Initial Rating Explained in Plain Language

When a completely unrated player enters their first official chess tournament, the Elo system has a unique problem to solve: How do you assign a starting rating to someone when the mathematical formula explicitly requires a pre-existing number to function?

If a player has no rating, the standard Expected Score equation simply cannot be run. Instead, federations use entirely separate mathematical procedures to grant an Initial Rating (often called a provisional rating). While normal rating updates involve step-by-step additions to an existing number after every single game, the initial rating process works by evaluating an entire block of games at once.

If you are trying to estimate where you or a student will land on the next rating list, you have to understand how this block-evaluation actually works.

The Core Ingredients of an Initial Rating

To generate an initial rating, the system doesn’t look at individual games in isolation. It needs three aggregate numbers from your first event (or series of events):

1. Number of Games (N): The total number of games played against opponents who already have published ratings. Unrated vs. Unrated games are mathematically meaningless and are entirely ignored by the system.
2. Scored Points (W): Your total score against those published opponents (Win = 1, Draw = 0.5, Loss = 0).
3. Average Opponent Rating (Rc): The mathematical mean of the ratings of all your rated opponents.

Once the system has these three numbers, it determines your initial baseline. If you scored exactly 50% (e.g., 2.5/5), your initial rating will simply be the exact Average Opponent Rating (Rc). Earning an even score proves you belong exactly at the mathematical center of that specific rating pool.

However, if you score above or below 50%, the system has to apply an adjustment.

The FIDE-Style Adjustment Table

If your score deviates from 50%, standard federations like FIDE apply a mathematical adjustment based heavily on a specific fractional table. This table translates your win percentage into a precise point value (known as dp), which is then added to or subtracted from your Average Opponent Rating (Rc).

For example:

• If you score 70% against a 1600 average field, the table dictates that a 70% performance translates to a +149 point adjustment. Your initial rating will be published at 1749.
• If you score 20% against a 1600 average field, the table applies a -240 point penalty. Your initial rating will be published at 1360.

This process is highly structured, but recently, FIDE added a controversial modifier to stabilize the system against extreme outliers.

The “Two Hypothetical Drawn Games” Rule

In March 2024, FIDE introduced a new element to their initial rating formula to prevent unrated players from receiving wildly inflated numbers after one lucky event.

The system now surgically attaches two hypothetical drawn games against 1800-rated opponents to your actual tournament performance before running the math.

If you play 5 games and score a perfect 5.0 against an average rating of 1400, the system does not calculate an initial rating based on a 5.0/5 performance. It calculates your rating based on a 6.0/7 performance against an adjusted average rating. Because those two phantom 1800-rated draws are baked into the equation, your final published rating is pulled closer to the center, protecting the wider rating pool from inflation.

(Note: If calculating initial ratings manually feels too abstract, our tool automatically implements the FIDE 1800-draw adjustment. You can run the numbers directly in our Initial Rating Calculator.)

Publication Thresholds: Why You Might Not Let a Rating Yet

Even if you successfully calculate your initial performance, it does not mean you will appear on the next rating list.

Federations require a minimum threshold of data before they publish a number. Under FIDE rules, you must complete at least 5 games against rated opponents to qualify. If you only played 4 rated opponents at a weekend swiss, your results are held in the database until you play another rated opponent in a future event.

Furthermore, federations implement a hard rating floor for initial publication. For FIDE, the minimum possible published rating is 1400. If you complete 5 games but the mathematical output places you at a 1320 performance, the federation simply will not publish the rating. Your results remain unrated until you play another event and your aggregate performance climbs above the 1400 threshold.