The science of scouting in FRC

FRC scouting is the disciplined study of how robots perform, not merely how many points they score. Each match offers evidence about speed, scoring routes, autonomous reliability, defence and failure recovery. Used well, that evidence helps an alliance choose partners, prepare game plans and decide what to practise.

A strong scouting program combines match records with video, observer notes and knowledge of the current game manual. It separates what happened from why it probably happened, so an alliance does not build strategy around a lucky cycle or a single misleading performance. The aim is better prediction, not more paperwork.

The science lies in repeatable observation. Definitions stay consistent, measurements use the same units, unusual results are flagged, and confidence is recorded. This reduces bias while preserving uncertainty, giving captains a clearer basis for selection than instinct alone.

Australian teams face real logistical constraints. Events are commonly centred on cities such as Melbourne, Sydney, Brisbane, Adelaide and Perth, and interstate travel can dominate a competition weekend. Scouting routines must therefore work with compressed schedules, limited rehearsal time and consistent formats across very different venues.

Turn match observation into comparable evidence

Start with a small set of comparable observations rather than unstructured impressions. The same fields should be used for every robot, and the game manual should determine what counts as a score, cycle or penalty.

Metric What it reveals Useful record Strategic use
Qualification points Overall match output Points by phase Baseline ranking
Scoring cycle Field efficiency Start, finish and elapsed time Match tempo
Autonomous consistency Repeatability Points or tasks and attempts Opening advantage
Fouls and failures Risk under pressure Count, cause and recovery Reliability
Defence Pressure on opponents Duration and legal impact Matchup fit

These measures answer different questions. Qualification points describe outcomes, while cycle time exposes efficiency. Autonomous consistency can be more valuable than a high ceiling if a robot completes the same task every match. Compare like with like, and never combine totals from different game seasons.

Context prevents misleading comparisons. A robot may score fewer points because it protects a lead, lacks access to a preferred field location or has been assigned a defensive role. Record match circumstances before assigning a label such as fast, elite or inconsistent.

Build a consistent scouting system

Use one shared scouting sheet for every observer, with separate columns for raw observations, interpretations and confidence. Record the match number, alliance, starting position and notable events. Define ambiguous terms in advance, particularly successful cycle, missed score, defence and reset, so two volunteers are not effectively measuring different things.

Before a competition, train observers by having them watch the same match independently. Compare their notes, resolve disagreements and agree on examples. Short calibration sessions improve consistency and prevent fatigue, shifting attention or the excitement of a close match from distorting the evidence.

Data handling and workshop safety both have an Australian legal context. The Privacy Act 1988 and Australian Privacy Principles can apply when scouting records contain personal information, while state and school requirements may add further conditions. Safe Work Australia guidance and relevant state work health and safety duties also matter when teams modify battery packs, guards or electrical systems.

Build the routine around Australian school terms, public and school holidays, and the weekend commitments many students retain for cricket or Australian rules football. Regional teams may have fewer practice hours than city programs, making a simple, rehearsed form more useful than a sophisticated app nobody can complete.

Measure scoring, mobility and reliability

Measure a scoring cycle from a defined start and finish, such as possession of a game piece to the completed scoring action. Record failed attempts separately from successful cycles. Comparing different scoring actions without normalising them can create a false impression of speed, especially when the current game includes multiple ways to score.

Mobility notes should describe more than top speed. Acceleration, stability over uneven carpet, turning precision, defensive control and the predictability of a robot’s path all affect alliance performance. Consistent motion can be more valuable than greater speed that repeatedly causes missed positioning or collisions. Track whether the same failures recur across matches.

Reliability data also influence purchasing decisions. The Australian robotics parts market may include local distributors, while many drive and control components arrive through international supply chains. Recording repeated failure modes helps teams order the right spares before leaving home, with Australian-dollar budgets allowing for GST, freight, currency movements and long lead times.

Read patterns across qualification matches

A single match is evidence, not a verdict. Look for patterns across several performances: median scoring rate, variation in cycle time, the percentage of autonomous attempts completed, and whether failures cluster in particular situations. Confidence should increase with repeated evidence, while a sudden strategy change or unusual field condition should be flagged rather than treated as a permanent weakness.

Schedule also affects what a team appears to be doing. Early qualification matches may reveal a robot still under development, while later matches can reflect deliberate changes. Scout as many qualification matches as practical, prioritise likely alliance partners and then revisit close decisions using the full record. Opposition analysis is strongest when it combines results with the match conditions surrounding them.

Communication under pressure is difficult to score but matters during alliance selection. Observable signs include concise callouts, calm recovery, willingness to share information and consistent support for teammates. The ideas in predict team leadership can complement a scouting process by encouraging attention to sustained behaviour and team contribution rather than one confident moment.

Match robots to alliance needs

Qualification ranking provides a baseline, but the strongest alliance is not always the one containing the highest-ranked robots. Consider complementary capabilities, shared cycle times, starting-location conflicts, autonomous routines, defensive styles and reliability. A slightly lower-ranked robot that consistently completes its assigned role may create a better matchup than a higher-ranked machine whose success is unpredictable.

Build alliance configurations around outcomes the robots can produce together. For example, one robot may create scoring opportunities while another converts them efficiently, provided their routes and human-player responsibilities do not interfere. Account for both the likely opponent and the alliance’s internal rhythm, including the seconds lost when one robot resets or cycles in a different location to another.

Opponent-specific preparation should remain adaptable. Robots can change behaviour after a match, and defensive pressure may force errors that never appeared in earlier scouting. Use tendencies to form a starting hypothesis, then update it after direct observation. Defence must also be legal: preventing access without avoidable contact or fouls is more valuable than an aggressive approach that reverses the result through penalties.

Run a repeatable scouting workflow

A repeatable workflow turns scattered notes into useful decisions. Assign one person to maintain the shared record, rotate observers across alliances and hold a short debrief after each qualification match. For teams that also publish event updates, Maryland FIRST’s guide to a team social media strategy can help separate engaging storytelling from confidential performance analysis.

A practical scouting routine should include these actions:

At the next team meeting, appoint the data lead and two observers, choose the measurements that can change an alliance decision and schedule the first review. Make the following practice count by building two alliance configurations, rehearsing a response to the most likely opposition and documenting one specific action for every robot.