" my theoretical "67%" chances. Instead, I would look for an endplay, squeeze, or maybe a combination play (e.g. a drop play for 3-3 distribution in a key suit followed by a finesse.))

The IMP scoring table allows bidders to come out ahead if they make as many as 40% of the game contracts they bid. Hence, many players will now bid a speculative game with 22-24 points (instead of 26). More to the point, this result is occurring more frequently because of pre-emptive and "competitive" bidding, where the slightly stronger partnership is pushed into a game contract, instead of being allowed to stop at the three level. In this situation, a 50-50 shot at a successful finesse will "average up" their chances.

Do good players play finesses more frequently to "shorten" the odds when they are in long odds contracts with less than 25-26 points? In finance terms, do good players regard finesses as "dilutive" when they have 26+ points and "accretive" when they have 22?

  • Tom - You really need to read more on the history of the game. Culbertson, Statyman, Goren, Sobel, Jacoby were all bidding aggressively to game from the 1930's and 1940's on. You are being misled by the relatively passive bidding styles that they recommended to students, who wished to make their contracts with weaker playing skills. Commented Apr 16, 2016 at 20:00
  • @Aryabhata: If a finesse is is 50-50, are you more likely to play one when you have 22 points and an (otherwise) 40% chance of making your contract, or when you have 26 points and a 70% chance of making your contract? So if competitive bidding systems put you in more 22 point contracts than before, wouldn't you use the finesse more?
    – Tom Au
    Commented Apr 21, 2016 at 19:37
  • Your question is too vague. Sorry. I am going to vote to close.
    – Aryabhata
    Commented Apr 21, 2016 at 19:41
  • 2
    @Aryabhata. I think Tom's question is perfectly clear, and actually quite a good question. Commented Apr 23, 2016 at 14:35
  • 1
    @northernman: Sure. What constitutes a good question for the site is subjective. If you have an answer, please feel free to add it.
    – Aryabhata
    Commented Apr 25, 2016 at 20:29

3 Answers 3


Finesses, now as always, are worth half a trick (absent information about opponents' distributions). In IMPs scoring, it is usually appropriate to take a finesse if that is the most likely way to make your contract, and usually inappropriate to take a finesse if it risks the contract for a chance at an overtrick.

Just because a contract is less likely to make, it doesn't change the value of a finesse. If a squeeze, endplay, or combination play gives you better chances, you should do that instead of finessing. If no better alternatives exist, you should finesse.

It may be that finesses are more commonly employed by players who bid lighter games, but that doesn't reflect on their desirability.

  • My theory is that players who bid lighter games have fewer chances of squeezes,endplays, and combination plays but I could be wrong of course. To a large extent, that's what the question is about; do light bidders need to rely on finesses, and do finesses give them a good fallback position by "averaging up" results? Thanks for answering, though.
    – Tom Au
    Commented Apr 17, 2016 at 19:27

You are essentially asking for statistics on the tactics chosen by expert players over the course of playing many games, correlated with the ease and perceived necessity of making their desired score in those games. While fascinating, i don't think you can find this out other than by asking experts directly for several reasons:

  • stats of this kind are hard to find or not produced
  • it depends on the player and their "style"
  • good players aren't playing to make their contract anyhow; they are playing to equal (defensive bidding) or marginally beat the field, so you can't correlate play tactics to actual bids, you have to correlate to their estimate of the tricks they feel they need and the risks they feel safe in taking to obtain them in that game
  • good players may perceive extra information/insights which guide them to more effective plays (if you can better read the lay of cards maybe you don't need to guess/finesse 100% of the time with 50% error but only 60% of the time with a 20% error, or you are much more sure than 50% so they aren't a 50/50 chance any more, or something?)
  • Welcome to the site. An interesting answer.
    – Tom Au
    Commented May 21, 2016 at 0:26

I broadly agree with ruds, but want to give some thoughts based on Tom Au's comment on that answer and they didn't fit in a comment.

There is a confusion about a priori (prior) and a poserteriori (post-facto) knowledge going on here. The fact that bidding 4H on 22 points will not make most of the time doesn't necessarily mean that finesses are better plays when they are available, it could be that most 22 point hands simply can't make 4H, or can only make when the lie of the cards is highly favorable. When the game will make, it might well be that finessing is still a bad idea a lot of the time, because in these scenarios there is also a higher percentage line of play. There may simply be fewer opportunities to finesse with 22 points. Similarly, in 27 point games, it might be the case that when a finesse is available it is actually likely to be the right play because they are only commonly available in scenarios where other, higher probability plays, are unlikely. What makes for a good line of play is based on the hand you see, not the unknown of the bidding sequence.

I don't really think the latter is true, the former might be (experience tells me that cross roughing might be that competing line) but instead what I'm saying is that the analysis needs to be based on sampling a large number of hands, it can't be based on the sort of probabilistic argument you're proposing.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .