The three most important pieces of information you can convey in an auction are game-forcing values, a fit for partner’s suit, and shortness. Splinters show all three things with one bid. That’s why they’re my favorite convention in bridge.
Splinters always show three things:
- Game-forcing values.
- 4‑card (or longer) support for partner’s suit.
- Shortness (singleton or void) in the suit being bid.
A splinter is generally a double jump into an unbid suit. A classic splinter auction is 1♠ – 4♣. In some auctions, when a natural jump-shift is not necessary, a splinter can be made with a single jump.
A splinter is always game-forcing, but the strength for a splinter varies based on how strong a hand partner has shown: you need enough strength to create a game-force opposite partner’s weakest possible hand. Because you have a fit and shortness, you can force to game with fewer HCP than you would with a balanced hand or without a fit.
- If you are responder, your partner has an opening hand, so you only need about 10 HCP. (You can think of this like a Limit Raise that has upgraded itself [or accepted its own invitation] because of its ruffing value.)
- If you are opener and your partner has made a one-over-one response (for example 1♦‑1♠), she has only promised 5 or 6 HCP, so you need a hand that evaluates to around 20 points – around 17 HCP plus the singleton) to force to game.
- If you are already in a game-force, no extra values are needed.
As you know, voids are more powerful than singletons, so you can be a couple HCP weaker to splinter with a void.
Always having 4‑card support for partner when you splinter is very important. Your splinter is announcing ruffing potential, and that potential is significantly less with only 3‑card trump support. Partner needs to know she can count on you for that fourth trump.
It’s also very important that splinters have a narrow strength range. Bids that take up this much room need to be very tightly defined, otherwise they cause more problems that they solve.
- Responder’s splinter shows around 10‑12 HCP.
- Opener’s splinter shows around 17‑20 HCP. (With more, opener likely would have opened 2♣.)
- Splinters in GF auctions can have a wider range, since they take up much less space.
Here are some example 4♣ splinters for both opener and responder in support of spades:
Responder Opener
♠ KT65 ♥ AJ65 ♦ K543 ♣ 3 ♠ KQ54 ♥ AKJ3 ♦ AQ63 ♣ 2
♠ J983 ♥ K43 ♦ AK976 ♣ 7 ♠ AJ32 ♥ AQJ643 ♦ K43 ♣ —
♠ A652 ♥ KJ874 ♦ T652 ♣ — ♠ AT98 ♥ A62 ♦ AKQ95 ♣ 6
The Splinter Bid
The Classic Splinter
The classic Splinter bid is a double jump in a new suit after partner has bid 1M. Usually this means jumping to the 4‑level, though when your short suit is spades the splinter bid will be at the 3‑level. Both opener and responder can splinter – opener’s splinter just shows a much stronger hand since responder could be very weak.
A few examples:
Opener |
Responder |
|
Opener |
Responder |
|
|
|
|
|
1♠ |
4♣ |
|
1♥ |
1♠
|
|
|
|
4♦ |
|
Opener |
Responder |
|
Opener |
Responder |
|
|
|
|
|
1♥ |
3♠ |
|
1♦ |
1♥
|
|
|
|
3♣ |
Not a splinter! (Only a single jump) |
Splinters in GF Auctions
Notice that in the classic splinter auctions the splinter bid is the cheapest call in that suit for which we do not need a natural definition. Take the auction 1♥-1♠; ? Say opener wants to bid clubs. We need natural definitions for both 2♣ (second suit, minimum or intermediate hand) and 3♣ (GF jump-shift); the first available bid in clubs that isn’t needed to show a club suit is 4♣, so that’s what we use for the splinter. When a single jump isn’t needed as a natural bid, we can use it as a splinter, saving some valuable room.
The most common time when a jump-shift isn’t needed as a natural bid is when we are in a game-forcing auction. One of the principal reasons we strive so hard to set a game-force at a low level (via 2/1, Fourth Suit Forcing, etc.) is so that we can save room when we make natural bids by not jumping, leaving us as much space as possible to explore for game or slam. Playing 2/1 GF, there is no need for opener to ever jump to 3♥ to show a heart suit after 1♠-2♦; she can just bid 2♥ and keep the auction low. Maybe her partner has spade support and is planning to bid 2♠. Or perhaps her diamonds are excellent and she intends to rebid 3♦. However the auction proceeds, there is very little to be gained by taking up an entire extra level of bidding with a natural bid.
Once we see that 3♥ is not needed as a natural bid, it becomes available for an artificial definition. And the most valuable definition we can give it is a splinter. So in GF auctions – when we do not need natural jump-shifts – we can splinter with a single jump. In this case, a 3♥ bid would show 4+ card diamond support and a singleton or void in hearts. This consumes one extra step of bidding space (without the splinter available, opener would make a natural raise of 3♦) but communicates not just the fit but also the essential information of shortness. Because the single jump takes up so much less space, it becomes much more practical to splinter in support of a minor suit.
Examples:
Opener |
Responder |
|
Opener |
Responder |
|
|
|
|
|
1♠ |
2♥ |
|
1♥ |
2♣
|
4♣ |
|
|
3♠ |
|
Opener |
Responder |
|
Opener |
Responder |
|
|
|
|
|
2♣ |
2♦ (GF) |
|
1♥ |
1♠
|
2♠ |
4♣ |
|
2♣ |
3♦
Not a splinter! (Not in GF) |
Minor-Suit Splinters
It’s possible to splinter in response to a 1m opening bid, but this agreement is much less common – be sure to discuss it with your partner before trying it out. Minor-suit splinters promise better trump support – at least 5 cards – since opener’s suit could be so much shorter than after a 1M opening. They also require a much stronger hand, since the most likely game after a minor-suit splinter is 3NT. Now we need not only full game HCP values, but usually a little extra, since the hand is a misfit (opener must have values opposite responder’s shortness if she is bidding 3NT), and misfits require extra strength. Responder should aim to have at least 15 HCP for this splinter. The auction 1♦-4♣ is especially problematic, since it bypasses 3NT; responder should have serious slam expectations to make this bid. Opener can make a minor-suit splinter after the auction starts 1♣-1♦; this bid is quite rare, but there is no other logical meaning for 3M.
Examples:
Opener |
Responder |
|
Opener |
Responder |
|
|
|
|
|
1♦ |
3♠ |
|
1♣ |
1♦
|
|
|
|
3♠ |
|
Auto-Splinters
On rare occasions, we can splinter in support of our own suit. This only happens when partner has shown a balanced hand, which lets us be confident of a fit when we have a 6‑card suit. The most common example of an auto-splinter is after a Jacoby Transfer:
Opener |
Responder |
|
|
1NT |
2♥
|
2♠ |
4♦ |
Responder is showing slam interest with 6+ spades and shortness in diamonds.
Evaluating After a Splinter
Splinters are a slam-exploration tool. They are not there to help us get to game (they force us to game) or to help us find a fit (they establish a fit); they are there to help us find good slams (and stay out of bad ones). One of the most common types of “light” slams (i.e., slams that make with less than the normal 33 HCP) is when we have a good trump fit and one (or both) hands have useful shortness; that’s exactly what splinters are there to find.
Wasted Values
The most important word in the last sentence – and the one most easily overlooked – is “useful.” All singletons are not created equal: their value lies in what partner holds in the suit. Say partner has a singleton club and your clubs as ♣ 8432. That’s great! His singleton has turned your four losers into one. That’s very useful. Say instead your clubs are ♣ KQJT. Not useful at all: you still have the same one loser you had to start with.
We refer to honors (other than aces) opposite partner’s shortness as “wasted.” It’s not necessarily that they’re worthless – in the example above when you hold ♣ KQJT, you’re still going to knock out the ♣ A and have three club tricks. But they’re wasted in the sense that they represent an over-evaluation of the partnership’s strength. Both you and partner have given a positive value to your club holdings: you counted your clubs as 6 HCP, while partner counted his singleton and fit as valuable assets and upgraded his hand – if he splintered as responder, he should have Limit-Raise HCP values that he has upgraded to a game-force based on the club singleton. Hand evaluation – whether counting HCP or upgrading a hand based on shape – is an attempt to predict a hand’s value at taking tricks. Tricks can only be taken one way – either by high cards or by ruffing – but not both. So your club assets are competing with each other, rather than complementing each other: they give you two ways to win the second round of clubs (you could cash your ♣ Q or ruff it in the dummy), but that’s not particularly useful.
Put another way, in both the examples above (when you held ♣ 8432 and ♣ KQJT opposite partner’s singleton) you have one club loser. In one you are using 0 HCP to hold the suit to one loser, in the other you are using 6 HCP. When you’re looking for a light slam you need all of your honors to be working together; the problem with having all those club honors is that 6 of your precious HCP are used up in clubs, meaning they’re not in the other suits, where partner’s honors must be. We all know that honors in combination are more valuable than honors by themselves. (Would you rather have your major suits be ♠ KJxx ♥ Qxx or ♠ KQJx ♥ xxx?) The “wasted” club values don’t mean you won’t take club tricks with your honors, but they mean your honors are in the wrong place, isolated from partner’s honors (in the other suits) rather than in combination with them. The ♣ K doesn’t lose value because it’s opposite the singleton, it loses value because it’s not the ♦ K. Let’s look at the full hands:
Responder’s hand: ♠ QJ32 ♥ 432 ♦ AQJ32 ♣ 2
Opener’s hands: 1. ♠ AK987 ♥ A8 ♦ K7 ♣ 8432 2. ♠ AK987 ♥ A8 ♦ 87 ♣ KQJT
Hand 1 is 3 HCP weaker, but it has the right HCP – the ♦ K, filling in responder’s suit, rather than the ♣ KQJ opposite responder’s shortness. Slam with hand 1 is excellent – barring awful splits declarer has 5 spades, 5 diamonds and 1 heart, and can ruff a club for his 12th trick. With hand 2, slam requires the diamond finesse (assuming they lead a heart). Not awful, but not nearly as good.
Your primary evaluation when partner shows shortness should be about wasted values. If you have wasted values, you will need significant extra values for a slam; if you have minimal wastage, slam might be possible with less than the normal 33 HCP.
The best holding you can have opposite a splinter is xxx(x). Absolutely nothing wasted. Next best is Axx(x). The ace would be more valuable in another suit, combining with partner’s cards, but it isn’t a duplication of values, since your holding wins the first round of the suit and partner’s singleton controls the second round (and beyond).
The next best holdings would be Jxx(x) and AJx(x). Just 1 HCP wasted. Holdings get worse as you add queens and kings, with KJx(x), KQx(x), KQJ(x) being the worst.
The 34-Point Deck
When you don’t have wasted values opposite partner’s shortness, it means all of the partnership’s high cards are concentrated in the other three suits and usually complementing each other. The opponents might have 10 HCP in the short suit, but that’s only worth 1 trick. We call this concept the 34‑Point Deck. A normal deck has 40 points. When you have a fit and shortness, the king, queen, and jack of your short suit are basically valueless; there are only 34 relevant HCP – the ace of the short suit and the 10 HCP in the three other suits.
In a normal 40‑point deck situation you need about 33 HCP for slam. That’s 82.5% of the HCP. When you are in a 34‑point deck situation, you don’t need the same number of HCP, you need the same percentage of relevant HCP: you need about 82% of the 34 relevant HCP. That’s about 28 HCP. (28 is 82.35% of 34).
When you find a good trump fit and have shortness opposite no wastage, the magic number you’re hoping to find for slam isn’t 33 HCP, it’s 28. That means just a very little bit extra is needed to get from GF values to slam values. You should be very aggressive in moving toward slam when you have no or limited wastage opposite partner’s splinter.
To communicate your interest in slam to partner, you control-bid (or ask for keycards). To communicate your wasted values and lack of interest, bid game immediately (Fast Arrival).
Common Splinter Questions
Are Splinters on in Competition?
Because they are both very useful in competition and usually bids that have no other logical meaning, it’s a good idea to leave splinters on when the opponents enter the auction. Splinters are especially valuable in suits the opponents bid naturally. The biggest caveat here is that the bid you intend as a splinter must not be needed for a natural bid. For example, after 1♥ (X) or 1♥ (1♠), a jump to 4♣ has no other useful meaning and should still be a splinter. After 1♥ (3♦), 4♣ can’t be a splinter, since you need it to show clubs.
Should I Splinter with an Honor?
This is a classic conundrum: should you splinter when your singleton is an honor. I prefer not to. The point of a splinter is to help partner evaluate her honors; if you splinter with a singleton ace, partner will think her king in that suit is wasted, when it actually is combining with your ace to have some extra value. That said, that king would almost certainly be more valuable in another suit, so her considering it “wasted” isn’t necessarily wrong. Splintering with singleton aces is much better than with singleton kings and queens. Jacks are less problematic, and splintering with a singleton jack is usually just fine.
Are Splinters Alertable?
Yes, all splinters are alertable. But some aren’t alertable until after the auction. ACBL’s alert rules are that bids at the 4‑level and above don’t get alerted during the auction – they require a “delayed alert,” which means you wait until the auction is over and, assuming your side is declaring, explain the bid before the opening lead. The reason for this is that alerts at this level are more likely to help the alerting side (“Oh good, partner remembered that” or “She alerted that? I wonder what she thinks it means”) than the opponents. However, there is one exception to this rule: on the first round of the bidding (i.e., until opener has taken a second call) there are no delayed alerts – all alertable bids are alerted normally.
This means that all of responder’s first-round splinters are alertable, because they happen during the first round of the auction. (eg. 1♠‑4♦). Opener’s splinters are only alertable at the 3‑level. (So you would alert 1♣-1♥-3♠, but would make a delayed alert for 1♦-1♠-4♣.) Splinters in GF auctions are alertable at the 3‑level (eg. 1♠-2♦-3♥) but not at the 4‑level (eg 1♠-2♥-4♣). I wish it weren’t this confusing, but these are the rules. When in doubt, err on the side of over-alerting rather than under-alerting.
How Do I tell the Difference between a Singleton and a Void?
Initially, you can’t. As the auction progresses, you might have the chance to clarify. If the splinterer later control-bids her suit, it shows a void (not a singleton ace). If you have agreements about how to show a void when partner asks for keycards, you can do that. Singletons are so much more common than voids that when partner splinters you should assume it is a singleton until proven otherwise.
Is 1♠-4♥ a Splinter?
Yes, but this is an important thing to discuss with your partner. Some people feel very strongly that anyone’s first bid of 4M should be to play. For me, since we already have a way to bid that hand – start with 2♥ and then bid 4♥ – the splinter is the more important bid. As always, a sound partnership agreement is more important than having the “right” agreement.
Can a Passed Hand Splinter?
It’s very rare that a passed hand would want to force to game, especially since partner’s opening bid in 3rd or 4th seat could be a little light. Usually you’re better off using Drury and not hanging partner if she’s a little light. However, these bids don’t have any other meaning, so while it’s very rare, I would assume that P-1♠-4♣ was a splinter. I’d expect a splinter by a passed hand to usually be a void rather than a singleton; that’s the best reason I can think of for a passed hand to force to game.
What Is a “Mini-Splinter”?
A Mini-Splinter is a splinter that is not forcing to game. The only time these exist in our system is when opener makes a Jump-Reverse – a single jump in a suit higher ranking than the one she opened (i.e., a suit in which she could have reversed). For example:
Opener |
Responder |
|
|
1♣ |
1♥
|
3♦ |
|
2♦ would be a reverse, so we don’t need 3♦ as a natural bid. (Opener has one forcing way to bid each of the unbid suits – a reverse or a jump-shift – but doesn’t get two; one is enough.) We define this bid as a Mini-Splinter: splinter shape (i.e., 4-card support, singleton or void in the splinter suit) but not necessarily GF values. An example hand for opener in this auction would be ♠ K65 ♥ AQ76 ♦ 3 ♣ AQ873.
Responder can sign off in 3♥ if she does not want to go to game; anything else is GF. Opener can have a stronger hand and continue on after the signoff.
What would 4♦ mean instead? It’s a GF splinter. It’s stronger than the Mini-Splinter, since it forces us to game. Some play it as showing a void, which seems like a good agreement to me.