01 · The Tool All Levels
📊 What Is VRC Data Analysis?
VRC Data Analysis (vrc-data-analysis.com ↗) is a free website that gives every VRC team a score — called a TrueSkill rating — based on their match results. The higher the score, the stronger their track record.
It's like a leaderboard, but smarter. Beating a strong team raises your score more than beating a weak team. That makes the ratings more meaningful than a simple win/loss count.
🏆
TrueSkill Rating
A score that reflects how strong a team is based on who they've beaten.
📈
Match History
A full list of every match result so you can spot patterns.
🌍
Rankings
Sort teams by state, region, or worldwide to see where teams stand.
📋
Event Pages
See every team registered for your tournament and their ratings before you arrive.
📌 Most important thing to know — for every level.
This tool shows you history. It cannot tell you whether a robot is working right now, or how a driver performs today. Always use it alongside your own eyes.
team page URL format
https://vrc-data-analysis.com/team/[TEAM NUMBER]
Example — Team 2473B:
https://vrc-data-analysis.com/team/2473B
📱 Works on your phone. No login needed.
Open this at the event. Pull up any team number during matches. It's public data.
02 · First Move All Levels
🔑 Start With the Event ID
VRC Data Analysis tracks every registered VRC team worldwide. Before you start looking up individual teams, you need to narrow that giant database down to only the teams that are actually at your event this weekend.
The way to do that is the RobotEvents event ID — a number that identifies your specific tournament. Paste it into VRC Data Analysis and it shows you a filtered list of only those teams. That filtered list becomes the foundation for everything else on this page.
✅ This is the first move — for every level.
Shadows use it to get 2 teams to observe. Rookies use it to build a watchlist. Veterans use it to draft the pre-event brief. All three start from the same event page.
⚡ Why the Event ID First?
❌ Without Event ID
- You're looking at hundreds of global teams
- Random lookups waste time on absent teams
- Easy to miss strong teams actually attending
- Beginners don't know where to start
- No shared starting list for the whole team
✅ With Event ID
- Only the ~30–60 teams at your event
- Everyone starts from the same list
- Shadows, Rookies, and Veterans can divide teams
- No time wasted on teams who aren't there
- Pre-event brief builds itself from one page
📋 How to Find Your Event ID
The event ID comes from RobotEvents.com — the official VEX competition system. Every tournament has a unique ID number in its URL.
RobotEvents event URL pattern
https://www.robotevents.com/robot-competitions/vex-robotics-competition/RE-VRC-XX-XXXXX.html
The event ID is the part at the end — like RE-VRC-25-1234
Your coach or EP will have this, or find it in the event confirmation email.
🗺 How to Open the Event Page in VRC Data Analysis
1
Get the RobotEvents event ID for your tournament
Find it in the event URL on robotevents.com, in your confirmation email, or ask your coach. It looks like
RE-VRC-25-XXXXX.
Before event day
2
Open the event page in VRC Data Analysis
Navigate to
vrc-data-analysis.com/event/[EVENT-ID] — or use the search/event lookup on the site. You'll see every registered team for that tournament.
Night before or morning of
3
Screenshot or note the full team list
This is your master list. Every scouting decision this weekend comes from here — not from random team lookups.
~2 min
4
Sort by TrueSkill
The event page lets you sort by rating. A quick scan from top to bottom shows you the strongest teams, the unknowns, and the teams to watch.
~1 min
5
Assign teams to your scouting list
From this filtered event list, assign specific teams to each scout. Shadows get 1–2. Rookies get 3–5. Veterans take the full top tier. Now everyone has a focused job.
~5 min as a team
event page URL format
https://vrc-data-analysis.com/event/[ROBOTEVENTS-EVENT-ID]
Example placeholder (not a real event):
https://vrc-data-analysis.com/event/RE-VRC-25-1234
🖼️
Screenshot Placeholder: VRC Data Analysis event page — sorted team list
All teams registered for one tournament, sorted by TrueSkill. This is the starting list for all three tracks.
vrc-data-analysis.com/event/RE-VRC-25-XXXX
Event · Push Back Regional · 42 Teams Registered · Sorted by TrueSkill
9999AIron Eagles31.40.7↑ +2.1
7777XApex Robotics29.80.9↑ +0.8
2473BSpartan Design27.21.1↑ +1.4
5555CCircuit Breakers24.04.2— 0.0
3311ZDelta Force21.61.3↓ −1.2
8800BNova Build18.35.1— 0.0
↑Top 3 (bold): Likely alliance captains. Research these in depth.
↑High σ (4+): 5555C and 8800B are unknowns — few events, unreliable rating.
↑↑ Trend: 2473B trending up +1.4 — improving team, may be stronger than rank suggests.
↑↓ Trend: 3311Z trending down — investigate why before considering as a pick.
📍 What Each Level Does With the Event Page
👁
Shadow: A coach or veteran picks 1–2 teams from the event list and assigns them to you. Your job is to look up just those teams and notice 3 things about each. Don't search randomly — start from the list.
🌱
Rookie: Open the event page, pick out the top 5–8 teams by TrueSkill, and build a small watchlist. Then open each team's individual page for more detail. The event page is the filter — team pages are the detail view.
⭐
Veteran: The event page is your pre-event brief starting point. Scan the full list. Identify likely alliance captains (top 8), flag high-σ unknowns, note upward-trending teams. This is step one of the 20-minute prep routine.
03 · Entry Point
🗺 Choose Your Track
This guide has three paths. Pick the one that matches your role right now. You can always come back and read a more advanced section later.
👁 Shadow Track
New to the team
You're observing, learning the vocabulary, and starting to notice what matters. You're not responsible for decisions yet.
Focus on: 3 things to notice, green/caution signs, writing one observation
Skip for now: μ, σ, alliance selection
Start here →
🌱 Rookie Track
Starting to scout
You're helping with real scouting and match prep. You can read a team page and write useful notes, but want a clear structure.
Focus on: Reading a page, comparing teams, before-match routine
Skip for now: Deep alliance strategy, STEM analysis
Start here →
⭐ Veteran Track
Running strategy
You're leading event prep, building pick lists, and making real alliance decisions. You need the full tool.
Focus on: Full team-page reading, trend analysis, alliance decision tree
All sections relevant.
Start here →
Shadow
🎯
Your job: start from the event list, get your 1–2 assigned teams, look them up, notice 3 things, write one sentence. That's it.
📋 Step 0 — Start From the Event List, Not a Random Team
Before you look up any team, your coach or veteran should open the event page on VRC Data Analysis and assign you 1–2 specific team numbers from it. Don't search randomly. Starting from the event list means you're researching teams who are actually at your event.
Wait for your assignment.
Ask your coach or veteran: "Which teams should I look up?" They'll give you 1–2 numbers from the event list. Then follow the steps below for each one.
🔎 Step 1 — Open the Team Page
Once you have a team number, go to vrc-data-analysis.com/team/[team number].
Practice example: Team 2473B
Go to vrc-data-analysis.com/team/2473B. You'll see their name, a rating number, and a list of match results. Don't worry about all the numbers yet — just look at the page.
👀 Step 2 — The 3 Things to Notice
When you look at a team page, focus on only these three things. Ignore everything else for now.
1
Their score (the big number)
A higher number means the system thinks they're a stronger team. Don't memorize the exact number — just notice if it's high, medium, or low compared to other teams you look up.
2
How many events they've been to
More events = more experience, and a more reliable score. A team that's only gone to one event is harder to judge than one with five events of data.
3
Is the score going up or down?
Scroll down to their recent results. Are they winning lately or losing? A team that's been winning more recently is probably stronger now than they were earlier in the season.
vrc-data-analysis.com/team/2473B
Team
2473B · Spartan Design
TrueSkill Rating
27.2
↑Shadow — notice this first: Is this number higher or lower than other teams you've looked up? Higher = stronger track record.
Events This Season
4 events
↑Shadow — notice this second: 4 events = good amount of data. A team with 1 event is much harder to judge.
Recent Trend
↑ Improving
↑Shadow — notice this third: Going up = they've been winning more lately. Going down = worth watching more carefully.
🚦 Step 3 — Green Signs and Warning Signs
After looking at those 3 things, ask yourself: does this team look reliable or questionable?
✅ Looks Reliable
- Been to 3 or more events
- Score is high on the list
- Winning recently
- Consistent results (mostly wins)
⚠️ Look More Carefully
- Only been to 1 event
- Score is low or unclear
- Losing recently
- Up-and-down results
📝 Step 4 — Write One Sentence
After looking at the page, write one sentence about what you saw. You don't need to be precise — just record what you noticed.
One-sentence format for Shadows:
"Team [number] — their score is [high/medium/low], they've been to [X] events, and they've been [winning/losing/mixed] lately."
Example · Team 2473B
What would a Shadow write?
"Team 2473B — their rating looks solid, they've been to 4 events, and they seem to be improving over the season. I'll watch their next match to see if that matches what I see."
👁 Step 5 — Compare What the Screen Says to What You See
This is the most important habit for any level. After looking up a team, go watch them compete. Then ask yourself:
- →Does the robot look as strong as the score suggested?
- →Is it running cleanly, or are there visible problems?
- →Does the driver look in control?
If something doesn't match — trust what you see.
A high score on the website doesn't mean the robot is working well today. If a top-rated team looks broken in their match, that matters more than any number.
⏭
That's your whole Shadow workflow. When you're ready to go deeper — reading the specific numbers, comparing teams, or preparing for scouting duties — move to the Rookie track.
🎯
Your job: open the event page first, build a small watchlist from it, then research each team individually. Write notes your team can actually use.
📋 Step 0 — Open the Event Page and Build Your Watchlist
Before you look up any individual team, open the event page for your tournament in VRC Data Analysis. This filters the database down to only the teams at your event — the only ones that actually matter this weekend.
1
Get the event ID from your coach or the RobotEvents page
It looks like
RE-VRC-25-XXXXX. Your coach will have it, or find it in the tournament confirmation.
Night before
2
Open vrc-data-analysis.com/event/[EVENT-ID]
You'll see every registered team for your tournament sorted by TrueSkill. This is your starting list — not the global rankings.
~1 min
3
Pick your watchlist — top 5 to 8 teams
From the event list, write down the 5–8 highest-rated teams. These are the ones worth researching individually. Ignore the rest for now.
~2 min
4
Also flag any teams with very few events attended
These are unknowns — their rating might not be reliable. Add them to a separate "watch closely" list.
~1 min
5
Now open each watchlist team's individual page
Use the 3-minute research workflow below for each one. The event page is your filter — individual pages are your detail view.
3 min per team
✅ This keeps you focused on what matters.
"This is the fastest way to turn a giant database into a useful scouting list. You're not wasting time on teams who aren't even at your event."
🔢 Reading the Numbers — In Plain Language
Once you have your watchlist from the event page, open each team's individual page. Here's what the three main numbers mean:
μ (mu) — the skill estimate
Simple: The system's best guess at how skilled this team is. Higher = stronger team.
More exact: It's the center of a probability range — the most likely value for this team's true skill level, based on their match history.
σ (sigma) — the confidence level
Simple: How sure the system is about that estimate. Low σ = confident. High σ = not enough data yet.
More exact: σ tells you how wide the uncertainty range is. A team with σ = 0.8 has a tight, reliable rating. σ = 4 means the system isn't certain yet.
Conservative rating — the leaderboard number
Simple: The number used to rank teams. It's lower than μ on purpose — a team has to prove themselves over many matches before ranking highly.
More exact: Calculated as μ − 3σ. This prevents teams from ranking high based on limited data.
Quick check: are μ and the conservative rating close together?
If they're close (within 3–5 points), the team has played enough matches that the data is reliable. If there's a big gap, the team is still building their track record — treat their rating with more caution.
📉 How to Read the Trend
The number isn't the whole story. Direction matters more. A team trending upward is building form. A team whose score dropped after their last event has something to investigate.
✅ Strong Signs
- Rating trending up over last 3 events
- σ is small — lots of match data
- Conservative rating close to μ
- 3+ events attended this season
⚠️ Be Careful
- Only 1–2 events attended
- High σ — not enough data yet
- Rating dropped after last event
- Very early in the season (Oct–Nov)
⏱ The 3-Minute Team Research Workflow
Use this before or during an event when you need a quick read on one team.
1
Open the team page
Go to vrc-data-analysis.com/team/[NUMBER]. Glance at μ (the skill estimate).
~20 sec
2
Check σ
Is σ below 2? Rating is reliable. Above 4? Treat it with caution — not enough match data yet.
~10 sec
3
Look at the trend
Scroll to recent results. Going up? Going down? Flat? Up = improving. Down = investigate.
~40 sec
4
Count events this season
More events = more reliable data and a more battle-tested robot.
~10 sec
5
Write one sentence in your notes
"Team X — rating [high/med/low], σ [small/large], [X] events, trending [up/down]. Watch their [autonomous/driver/end game]."
~40 sec
🔬 Comparing Two Teams
When you need to compare two teams quickly — for an upcoming match or a potential pick:
comparison example
Team A μ = 26.4 · σ = 1.1 · 4 events · trending ↑
Team B μ = 27.8 · σ = 3.8 · 1 event · trending —
Higher score: B
More reliable data: A (σ much lower, way more events)
If you had to pick one: A. More data = more trust.
The team with the higher score isn't always the better pick.
A lower rating built from many events can be more trustworthy than a higher rating from a single tournament. Always check σ and event count alongside the score.
🏁 Before a Qualification Match — Rookie Routine
Use this before each match you're assigned to watch or strategize for.
1
Look up both opponents
Pull up both teams on the site. Note who has the higher rating and more reliable data.
2
Identify the bigger threat
Which opponent is more likely to cause problems based on their score and trend? Focus your scouting attention on them.
3
Watch the match, then update your notes
Did the robot perform the way the data suggested? Write "confirmed" or "different than expected" and what you actually saw.
🔗 Combine This With Your Scouting Sheet
VRC Data Analysis gives you the starting picture. Your Match Scouting Sheet → captures what you see in real time. Use them together.
how the two tools work together
VRC Data Analysis → What history says about this team
Scouting sheet → What you see today at this event
Together → A complete, two-source picture
Example · Team 2473B
What should you still verify in person?
Even with a strong TrueSkill rating, you should still watch 2473B's matches to confirm: Is their autonomous running today? Is the robot healthy? A team that rebuilt their robot between events might be stronger or weaker than the number says. Always check.
⏭
When you're ready to build full pick lists, run 20-minute event prep sessions, and drive alliance selection decisions — move to the Veteran track.
🎯
Your job: start from the event ID, build a pre-event brief from the event page, identify alliance targets, and make confident picks when the board opens.
🔍 Reading a Team Page at Full Depth
🖼️
Screenshot Placeholder: vrc-data-analysis.com/team/2473B
Annotated: μ, σ, conservative rating, event count, trend arrow
| Number | What It Means | What You Want |
|---|
μ (mu)
skill estimate |
The system's best guess at this team's skill. Center of their probability distribution. |
High — indicates strong match history |
σ (sigma)
confidence |
Uncertainty range. High σ = few matches played, volatile estimate. Low σ = well-established rating. |
Low — tighter σ means higher confidence |
Conservative rating
leaderboard |
μ − 3σ. Used for rankings. Prevents over-crediting teams with limited data. |
High + close to μ = proven across many events |
Example · Team 2473B
What does the gap between μ and conservative rating tell you?
A team with μ = 28 and σ = 0.8 has a conservative rating of 25.6 — very close. The system is confident. A team with μ = 28 and σ = 4 has a conservative rating of 16 — the same midpoint, but wildly uncertain. Same headline number, completely different reliability. Always check σ.
📉 Trend Interpretation — Beyond Up/Down
When reading a trend, ask why, not just which direction.
- ✓Rising after attending tougher events: Real improvement. This team is proving themselves against better competition.
- ✓Rising after only weak opponents: Soft schedule. Verify their quality against stronger teams before trusting the trend.
- ✓Falling after a rough event: Could be a bad day, a mechanism failure, or a real skill gap being revealed. Pit scouting will tell you which.
- ✓Flat σ after many events: Rating has stabilized — the system is confident this is their real level. Treat this as a reliable benchmark.
- ✓Recent events count more. A strong event in October matters less than a strong event in February. Scroll to the bottom of their match list.
📅 The Night Before — 20-Minute Event Prep Routine
This routine starts with the event ID — not a random team search. Everything else builds from the event page you open in Step 1.
1
Get the RobotEvents event ID — open the event page
Navigate to vrc-data-analysis.com/event/[EVENT-ID]. This filters the full database to only teams at your tournament. Sort by TrueSkill. Everything this season prep session produces comes from this list — not from random team lookups.
2
Identify the top 8–10 teams — likely alliance captains
From the event-specific list, note who's ranked highest. These are your probable 1st–8th seeds. Build your initial pick list order from them.
3
Flag high-σ teams as unknowns
From the same event list, find teams with high σ — few events attended. They're wildcards: could be much stronger or weaker than listed. Mark them for mandatory live verification.
4
Identify upward-trending teams — sleepers
A team ranked 15th overall who improved +5 in the last month may outperform their rank today. Flag them in your brief.
5
Check your own team page
Know your own rating and regional standing before you walk in. Sets realistic expectations for seeding and who you'll likely face.
6
Write the pre-event brief
5–6 teams worth watching (from the event list). 3 potential alliance targets. Your own position. The event ID and date at the top so this brief is traceable. Bring it. Update it live.
Example · Pre-Event Brief Structure
What does a veteran bring to the event?
A one-page document with: Event ID + name + date at the top. Top 8 teams from the event page sorted by TrueSkill. 2–3 flagged unknowns (high σ). 1–2 sleepers (upward trend). Your own team's current rating. 3 first-pick candidates with a one-line justification each. This is built entirely from the event page — no random searching involved.
vrc-data-analysis.com/event/RE-VRC-25-XXXX · sorted by TrueSkill
9999AIron Eagles31.40.7↑ +2.1
7777XApex Robotics29.80.9↑ +0.8
2473BSpartan Design27.21.1↑ +1.4
5555CCircuit Breakers24.04.2— 0.0
3311ZDelta Force21.61.3↓ −1.2
8800BNova Build18.35.1— 0.0
⭐Veteran Step 1: Top 3 are your likely alliance captains. Open each individually for deep reading.
⚠️Flag these: 5555C σ=4.2, 8800B σ=5.1 — unknowns. Mark for mandatory live verification.
👁Sleeper: 2473B +1.4 trend — improving faster than their rank suggests. Watch closely.
📉Declining: 3311Z −1.2 — data says dropping. Investigate before considering as a pick.
🖨 Pre-Event Brief Template
Fill this out the night before. Hit Print Brief to get a paper copy for the event, or keep it on your phone. All fields are optional — fill in what you have.
🤝 Alliance Selection — Using TrueSkill to Build Your Pick List
1
Open the event page — sort by TrueSkill
If you ran the 20-minute prep the night before, you already have this. If not: open vrc-data-analysis.com/event/[EVENT-ID] now. This event-specific list is your draft pick list — do not use the global rankings, which include teams not at your event.
2
Filter out high-σ teams without live verification
Unless you've watched them today and they confirmed their rating, high-σ teams carry too much uncertainty for a high-stakes pick.
3
Layer in your live scouting
For each candidate, recall what you actually saw in quals. Robot health? Autonomous consistency? Move teams up or down based on today's evidence.
4
Find complementary strengths
Strong in driver control? Find a partner with a reliable autonomous. Don't pick a team identical to yours — fill the gap in your alliance's skill profile.
5
Talk to your top picks before the board opens
A warm pick is a confident pick. Walk to their pit. Introduce yourself. A team that knows you're interested is more likely to accept without hesitation.
⚡ Alliance Selection Decision Tree
Is this team's σ small? (Is the rating reliable?)
YES → Rating is trustworthy. Move to next question.
NO → Treat rating as uncertain. Requires direct live observation before considering.
Did you watch them compete at this event?
YES + performed well → High-confidence pick. Add to list.
YES + struggled → Investigate before picking. Data says strong; behavior says risky.
NO → Try to watch 1 more match before selection. Data alone isn't sufficient.
Do their strengths complement yours?
COMPLEMENTARY → Strong pick. You cover each other's weaknesses.
DUPLICATE → Lower priority. Look for a team that fills a gap in your skill profile.
❌ Common Veteran Mistakes
- ✗Ranking by μ alone, ignoring σ. A high rating with σ = 4 from one event is not equivalent to the same rating with σ = 0.9 from eight events.
- ✗Picking teams with identical strengths. Two offensive powerhouses don't always outperform one offensive + one reliable autonomous partner.
- ✗Over-trusting early-season data. October ratings don't reflect February robots. Always weight recent trend heavily.
- ✗Not talking to picks in advance. Cold selection during the board is riskier. A 5-minute pit conversation before eliminates hesitation.
- ✗Ignoring what you saw live. A historically strong team that broke down in three quals is a liability regardless of rating.
🗺 Pre-Match Plan Template
TrueSkill is most useful when it drives a specific decision, not just a general opinion. Before each elimination match:
pre-match plan format
"Our strongest opponent is [TEAM] (TrueSkill XX.X, σ X.X).
Based on scouting, their weakness is [OBSERVATION].
Our alliance plan is to [SPECIFIC STRATEGY]."
⚙ STEM Highlight · Decision Theory & Risk Management
Alliance selection is a multi-criteria decision problem under uncertainty. TrueSkill gives expected performance (μ), but σ is the risk attached to that estimate. In decision theory, minimizing worst-case outcomes over elimination rounds often beats maximizing best-case ceiling. A team with μ = 27, σ = 0.9 is a safer pick than μ = 29, σ = 4.5 when you need consistency across 3 matches, not a single spectacular run.
🎤 Interview line: "We use TrueSkill data as the foundation for alliance selection, but we specifically weight σ alongside μ — a high rating with low certainty gets less weight than a slightly lower rating with extensive match evidence behind it. We verify every data pick with live observation at the event."
🔬 Check for Understanding
You're the 2nd seed at alliance selection. Team A: μ = 30, σ = 4.1, one event attended. Team B: μ = 27, σ = 0.8, five events, strong quals day today. Which is the better first pick?
ATeam A — higher μ means higher ceiling
BTeam B — lower σ and live verification makes their rating far more trustworthy; you've confirmed their form today
CNeither — need more data before picking
DTeam A — always pick highest available TrueSkill
Shared All Levels
⚠️ What This Tool Cannot Tell You
This applies to every experience level. Knowing these limits makes you a better strategist.
✅ What it DOES tell you
- Historical match performance across the season
- How a team ranks relative to others in the region
- Whether a team has been consistently strong or inconsistent
- How their rating has changed over the season
- Which teams have the most battle-tested results
🚫 What it DOES NOT tell you
- Whether the robot is working today
- How the driver is performing this specific event
- What autonomous they're running this tournament
- Whether they've rebuilt since their last event
- Team health, morale, or coaching situation
⚠️ When data and observation conflict — trust what you see.
TrueSkill reflects the past. If a historically strong team is clearly struggling at this event, your live observation outweighs any historical number. Note it in your scouting sheet and adjust your pick list accordingly.
🤝 The Three-Source Rule
📊
VRC Data Analysis
Historical strength, trend, confidence level. Research the night before.
👁️
Live Match Watching
Current form, robot health, driver quality. Observe during quals.
🔧
Pit Scouting
Robot design, mechanism choices, team state. Visit during lunch.
Shared All Levels
📅 Event-Day Workflow
1
Evening before — open the event page via event ID
Go to
vrc-data-analysis.com/event/[EVENT-ID]. Screenshot the team list sorted by TrueSkill. This is your scoped starting point — not the global leaderboard.
~10 min
2
Arrival — cross-reference with your match schedule
Mark when you play high-rated teams. Identify windows during other matches to observe alliance targets.
~10 min before quals
3
Quals — watch more than you play
When not on field: watch, update your scouting sheet, cross-check TrueSkill data against what you see live.
Ongoing
4
Lunch — pit visits
Visit top 3 alliance candidates. 5-minute conversation each. You learn things no database shows.
~20 min
5
Final quals — finalize pick list
Ranked by: TrueSkill + what you saw + complementary fit. 6–8 options in priority order.
Last few rounds
6
Alliance selection — lead with confidence
You've done the work. You know who you want. Make the pick without hesitation.
Selection board
✅ Before You Leave — Quick Checklist
- ✓Have the RobotEvents event ID for your tournament
- ✓Opened the event page on vrc-data-analysis.com and pulled the team list
- ✓Know the top 8–10 teams by TrueSkill from that event list
- ✓Know your own team's current rating and regional rank
- ✓Flagged 2–3 high-σ teams from the event list to watch closely
- ✓Flagged 2–3 upward-trending teams from the event list (potential sleepers)
- ✓Team assignments made — each scout knows their 1–5 teams from the event list
- ✓Scouting sheet open and ready on your phone
For Coaches Vet / Mentor
👨🏫 For Coaches & Experienced Students
Teaching Each Level
- ✓Shadows: Give them one team to look up before an event. Ask: "What did you notice?" Don't correct — build the habit of observation first.
- ✓Rookies: After each event, have them compare what TrueSkill predicted vs what actually happened. The gap is the lesson.
- ✓Veterans: Require them to explain every pick in full: rating, σ, live observation, and why the strengths are complementary. "High TrueSkill" is not a justification.
📋 Assign From the Event Page — Don't Let Students Search Randomly
Before an event, open the event page yourself and divide the team list among your students. Give each Shadow 1–2 teams. Give each Rookie a watchlist of 4–6. Veterans take the full strategic tier.
This makes training structured and measurable.
When every student is researching from the same event-scoped list, you can debrief as a team. "What did you find about your 2 teams?" is a much better training question than "did you look anything up?" Students also can't spend time on teams who aren't even attending.
coach assignment model · night before
Open event page → sort by TrueSkill
Shadows → assign teams ranked 15–20 (enough data, not intimidating)
Rookies → assign teams ranked 5–15 (core scouting pool)
Veterans → take teams ranked 1–8 (likely alliance captains) + full list ownership
🗂 Match-Day Assignment Tracker
Paste team numbers from the event page below. Hit Assign and the tracker distributes them across your scouting team automatically. Drag any tag between buckets to reassign.
Avoiding Over-Reliance
Ask: "What would change your mind about this pick?"
If a student can't answer, they haven't thought it through. Push for a falsifiable answer: "If I saw their mechanism jam twice more, I'd move them down." That's strategic thinking — not data reading.
Turning Data into a Match Plan
coaching prompt before each match
"Who is our strongest opponent right now? [TEAM]
What does TrueSkill say about them? Rating, σ, trend.
What did you actually see in their last match? [OBSERVATION]
What does that mean for how we play today? [PLAN]"
If a student can answer all four lines using a combination of data and live observation, they're doing strategic preparation correctly.
Notebook Vet Level
📝 Notebook & Interview Connection
Using this tool as part of event preparation is a documentable process — exactly the kind of organized, data-informed decision-making judges want to see.
What to Write
- ✓Event ID + pre-event team list: Document the RobotEvents event ID you used, the date, and the list of teams you pulled from the event page. Include the top 5–6 teams with their TrueSkill ratings at the time of research. This makes your process traceable and shows methodical preparation.
- ✓Watchlist creation: Note how you narrowed the event list to your scouting focus. Which teams did you assign to each scout? How did you decide who to prioritize?
- ✓Post-event reflection: Compare what the pre-event team list predicted to what actually happened at the tournament. Where did TrueSkill align with results? Where did it miss? This analysis is exactly what judges look for.
- ✓Alliance selection reasoning: Connect your pick to both TrueSkill data from the event list and your live observations. Two sources = stronger documentation.
- ✓Season benchmarking: Track your own rating across events. Did an autonomous upgrade move your rating? Connect performance data to design decisions.
⚙ STEM Highlight · Data Analysis & Engineering Design Process
Using external performance data to benchmark your team, identify gaps, and design targeted improvements is a core EDP loop. TrueSkill gives you a quantitative external signal — it's not just a ranking, it's evidence that your robot improvements are translating into competitive results.
🎤 Interview line: "Before each competition, we get the RobotEvents event ID and open that specific event page in VRC Data Analysis, which applies a Bayesian TrueSkill model to rate teams by match performance quality. That gives us a filtered list of only the teams at our tournament — we don't search randomly. We combine that event-scoped list with match scouting at the event so our alliance selection is always built on two verified sources."
Shared All Levels
⚖️ Using Data Responsibly
- ✓Use it to prepare, not to dismiss. A low rating means "less likely to be a top threat" — not "ignore them." They might surprise you.
- ✓Rate robots, not people. TrueSkill reflects match results. Never use it to talk down about a team or treat them differently as people.
- ✓Verify with current-season data only. Last season's rating carries no weight. Always confirm the data is from this year.
- ✓Don't write off improving teams. A team that was 30th at their first event can be top-10 by Regionals. Trend matters more than overall rank.
- ✓Be honest about uncertainty. "Their TrueSkill is high but I haven't watched them today — they're unverified" is better strategy than false confidence.
✅ The best strategists are humble about data.
They look things up. They verify what they see. And they stay open to being wrong — because a surprise at a competition is always better than being wrong at alliance selection.
▶ Next Step
Data in hand. Now record what you observe at the event with your live scouting sheet.
📊 Match Scouting Sheet →