I still search for ways to play without losing to much, so i can collect data from the wheel.

This is one way i still test and are curios about.

I quote Snowman:

Some roulette facts:

Here's the real deal on past spins influencing future spins. The following information is based on a REAL ROULETTE WHEEL WITH A LIVE DEALER. On these wheels the rotor did NOT alternate spin direction- (every other spin like they do in EU).

The following information does NOT pertain to an RNG based wheel or the random game of roulette. In other words, no electronic machines.

I can tell you the facts, based on my experience, as they pertain to a real wheel. Here they are, like them or not.

1. If a number has hit within the last five spins, it really is slightly more likely to hit on the next spin.

2. Simply playing the last five numbers to have hit will slightly reduce the house edge, especially when the wheel speed is close to the speed of the previous spin. In rare cases, you can get a small edge on some wheels playing this way.

3. The dealer's pick up and release of the ball is not random, like someone on this board had stated. As a matter of fact, it will quickly spike out beyond five standard deviations in a short period of time for most dealers, if you take the time to measure it.

Here's a small example below.

30000 trials. (This is just one wheel. Each wheel and sample varies some. Overall, the effect is real.)(Wheel make is a Huxley Mark Series with moveable fret ring). (Date 2007 to 2009)(Right handed)

Betting the last seven numbers to have hit.

Edge reduces to -1.29

Max loss run 37

Max DD 4871

- - -

Betting the last five numbers to have hit.

Edge reduces to -1.79

Max loss run 52

Max DD 4209

- - -

Betting the last three numbers to have hit.

Edge .02

Max loss run 85

Max DD 1555

- - -

Betting the last number only.

Edge 5.17

Max loss run 85

Max DD 1555