Streets of Nine (Roulette): What the Progression Actually Does — and Doesn’t

At a glance

Strategy profile

Difficulty: Beginner
Bankroll demand: High
Volatility: Medium

The method places equal units on nine streets (27 numbers) while skipping three streets determined by recent spins, and uses a simple negative progression (increase unit by +1 after a loss); it’s operationally simple but creates rising exposure and significant drawdown risk.

Watch the original video on YouTube →

If you’ve seen a streamer or a chart saying “cover nine streets, skip three, up one on a loss,” that’s the Streets of Nine idea in a nutshell: cover 27 of 37 (or 27 of 38) numbers with street bets, and increase the per‑street unit by one after a loss. It’s straightforward to run at a live table, and because you cover most of the wheel it delivers frequent small wins — but the math underneath is immutable. This article breaks down the probabilities, the payouts, the expected value, and the real bankroll exposure so you can decide whether the trade‑offs make sense for you.

How the bet is structured

A “street” is a 3‑number inside bet; on most wheels a winning street pays 11:1 (you get 11 units profit plus your street stake back).
Streets of Nine: at baseline you place one unit on nine different streets (9 streets × 3 numbers = 27 numbers covered). You skip three streets determined by a rule (for example, the three streets associated with recent spins).
Progression rule in this version: if you lose, increase the per‑street unit by +1; if you win, reset (or drop back) to baseline. The streamer example used $10 as one unit, so an initial spin is 9 × $10 = $90 in action.

Probabilities, payouts, and expected value (the immutable math)

Let u be the stake per street. On any spin (European wheel, single zero):

Probability one of your covered streets hits: p = 27/37 ≈ 0.72973.
Probability of a miss (none of your streets hit): q = 10/37 ≈ 0.27027.

If one of your streets hits, you receive 11u profit on that street and lose the other 8u you had staked — net profit = 11u − 8u = 3u. If none of your streets hit, you lose 9u.

So expected value per spin = p*(+3u) + q*(−9u) = u*(3p − 9q). Plugging the European numbers: EV = u*(30.72973 − 90.27027) = u*(2.18919 − 2.43243) = −0.24324u.

Interpretation:

With u = $10, EV ≈ −$2.43 per spin.
As a percentage of the total money at risk that spin (9u), that’s −$2.43/($90) ≈ −2.7027% — exactly the wheel’s house edge for a European wheel.

On an American double‑zero wheel the same calculation gives p = 27/38, q = 11/38 and a house edge of ≈5.263% — the progression doesn’t alter the wheel’s edge.

Bottom line: covering 27 numbers increases win frequency and reduces the size of each winning pocket, but the expected loss per dollar wagered equals the wheel’s house edge. No staking plan can change that.

What the +1 after loss progression actually does (and a caution)

The progression increases your stake after every loss. That raises two things:

Your volatility (larger bets after a losing run), and
Your potential drawdown (your cumulative loss before recovery could be large).

A crucial and often-missed point: a single win after a losing run does not necessarily recover prior losses. If you lose k spins in a row, increasing units by 1 each time, and then win at the next level, the net result is:

Total net (in units of u) = 3(k+1) − 9 * [k(k+1)/2]

That formula comes from 3u for the final winning spin minus the sum of the preceding losing stakes (9u × (1+2+…+k)). Numerical examples (u = $10):

k = 0 (win first spin): net = +3u = +$30.
k = 1 (lose, then win): net = −3u = −$30.
k = 2 (two losses, then win): net = −18u = −$180.

So after one or more consecutive losses, a single subsequent win can leave you further in the hole. You need multiple wins at the higher unit to recover — and that is exactly where the risk lies.

Bankroll sizing and loss probabilities — real numbers

Because the miss probability q is ≈0.2703 (European), the chance of k straight misses is q^k. That falls quickly but not negligibly:

3 straight misses: q^3 ≈ 0.0197 (about 2%).
5 straight misses: q^5 ≈ 0.00145 (≈0.145%).

What does a 5‑loss run cost with u = $10? Losses per spin: 9u, 18u, 27u, 36u, 45u ⇒ cumulative lost = 9u*(1+2+3+4+5)/5? — compute directly: 90+180+270+360+450 = $1,350 lost before the 6th spin. The next bet (if following the +1 rule) would be 54u = 54 × $10 = $540. These are not hypothetical tiny numbers — they can blow past a casual bankroll quickly.

If you want to run this progression, you must be explicit about how many consecutive losses you can tolerate and size u so that your stop‑loss is acceptable. That’s bankroll management — it helps you control ruin risk — but it doesn’t change the underlying expected loss rate per dollar wagered.

Tink’s Toolchest can help you compare the likely loss magnitude for different unit sizes and calamity lengths so you can make an informed bankroll decision before you sit down.

Practical takeaways and alternatives

The Streets of Nine gives frequent small wins (net +3u on a hit) because you cover 27 numbers, but the house edge remains unchanged: approximately 2.70% on European wheels and ~5.26% on American wheels.
The +1 after loss progression increases the chance of a big drawdown. Don’t confuse longer streaks being unlikely with them being impossible — the payouts of the progression don’t guarantee recovery after a single win.
If you like the frequent action and small per‑win reward, consider using a flat stake (no progression) sized to your comfort level; you’ll get the same long‑run EV per dollar but with lower spike risk.
If you prefer to manage downside, explicitly set a cap on the number of increases you’ll accept and a clear session stop‑loss. That’s bankroll discipline, not a way to beat the house.

Responsible play reality check

This is a clear example of a system that’s easy to run at the table but still subject to the wheel’s math. Progressions change your variance and the shape of your wins and losses, not the house edge. If you enjoy the method as entertainment, set unit size and loss limits first, use Tink’s Toolchest to model worst‑case runs, and never wager money you can’t afford to lose.

Video: Streets of 9 Strategy - Session 2 #roulette by Compulsive Tinkerer.

← Back to the strategy lab