Everything you do, in one number.

Twenty-two raw points at a stacked Signature event should count for more than twenty-two at a quiet local one. This is where RFORCE weighs them. Every event earns a difficulty rating, and your raw points are scaled by it.

One event gave 1900H an Event RFORCE of 42.52. But a season is not one event. Your RFORCE, the number that ranks you, is the average of every Event RFORCE you have earned this season.

That is the whole calculation. There is no adjustment applied between the average and the headline number, no confidence penalty subtracted, and no separate internal figure the rankings are secretly built from. Add up your own Event RFORCEs, divide by how many events you played, and you get your published RFORCE back.

Every event you compete at joins that average. Early in the season the number swings around, because a single event is a large share of it. As events pile up it settles into a steady read on your performance, and the evidence window published beside it narrows to say so.

Before your first event of the season you have no RFORCE at all. That is published as an absence, never as a zero: a zero would mean you competed and scored nothing.

After 1900H's first event their RFORCE was just that event: 38.00. A 46.00 at their second event pulled the average up to 42.00. This third event's 42.52 nudged it to 42.17. Three events in, 1900H stands at 42.17 Rpts per event.

Now the almost-simple part. Teams are ranked from highest RFORCE to lowest. But two teams can sit so close together that calling one better than the other by pure RFORCE wouldn't be enough. So RFORCE has a tie rule.

Sort every team by their RFORCE, highest first, and you have the leaderboard.

When two teams' RFORCEs fall within 2% of each other, RFORCE treats them as tied, because a gap that small is within the margin of a single lucky or unlucky event. The 2% is measured from the top team in the group, not from the row above. From there the tie breaks in order, by things a team can see and earn:

1. 1The narrower evidence window. The window is how much uncertainty a rating carries, and it narrows both with more events and with more consistent results. So this single step rewards the rating that rests on better evidence, whichever way that evidence was earned.
2. 2More total Award Rpts across the season. This rewards the judged side of competition, the notebooks and design interviews those awards stand for.
3. 3The higher exact RFORCE, with the 2% grace removed. Two teams matching to full precision is almost impossible, so this all but always settles it.

Each step only matters when the one before it is still level.

If two teams are still equal after that last step, they genuinely tie, and they share the rank. RFORCE does not invent a further tiebreaker to separate them.

Look at 1900H and 6200K. 6200K's 42.80is higher than 1900H's 42.17, yet 1900H ranks above them. The two sit within 2% of each other, so they are tied. 1900H's evidence window is ±2.63 against 6200K's ±4.58, so 1900H's rating is the better established of the two and takes the higher rank. 6200K has played twice, and those two results were far apart; 1900H has played three times, closer together. Both of those facts are already inside the window, which is why there is no separate step for either.

Had their windows also matched, RFORCE would look next at their total Award Rpts, and then at their exact ratings. And notice what does nothappen: neither team's rating is reduced by its window. 6200K is still a 42.80team. The window decided an order, not a score.

And that is the whole journey. A single result at a single event became raw points. Those points were weighed by the real difficulty of the event. The event settled into a season number. And that number set 1900H's place in the standings. Every team on RFORCE travels the same four moves.

From here, the same engine branches out: into individual driver, programmer, builder, and coach ratings, and into rankings for whole organizations and countries. Those come next.

Everything so far has produced one number for a whole team. But a team is people, and RFORCE rates them too. Using the same engine, the RANK system and the award hierarchy, adjusted by event difficulty, RFORCE builds an individual rating for each role a person plays. Each role is scored from the parts of the team's performance that role actually drives.

On VAIRC, which runs fully autonomous with no human drivers, there is no driver role.

The RFORCE Notebook Score, covered further down, is deliberately not part of any of these. Role ratings stay reproducible from public VEX data, which is what lets them compare cleanly across seasons.

On a competitor's profile, their role rating is shown as a radar across the dimensions that define it, so you can see at a glance where a driver, programmer, or builder is strongest.

Each role board, driver, programmer, and builder, is ranked per platform by that role's RFORCE, and the same tie rule applies. When two competitors' role RFORCEs fall within 2% of each other, RFORCE treats them as tied. From there the tie breaks in order:

1. 1The narrower evidence window. For a role, that window comes from the spread of that role's own event-level results, so it narrows both with more events and with steadier ones. It REPLACES the old event-count step rather than joining it: the count is already inside the window.
2. 2More total Award Rpts that count for that role. Drivers count Excellence and Amaze, programmers add Think, and builders use the full judged set, the same split that builds the role rating in the first place.
3. 3The higher exact role RFORCE, with the 2% grace removed. Two competitors matching to full precision is almost impossible, so this all but always settles it.

Each step only matters when the one before it is still level.

If two competitors are still equal after that last step, they genuinely tie, and they share the rank. RFORCE does not invent a further tiebreaker to separate them.

Two people on the same team in the same role share every measure they earn, the same events, the same awards, the same skills rankings. So they will often land on exactly the same number and tie outright. That is expected, and it is correct: they did the same season together, and RFORCE will not invent a difference between them that the results do not show.

Zoom out one more level, from the people to the institutions they compete under.

RFORCE also ranks whole organizations and whole countries. The idea is simple: add up the RFORCEs of every team they field.

Cumulative RFORCE = the sum of every member team's RFORCE

Average RFORCE per team = cumulative RFORCE / number of member teams

These totals are kept separate for each platform. RFORCE measures every platform against its own season — its own typical field size and its own typical result — so a V5RC rating and a VIQRC rating sit on slightly different scales, and adding them into one sum would combine things that are not quite the same. An organization that competes in three platforms has three separate ranks, one per platform. A profile can still gather all of a school's teams into one view, but the ranked totals stay per platform.

1900H's 43.69 would drop into its organization's V5RC cumulative total, added alongside every sibling team the organization fields on that platform.

Organizations and countries are ranked by cumulative RFORCE, per platform, and the same tie rule applies. When two groups' cumulative RFORCEs fall within 2% of each other, RFORCE treats them as tied. From there the tie breaks in order:

1. 1The higher RFORCE per team. Cumulative RFORCE rewards breadth, so this step rewards the group that reached the same total with fewer teams. If those per-team averages are themselves within 2%, the two are still tied and the cascade continues.
2. 2The narrower evidence window on the cumulative total. A group's total is a sum, so its uncertainty is built up from its member teams' own windows. It REPLACES the old event-count step rather than joining it: the count is already inside each member's window.
3. 3The higher exact cumulative RFORCE, with the 2% grace removed. Two groups matching to full precision is almost impossible, so this all but always settles it.

Each step only matters when the one before it is still level.

If two groups are still equal after that last step, they genuinely tie, and they share the rank. RFORCE does not invent a further tiebreaker to separate them.

Two features sit alongside all of this: a notebook score that RFORCE assigns by hand, and Teamwork Rpts, a measure of Teamwork Challenge performance that RFORCE computes from its own data. Both are next.

This one is different from everything above. The RFORCE Notebook Score is a quality score for a team's engineering notebook, out of 100, assigned by RFORCE against its own rubric. It does not come from the VEX API, and it is not derived from any competition result. It is read and scored directly from a team's notebook.

VEX announces which teams win judged awards, but it never releases the scores behind them or the reasoning, so most teams have no idea how their notebook was actually evaluated or how to improve it. The Notebook Score exists to fill exactly that gap: concrete feedback on notebook quality, win or lose.

The rubric is its own instrument, not a copy of VEX's judging. It is built to give useful, original feedback rather than to guess at how VEX scores.

Eight categories add up to 100 points. Each criterion within them is scored on a five-level descriptive scale, from Exemplary down to Not Evident, so scoring stays consistent from one evaluator to the next. The final score maps to a named recognition level, shown on the team's profile. A notebook scoring 84, for example, earns an RFORCE Silver Notebook.

One more number appears on RFORCE profiles, and this one RFORCE computes itself.

How a team performs in the Teamwork Challenge, built from results RFORCE already has.

At a single event, RFORCE adds together three numbers already earned for that team: how it did in qualification, how it did in the finals bracket, and how it did in the grand final round, when the event has one. That sum is the team's teamwork Rpts for the event.

Teamwork Rpts (one event) = qualification + finals + grand final

Each of those three pieces caps at ten points, so a team's teamwork Rpts for any single event can never rise above thirty. That ceiling is not something a season happens to produce. It follows directly from how the three pieces are built: cap three numbers at ten apiece, and their sum cannot exceed thirty no matter how the event runs.

Two versions of the number appear on a team's profile, and they answer different questions. Add teamwork Rpts across every event a team competed in this season, and you get a total: how much the team has done. Divide that total by the number of events, and you get an average: how well the team does when it competes. A team that ran three events and a team that ran twenty cannot be compared fairly on the total alone, so the average is what stands next to another team's average. The total stays on the profile too, because a full season of work earns the right to be seen rather than folded into an average that hides it.

Teamwork Rpts applies to every platform RFORCE covers, head-to-head or cooperative. Qualification standing and finals placement are earned the same way whether or not there is an opposing alliance on the field, so there is no format this metric sits out.

Every term, symbol, and code name used across the RFORCE Index, in one place. Search for any of them.

RFORCE calculates at full precision and rounds once, at the moment a number is shown. Scores and evidence windows are shown to two decimals; coefficients to four, because a coefficient near 1 carries its information in the third and fourth. A hand check built from displayed figures reproduces a published one to the last published digit.

RFORCE also does not claim to be the only measure that matters. Other respected metrics each capture their own part of the picture, and RFORCE is meant to sit alongside them as one comprehensive lens, not the last word.

Questions about any of this? Every term is defined in the glossary above, and the RFORCE team is one message away on the Contact page.