Thorough Analysis of the GeoCities Roulette Archive

1. Executive Assessment

The GeoCities roulette archive is substantially more interesting as a historical collection of gambling-system literature than as a reliable source of profitable roulette strategies.

The fundamental distinction throughout this analysis is between describing probability and demonstrating a profitable betting edge.

A strategy can accurately calculate probabilities, identify streaks, organize historical data, and prescribe sophisticated bet sequences while still losing money in expectation.

2. What the Website Actually Is

The roulette page functions primarily as a directory or inventory of roulette systems. It presents a large collection of downloadable documents covering many different approaches to roulette.

The collection includes systems involving:

• Number selection.
• Repeat numbers.
• Hot and cold numbers.
• Dozens and columns.
• Even-money bets.
• Betting progressions.
• Martingale-type systems.
• Wheel sectors.
• Dealer and croupier behavior.
• Streaks and cycles.
• Probability calculations.
• Online roulette.
• Software-assisted strategies.

The result is less a coherent book than a catalogue of different gambling philosophies from different periods.

3. Historical Context

The current GeoCities.ws service should not be confused with the original Yahoo! GeoCities service. The original GeoCities platform was shut down by Yahoo in 2009. The present service describes itself as an archive intended to preserve recovered GeoCities material.

This matters because the material should generally be interpreted as archival content, not as a modern editorial publication in which every claim has been independently verified.

Consequently, the archive has considerable historical value even where individual systems have weak statistical support.

4. Major Strengths of the Archive

5. The Major Credibility Problem

The archive places extremely different kinds of claims beside one another without providing a standardized evidence hierarchy.

Some titles make exceptionally strong claims, using phrases such as “consistent profit,” “ultimate,” “greatest,” or even “infallible.” Such language is marketing language rather than statistical evidence.

The most important methodological rule is therefore:

A claim that a betting system wins consistently is not evidence that it wins consistently.

The evidence must come from independently reproducible results, clearly defined rules, adequate sample sizes and appropriate statistical controls.

6. The Mathematical Foundation: Roulette Has a House Edge

A standard single-zero European roulette wheel has 37 possible outcomes:

A straight-up number bet pays 35:1 while having a probability of approximately 1/37 of winning.

For a one-unit wager, the expected value is:

Thus, on a fair European wheel, a player betting one unit on a standard straight-up number loses approximately 2.70 cents per dollar wagered in long-run expectation.

American double-zero roulette has a substantially higher house edge, approximately 5.26% on standard bets.

Special rules such as La Partage or En Prison can reduce the house edge on certain even-money bets, but they do not normally turn standard roulette into a positive-expectation game.

7. The Ion Saliu Material

The Ion Saliu material is among the more intellectually interesting material in the collection because it attempts to formulate roulette systems using explicit probability concepts.

The documents discuss probability, recurrence, streaks, skips, sample sizes, simulations and what Saliu calls a “Fundamental Formula of Gambling.”

Why the mathematics is interesting

There is nothing inherently wrong with studying the probability of recurrence or the frequency of events across a sequence of roulette spins. Those are legitimate statistical questions.

Where the reasoning becomes problematic

The critical step is moving from a statement such as:

“An event becomes increasingly likely to occur within a sufficiently large number of trials.”

to a conclusion such as:

“Therefore a particular betting strategy based on that event has positive expected value.”

The second statement does not automatically follow from the first.

This distinction is central to evaluating Saliu's material and, more broadly, nearly all probability-based roulette systems.

8. Martingale and Betting Progressions

Several roulette systems rely on progressive betting. The classic Martingale is the simplest example.

A player might wager:

After each loss, the next bet doubles. A win is intended to recover the previous losses and produce a one-unit profit.

The attraction is obvious: most short sequences end in a win before the required bet becomes enormous.

The mathematical problem is the size of the exposure after a losing streak.

The progression therefore changes the distribution of outcomes, not the underlying expectation.

A strategy may produce many small winning sessions followed by an occasional devastating loss. A high percentage of winning sessions can therefore coexist with negative long-run profitability.

9. Streaks and Trend Systems

Documents such as Trend Tracker attempt to exploit streaks or trends in roulette outcomes.

Streaks are real statistical phenomena. A sequence such as:

is perfectly compatible with a random process.

The problem is the assumption that the existence of the previous streak necessarily makes the next outcome more likely to continue the streak.

On a fair wheel, assuming independent spins:

The previous sequence does not change that probability.

Therefore:

subject to the same standard assumptions as any other spin.

A trend system can only exploit streak information if the underlying roulette process contains a genuine dependency or physical effect.

10. Hot Numbers and Cold Numbers

Hot and cold numbers are another recurring theme.

It is perfectly legitimate to calculate which numbers have appeared most often during a historical sample.

For example, after 1,000 spins one can calculate:

• Frequency of every number.
• Average gap between appearances.
• Longest observed gap.
• Longest streak.
• Sector frequencies.
• Repeat frequencies.
• Clustering statistics.

These are descriptive statistics.

The difficult step is claiming that the historical frequency changes the probability of the next independent spin.

A number appearing unusually frequently in a finite sample does not, by itself, demonstrate that it has become intrinsically more likely.

11. Wheel Bias: The Important Exception

Wheel bias is fundamentally different from Martingale, hot-number or streak systems.

If a physical roulette wheel is genuinely biased, then the probability distribution may no longer be uniform.

Suppose, hypothetically, that a number has an actual probability of 4% rather than the theoretical 2.70%.

In such a situation, a sufficiently informed player could theoretically possess a positive expected-value opportunity.

However, demonstrating this requires much more than observing that a number has appeared frequently.

Evidence required for a credible wheel-bias claim

• A large sample of spins.
• Statistical significance.
• Controls against random fluctuation.
• Controls against multiple testing.
• Evidence that the bias persists over time.
• Evidence that the effect survives changes in dealer or conditions.
• Preferably, a plausible mechanical explanation.
• Out-of-sample confirmation.

This category is therefore theoretically legitimate but empirically demanding.

12. The Validation Problem

The greatest weakness of the collection is the lack of the kind of independent statistical validation that would be required to establish profitability.

A serious evaluation of a roulette system should include:

13. Copyright and Provenance

An important secondary issue is that some of the archived documents explicitly contain copyright notices.

For example, the 119-page Consistent Profit Roulette manual identifies Chuck Sutton as its developer and states that the material is copyrighted and may not be reproduced without permission.

The archive's existence therefore should not automatically be interpreted as evidence that every document is freely redistributable.

14. Classification of the Roulette Systems

15. Case Study: Consistent Profit Roulette

Consistent Profit Roulette is one of the most revealing documents in the collection.

The manual claims that users can win approximately 95–98% of their games and discusses very large hourly and multi-week profit figures. It describes a system involving wheel sections, waiting for specified numbers of misses, a “Coast Mode,” and an “Attack Mode.”

The document explicitly instructs players to divide the wheel into sections and wait for a specified sequence of misses before betting on the qualifying section. It then introduces different betting levels depending on subsequent outcomes.

The manual even presents a progression from small units toward increasingly large unit sizes, accompanied by projected profits.

These claims are extraordinary and therefore require extraordinary empirical evidence.

The critical issue

A reported winning percentage does not establish positive expected value.

Consider a hypothetical strategy:

This hypothetical example demonstrates why percentage of winning sessions and expected profitability are different measurements.

A proper evaluation of CPR would therefore need the complete wager history, every bet size, every loss, every win, maximum exposure, maximum drawdown and an independently replicated test.

16. What Would Convince Me That a System Works?

A rigorous research project could take one or more systems from the archive and convert them into formally testable algorithms.

Step 1 — Formalize the rules

Every instruction should be expressed as an exact algorithm. Subjective phrases such as “strong trend” or “good indication” should be eliminated or formally quantified.

Step 2 — Define the roulette environment

Specify:

• European or American roulette.
• Single or double zero.
• Payout structure.
• La Partage or En Prison rules.
• Table limits.
• Bankroll.

Step 3 — Run large simulations

Generate millions of independent spins using a well-defined random process.

Step 4 — Measure the complete economics

• Total amount wagered.
• Net profit or loss.
• Return on amount wagered.
• Maximum drawdown.
• Maximum stake.
• Bankroll requirement.
• Probability of ruin.
• Length of losing streaks.

Step 5 — Use out-of-sample data

The most important protection against overfitting is to test the final system on data that played no role in its development.

Step 6 — Repeat the experiment

A genuine statistical effect should not depend on one lucky simulation or one convenient starting point.

Step 7 — Test physical hypotheses separately

If a system claims to exploit wheel bias, its testing procedure should specifically measure physical-wheel behavior rather than merely treating random historical frequencies as evidence of bias.

17. Overall Assessment

18. Final Conclusion

The GeoCities roulette archive should be regarded primarily as a historical and analytical resource rather than as a verified manual for beating roulette.

Its greatest strength is its breadth. It preserves an unusually rich collection of roulette-system literature covering probability, progressions, pattern recognition, wheel sectors, dealer behavior, simulations and commercial gambling systems.

Its greatest weakness is that the material does not consistently distinguish between:

• mathematical facts;
• statistical observations;
• hypotheses;
• marketing claims;
• anecdotal results;
• independently demonstrated positive expected value.

That distinction is essential.

A fair roulette wheel has a mathematical house edge. Betting progressions, stop-win rules, hot-number systems and streak-following systems do not automatically eliminate that edge.

A genuine advantage is theoretically possible if the player possesses information about a real deviation from the assumed random process—for example, a persistent physical wheel bias—or if the payoff environment itself provides a positive expected value.

But those possibilities must be demonstrated empirically. They cannot simply be inferred from long manuals, elaborate charts, high historical win percentages or claims of consistent profitability.

In other words:

The most scientifically interesting next step is not to choose the “best-looking” system from the archive. It is to formalize several of the systems, simulate them under precisely defined roulette conditions, and compare their long-run distributions against the theoretical house edge.

19. Suggested Research Program

A particularly useful follow-up study would select five to ten of the archive's most prominent systems and subject all of them to the same experimental framework.

Such a study would be substantially more informative than simply comparing the authors' claimed winning percentages.

20. Sources Examined

• GeoCities.ws roulette archive — primary collection/index examined for the structure and scope of the roulette material.
• Consistent Profit Roulette — 119-page manual by Chuck Sutton, including its stated methodology, profitability claims and betting procedures.
• Trend Tracker Roulette — short roulette-system document centered on trend and streak concepts.
• Ion Saliu roulette material — probability-oriented roulette-system literature preserved in the archive.
• GeoCities.ws archive/terms information — relevant to the historical and provenance context of the material.

This document is an analytical review, not financial or gambling advice. The existence of a strategy in an archive does not constitute evidence that the strategy is profitable, safe, or suitable for real-money play.