There’s an incredible amount of nonsense written about roulette computers, so it’s difficult to know the truth. Most is written by manipulative roulette computer sellers, or online casino promoters who don’t know or care about accurate information. To know the truth, you can visit us and see a demonstration on any wheel, or test the free roulette computer explained here. It costs nothing and you can test from home.
The trial version provided here is about the equivalent to our Lite version computer. Our other versions are far more capable, and achieve earlier and more accurate predictions.
How To Use The Free Trial Computer
STEP 1
Visit www.roulettewheelanalysis.com/login and enter your login details with a WINDOWS PC (it may not work on other devices). You’ll then see a screen like below. If the website isn’t loading, let us know as the server may need rebooting.
The login details for the trial are:
Username: simple
Password: physics
Explanation of each part:
The Video: This is video of the wheel to test on. You can add the video of any wheel you want, but we add it for you on request. See instructions to request a different wheel.
Wheel profile: Here you can select the wheel to test on. The text area below the “Wheel profile” menu contains important information about the wheel. Whenever you select a new wheel profile, the “jump chart” is cleared, and the bankroll is reset to $2,000.
The betting table (click winning number): This is NOT where you place bets. Despite it being a betting table, it is only used to select the winning number after the ball lands. The simulated bets are displayed as “predicted numbers”, and they are displayed automatically for you.
Predicted numbers: These are the “tuned” numbers that you would actually bet on.
Raw prediction: This is the “UNTUNED” prediction. You only need to observe it if you are verifying the trial computer is not cheating somehow. We display it because you can expect manipulative computer sellers to claim the trial computer is somehow cheating.
Bankroll: This is your simulated bankroll. It will increase after wins, and decrease after losses.
Jump chart: This shows how far the ball landed from the “raw prediction”. A bar at a value of 0 means the ball landed right on the raw prediction. A value of 1 means the ball landed 1 pocket clockwise of the raw prediction, and so on. The idea is if the roulette computer is beating the wheel, specific areas of the chart will have groupings of high bars. These groupings are called “peaks”.
Mouse click area: For the computer to know the speed of the wheel and ball, you will click your mouse in this area when the green zero and ball are at a specific diamond on the wheel (the metal defector). See the wheel profile details so you know which diamond to use. If you use the wrong diamond, you will get poor results.
Instructions: This area tells you what to do at each stage in the prediction process, so you know what the computer is waiting for.
STEP 2
How to get predictions
1. Select the wheel profile to test with. The video will appear. But before playing the video, carefully read the information about the wheel. You must use the specified reference diamond to get accurate predictions.
2. When you are ready, play the video and place your mouse cursor in the “mouse click area”.
3. When the green zero is at the “reference diamond”, click the left mouse button once. Now wait for the wheel to complete one full revolution, and click once more when the green zero is at the same point. This completes the “timing” for the wheel (called the “rotor”).
4. If the ball is very fast (800ms/0.8s per revolution or less), wait for the ball to slow to a reasonable speed (about 0.8 seconds per revolution), then click in the mouse area each time the ball is at the same reference diamond. You will do FOUR ball clicks.
5. Now the computer will display the predictions. These are the numbers you have simulated betting on. We assume there is a $25 bet per predicted number. 6. When the ball lands, PAUSE the video (don’t fully stop it), then click the betting table to enter the winning number. You will see the bankroll change depending on whether you won or lost. You’ll also see the “jump chart” add an extra value.
Evaluating Accuracy
Firstly, understand that no computer can possibly be 100% accurate. But to have an edge over the casino, computers only need to achieve slight accuracy. Remember, the casino’s normal house edge over players is a relatively small -2.70%, but a computer can achieve an edge of +120% or more. This is an edge around 40 times greater than what casinos have over normal players. To evaluate accuracy, you could simply note your start and end bankroll. But proper testing needs much more as explained below:
It is normal that the “jump chart” has multiple peaks. These are multiple areas with groupings of high bars. These occur mostly because of different diamonds the ball hits.
Consider how clear the “peaks” are. The more noticeable they are, the more accurate predictions are.
Testing only a few spins is not proof of anything. In my public demos, I demonstrate around 60 spins for one direction, then repeat the test with another 60 spins. This is to demonstrate that the jump chart peaks are in the same location. If the results were “random”, then the peaks would NOT be in the same location.
If you know what to look for, you can test the accuracy of the computer’s basic functions by observing the raw prediction, and the wheel’s orientation when the ball falls.
Tips For Testing
- It takes the average player about 1 hour of practice to have reasonably accurate button clicks for timings. Your accuracy improves with practice, so don’t expect any accuracy until you are reasonably skilled. Do not expect immediate positive results without practice.
- Think in the longer term. Do not judge from 10 or so spins. You realistically need at least 60 spins for a reasonable test. Remember you have a long-term edge, but short term losses can and still do occur.
- Learn how to properly fast forward and rewind videos: On some videos you test just one wheel direction, and skip the other direction. Use the left and right arrow keys to move through the video quicker. You need to left click on the video with your mouse before using the arrow keys.
- To test how accurate your clicking is, try predicting the same spin repeatedly but when the ball and wheel are approximately the same speed. You should most predictions within the same 5 or so numbers. Otherwise: you need more practice, or the rotor is faster than normal, or you are starting clicking when the ball is too fast.
- The faster the wheel (the “rotor”), the less accurate your wheel timing will be. If the rotor is faster than 3 seconds per revolution, you can expect small clicking errors to produce significant errors. If your rotor timings are bad, predictions wont be accurate. Again it usually takes an hour of practice before your clicks are accurate enough.
- The faster the ball is when you make clicks, the less accurate predictions will be. In most cases, you should not start clicking for the ball until it is completing full revolutions at about 800ms per revolution (0.8 seconds).
- Don’t consider a test to be conclusive if you have too few spins. If you had just 30 spins, sometimes peaks look clear because short-term data isn’t evenly distributed. But further testing may show the peaks are co-incidental. More data is always better. I recommend 60+ spins per direction. If possible, 120 or so spins so you can do two separate groups of 60 spins and compare the location of peaks. This is what I do in private demonstrations so the proof is more conclusive. Usually you should ignore the charts until you have at least 60 spins saved.
- Adjust page zoom in your Internet browser to make the video larger: This is usually done by holding CRTL on your keyboard and using your mouse wheel to zoom in/out.
- Wait for the wheel to complete at least 3/4 of a revolution before you start clicking. This is because on wheels like the automated Slingshot, the wheel not have reached it’s target speed (it may still be speeding up when you take clicks)
- Although the free trial computer is basic compared to my other computer computers, it is capable of rejecting spins where the ball is predicted to fall less predictable. This feature maximizes accuracy, although it’s only enabled for some wheel profiles. If a spin is rejected, you’ll see a message indicating why.
- Some of the videos aren’t good quality and tend to skip (especially online casino recordings). This greatly degrades accuracy. I can add additional spins upon request.
Frequently Asked Questions
Q. How are your other computers different to the trial computer?
The free trial computer is about the equivalent of my Lite version computer. My more powerful versions such as the Uber and Hybrid are very different. The main differences with the Uber and Hybrid versions:
The Uber and Hybrid can get predictions 2-3 seconds after ball release. However, the free trial computer gets prediction only after 1 full wheel revolution (2 wheel clicks) AND 4 clicks for the ball. This is a total of 6 clicks and it may take 6 or more seconds to get a prediction. This is early enough for about 20% of casinos. But the Hybrid and Uber can get predictions early enough for about 95% of casinos. You can see a full comparison at the roulette computer comparison page.
The Uber and Hybrid versions eliminate significant “timing errors”: The Hybrid uses automated image recognition hardware to automatically measure wheel and ball speed, so there is virtually no significant error. And the Uber uses a variety of features to reduce errors, such as allowing multiple players taking timings at the same time.
The Uber and Hybrid are much more accurate: The free trial computer may only break even on some wheels, when the Hybrid or Uber may achieve an edge of +50% or more. This is achieved by much more advanced algorithms, such as calculating ball bounce based on which diamonds the ball is expected to hit.
Q. Can I test the computer on any wheel I want?
Yes for private demonstrations. But no for this particular trial computer: you can test only on the wheels available in the profile selection area. Every wheel has a different ball and rotor deceleration rate, and the trial computer only has profiles with pre-configured deceleration rates for specific wheels. Of course the full versions allow you to test on any wheel, but the evaluation version is only for evaluating. However, you can contact us to add new wheels to test on.
Q. How do I know you haven’t somehow made the computer accurate only for the pre-configured wheels?
The computer doesn’t know which spin you are testing on. It just processes the timings from your clicks and calculates a prediction. So you can easily run any test to prove the predictions are from a genuine algorithm, and that the results aren’t “rigged”. Perhaps randomly choose a few particular spins and test on them repeatedly. If your timings are reasonably accurate, you will get predictions in much the same area on the wheel for specific spins. So you can be sure about this, load the video outside the software so there’s no possible way to software knows what spin you are testing on. But play a different video inside the software interface. This allows you to start the prediction process with one spin playing in the software, but test on another in a completely different window.
Q. What is the algorithm being used?
It is one of our basic algorithms, and is almost identical to what’s used by our “Lite” version computer. The algorithm is simply the mathematics that calculates the predictions.
Q. If I purchased a computer from another vendor, can I test the evaluation computer on spins they provided?
YES. But you need to send me the video so I can enter the data for the wheel’s ball and rotor deceleration rates. Also I suggest not telling the vendor you are doing a comparison test, because it tends to upset them. By offering the free computer, it is not my intention to create friction. I simply want you to be able to easily know facts from fiction, without needing to rely on anyone’s word alone.
Q. In a real casino, would I need to bet on all predicted numbers?
NO, you simply bet on as many of them as you can. Usually betting only 1 or 2 of the numbers is sufficient. Generally the more numbers you bet, the quicker you’ll profit, and the less time you spend in the casino. Betting more numbers is beneficial because you bet only 1 or 2 numbers, you may have a lot of “near misses”. Betting more numbers means this is much less likely to happen.
Q. When do dealers normally call “no more bets” in casinos?
It varies between casinos, dealers and times of day (depending on how busy tables are). Usually you have about 8 or so seconds to get predictions and bet, then the ball falls about 10 seconds later. Casinos allow this because it means more bets on the table, and more revenue for them. Only in rare cases is “no more bets” called very early. It’s different with online casinos which now almost always close bets before the ball is released, because they were heavily attacked.
Q. Can the free trial computer give earlier predictions?
The free trial computer needs 2 clicks for the wheel, and 4 clicks for the ball before giving prediction. It is a comparatively simplistic version just for you to test. It is not capable of earlier predictions, unlike our other versions.
Q. Can the free trial computer predict spins in both clockwise and anticlockwise directions?
Yes but if you test a video with both directions, then there is a different profile for each direction. You cannot apply the same profile on both directions. Our full version computers all have the ability to predict spins of either direction, which is done by selecting the wheel direction during the prediction process.
Q. Can I use the free trial computer on any device like ipads?
Currently you can only use it on a windows pc, assuming it has a reasonably fast processor. You are unlikely to have accurate clicks on any touch pad pc. You should only use a mouse to make clicks.
Q. Why is there a slight delay to get predictions after the last click?
The timings are measured on your PC, and then sent to the server for calculation. The server then sends back the predictions, which is the cause of the delay. It is done this way to prevent reverse engineering of the software.
Which Wheels Can You Test On?
Every roulette wheel is different, so the settings for one wheel won’t work for another. So you can only test the computer on wheels in the “wheel profile” menu (in the software interface). However, you can request us to add video for any wheel you want. Contact us to request addition of a wheel, but please first see the requirements below.
Requirements for new wheels:
- 60+ spins per direction. If there are too few spins, the test results won’t prove anything. 60 spins per direction should be considered the minimum. But if you really want to be sure, I suggest around 120 spins per direction so you can compare results from the first and second 60 spins. You can test on as few as 30 spins but then results are more likely to be luck.
- Ideally the wheel should spin in only ONE direction: Any of our full version computers work with any direction, but the free trial computer is designed for one direction only. This is to simplify testing. We can upload videos with alternating wheel directions, but you should apply the computer on one direction at a time. So if you want to test both directions, first test all spins for one direction, and then test all for the other direction. You must clear the charts and reset the bankroll before testing different directions (just re-select the same profile).
- Good clear video that shows the whole spin including the winning number: We don’t want to use poor quality videos. And remember that other people using the free trial computer will see it too.
- At least 20 frames per second: video is jittery if less than 20 frames per second (FPS). It is still possible to test with 15 FPS video, but it should be at least 20 FPS to ensure smooth video and a clearly visible ball.
- Consecutive spins from the same wheel and same day: The ball will decelerate at a different rate on different days. This is because of variables like air pressure. The advanced computers easily adjust, but the trial version does not. So the spins you use should all be done consecutively and on the same day. The wheel must also be the same for all spins.
- Real dealer wheels: You can still often beat automated wheels, but usually results are best on typical wheels with real dealers. A good place to start is by screen recording spins from live online roulette casinos. You can also request videos of other roulette computers being demonstrated (from other roulette computer vendors). But we suggest NOT telling other vendors about it, as it tends to annoy them. However, it enables you to compare the accuracy of their computers, with the accuracy of a basic computer.