Research
A study of 25,380 deals: control bids that show any control, against control bids that show only aces and voids, and two ways of mixing them.
A control bid shows a control in the suit bid on the way to a slam. There are two ways to play it. In the first, a control bid shows any control: an ace, a king, a singleton or a void. Bridge World Standard plays it this way, and Dubois’ Bidding Notes describes it as Italian control bidding. In the second, the standard method in Dubois’ Bidding Notes, the first control bid in a suit shows an ace or a void; a king or a singleton can be shown only by bidding again a suit partner has already shown. In both, bidding past a suit denies a control in it.
BidMaestro changed to showing any control in July 2026, on the authority of those sources. This is the first time the choice has been measured.
The app dealt 1.2 million deals at random and bid them on three settings: the default, every convention switched on, and one club player’s own settings. On 25,380 of them a control bid could be made on at least one setting. Each deal that the two methods bid to different contracts was played 200 times, with the opponents’ cards redealt each time and all four hands visible.
| Settings | Deals bid differently | Aces only better / worse | Points a deal, aces only against any control |
|---|---|---|---|
| Default | 2,629 | 1,516 / 1,032 | 0 ± 8 |
| Every convention on | 2,323 | 1,302 / 944 | −12 ± 8 |
| One player’s settings | 2,880 | 1,498 / 1,299 | −19 ± 7 |
On the default settings there is no difference. On the other two, showing any control is slightly ahead. Showing only aces does better on more of the deals, mostly by stopping in game where showing any control went on to the five level and failed. Showing any control bids more slams, and a slam that makes is worth far more: on the default settings it bid a slam on 937 of these deals where aces-only stopped lower, and the reverse happened on 358.
One rule matters a great deal in the aces-only method. Bidding past a suit denies an ace in it, and with two small cards in that suit the partnership should not ask for aces (Dubois’ Bidding Notes makes the same point: with two small cards in an unbid suit, the hand should not normally bid 4NT). A first version of the method left this rule out and lost by 24 to 41 points a deal. The figures above include it.
Grouping the deals by the first bid where the two methods part (default settings):
That split suggests a mix: one hand shows only aces and voids, the other any control. Two ways of choosing the hands were tried. In the first, the first player on the side to bid shows any control, and partner only aces and voids. In the second, the first player on the side to bid the trump suit shows any control, and partner only aces and voids. They differ mainly after opener’s splinter: after 1♦–1♠–4♣ responder bid spades first, so under the second rule responder shows any control and the splinter bidder only aces and voids.
Because the mix was chosen by looking at these deals, both versions were also tried on 25,201 new deals from a fresh 1.2 million, which they had not seen.
| Points a deal, against showing any control | Same deals, default | New deals, default | New deals, every convention on | New deals, player’s settings |
|---|---|---|---|---|
| First bidder shows any control | +46 ± 9 | +60 ± 9 | +53 ± 10 | +38 ± 9 |
| First to bid the trump suit shows any control | +71 ± 9 | +76 ± 9 | +72 ± 10 | +53 ± 9 |
Each figure is an average over the deals where the mix bid differently from showing any control: between 1,296 and 1,663 deals, depending on the version and the settings. Both versions gain, on the new deals as well. The gain comes from the hand that shows only aces and voids when it continues the control bidding, or makes the first control bid after Jacoby 2NT: the mix stops in game where a king or singleton shown there carried the partnership to a five-level contract that failed.
The trump-suit version is the better of the two. Against the first-bidder version it gains 112 ± 30 points a deal on the 162 new deals where the two bid differently (default settings), and 120 ± 32 on 146 on the player’s settings. Almost all of that comes after opener’s splinter, where responder showing any control finds slams.
BidMaestro shows any control, an ace, king, singleton or void, with either hand.
← Back to research