We have been busy, and a few things are finally ready to show. The headline: BridgeDummy can now tell you, with actual numbers, which hands justify which bid. Beyond that, the template library keeps growing, bidding with a real partner instead of a robot is on the way, and the mobile app is close enough to release that we are looking for Android testers. Details on all of it below.
Hand Insights: What the Simulations Actually Say
If you have used our simulation templates, you know the drill: pick an auction, generate deals that match it, see how your choice fares. Hand Insights is what happens when we run that process at scale and keep notes.
For a template's decision point, the research engine deals hundreds of different hands you could hold and plays each one across 50 simulated layouts. Then it digs through the results looking for what actually separates the winners: does point count matter here? Shape? The number of aces and kings? When a pattern holds up across enough hands, it goes into the report as a plain sentence a bridge player would say, not a statistics table. Every claim comes with its sample size, so you can see how much weight it deserves.
Each report also includes an explorer where you set the conditions yourself. Curious how balanced minimums did? Hands with both majors? Pick the conditions and it shows you the record for hands like that, next to the base rate for everyone else.
One Question, 500 Hands: The Quantitative 4NT
You open 1NT, partner invites slam with a quantitative 4NT. Pass or bid 6NT?
Every teacher gives the same advice here: pass with a minimum, bid slam with a maximum. Fine, but where exactly is the line?
We put 500 different 1NT openings through the engine, 50 deals each. Taken as a whole it is nearly a coin flip: passing was the winning action on 51% of hands, 6NT on 49%, and 24 hands were too close to call either way. The interesting part is what separates the two groups:
| Your hand | Winner | Record |
|---|---|---|
| 16+ HCP with a five-card suit | 6NT | 63 of 68 hands (93%) |
| 16+ HCP with 6+ controls | 6NT | 86 of 95 hands (91%) |
| 16+ HCP, no other requirement | 6NT | 166 of 207 hands (80%) |
| 15 or fewer HCP with 4-3-3-3 shape | Pass | 53 of 55 hands (96%) |
| 3 or fewer controls | Pass | 24 of 25 hands (96%) |
(Controls are aces and kings, weighted: an ace counts 2, a king counts 1.)
Two things jump out. A five-card suit is worth nearly as much as the extra point, because it is a source of tricks that 4-3-3-3 simply does not have: 16 points alone wins 80% of the time, but 16 points with a five-card suit wins 93%. And on the pass side, the flattest hands are the clearest passes of all, even before you count the points.
This table is a taste of one report. The full 4NT report is live right now: every pattern the engine found, confidence bounds on each one, and real example hands from both groups. You will need a premium account to open it.
TipTen templates have published reports so far. Look for the gem icon on a template's row in the Templates library on the Simulations page: gem means a report exists. More are on the way.
The Template Library Passed 150 Scenarios
The simulation library just passed 150 scenarios, 45 of them added since the start of June. It is not only bidding decisions anymore: the newest batch leans heavily on opening leads after competitive auctions, the messy real-life kind. A few from this week:
- Your lead after 1♥-2♠-X-4♠-Pass-Pass-5♥: they pushed on over your side's 4♠, now find the killing lead.
- Your lead after Pass-Pass-2♣-2♠-X-Pass-3NT: they landed in 3NT with your overcall. Do you lead your suit?
- Pass, 2♣ or 2♦ after 1♦-Pass-1NT-Pass: the quiet rebid problem every minor-suit opener knows.
If there is an auction your partnership keeps mangling, odds are decent a template for it already exists, and if not, tell us and we may build it.
Coming Soon: Bidding With Your Actual Partner
Robots are patient, but they are not your partner. Partnership bidding puts the two of you at the same table: you each see only your own hand and bid a shared set of boards whenever you have a spare minute. Nobody has to be online at the same time, and every board has its own chat, so when the auction lands somewhere strange you can argue about it right there, next to the evidence.
We built this because it is the practice we always wanted with our own partners: real sequences, real misunderstandings, no scheduling. It is in the final stretch of testing now - watch this space.
The Mobile App Is Almost Here (Android Testers Wanted)
The BridgeDummy app has been in the works for a while, and it is finally close: puzzles, the daily challenge, and your stats on your phone, built as a proper app rather than a browser tab.
Before it goes public we want real players banging on it. If you have an Android phone and do not mind the occasional rough edge, we would love to have you on the testers list: you get the app before everyone else, we get your bug reports and complaints. Both sides win.
To join, send us a note through the contact page: just say you want in on the Android test and we will add you to the list.
TipAndroid tester signup: one message through the contact page. Say you want in, we handle the rest.