Day 10
Season 1 Analysis Block
Season 1 · Analysis Block · 25 min
Time Left
25:00

Your Season 1 Data

Enter your Error values, then answer three questions about the pattern.

8–10 min
Add today's matches to the audit from before Round 1. You now have the complete Season 1 Error field.

Error = Predicted Goals − Actual Goals. Positive = over-predicted. Negative = under-predicted.
What is Error?
EXAMPLE 1 — PREDICTED TOO HIGH PREDICTED 7 goals − ACTUAL 5 goals = ERROR +2 over-predicted bot scored LESS than expected EXAMPLE 2 — PREDICTED TOO LOW PREDICTED 3 goals − ACTUAL 5 goals = ERROR −2 under-predicted (negative number) bot scored MORE than expected
Season 1 Error Values — All Matches
Match
Error
Velocity Arena bot
Error = Predicted − Actual. Keep negatives negative. Predicted 5, scored 7 → Error = −2. Leave unused rows blank.
Step 1 — Seasonal Average Error
Sum of Errors
—
Match Count
—
Average Error — Bias
—
tells whether your model usually predicts too high or too low. Do not call it "accuracy."
Keep negatives negative. Do not turn every error positive when calculating the average. Average error measures bias — dropping the sign destroys the meaning.
Cancellation warning: +4 and −4 average to 0, but both were large misses. Use trend and biggest errors too — average error alone can hide a bad model.
Step 2 — Sign of Average Error
Based on the average above — choose one.
Positive
Consistently over-predicted
Negative
Consistently under-predicted
Near Zero
Prediction was balanced
Step 3 — Error Trend Across Season
Compare the size of your first Day 7 Error to the size of your last match Error today. Ignore + or − signs — only the magnitude matters.
First Error size (Day 7 · R1) — no minus sign
Last Error size (Today · R3) — no minus sign
Enter the size only — always positive. If your error was −3.5, enter 3.5.
Step 1
0 / 3

Find the Pattern

Name the input variable your model systematically gets wrong.

6–8 min
Look at your three biggest Errors — matches where prediction was furthest from result. For each, check your "What happened" notes. What was your opponent's dominant stat?
Your Season 1 Error Data
Hide ▲
Match Error
Top 3 Biggest Errors — by Size
Look at the error data table above. Find the 3 matches where your error number was largest — ignore + or − signs, just pick the biggest numbers. Copy each error value below, then check your match notes to answer: which robot stat was that opponent especially strong in?
Robot stats to look for in your notes: Speed, Turning, Power, Endurance, Kicking
1
Which stat was that opponent strong in?
Your error for that match
What happened? (from your match notes)
2
Which stat was that opponent strong in?
Your error for that match
What happened? (from your match notes)
3
Which stat was that opponent strong in?
Your error for that match
What happened? (from your match notes)
Write Your Sentence
Must name an INPUT variable — not a result. "We lost to fast teams" = result. "Turning is missing from our equation" = input.
My model consistently because it .
That looks like an outcome, not an input variable. Examples of input variables: Speed, Turning, Power, Endurance, distance, launch angle. Do not write: lost, won, goals, score, opponent, match.
Most Common Error

What you might write: "We lost to fast teams." — That is a result.

What you need: "Our equation uses only Speed, but high-Turning opponents scored from angles our bot didn't cover — Turning is missing from the model."

Ask: Which variable that actually affected the match is NOT in y = mx + b?

Step 2
0 / 2

Build Your Hypothesis

A calculation, not a preference. Your scatter plot equation must appear.

6–7 min
Open your scatter plot equation from Day 8 — y = mx + b, where x = Speed allocation and y = predicted goals. Plug in your proposed new value. The math must appear in your hypothesis.
What each part of y = mx + b means
y = mx + b y Predicted goals m Slope (rate) x Your Speed value b Intercept (starting point)
Does Your Stat Change Fix The Problem?
Your Step 2 analysis named a variable your model is missing. Your hypothesis must change a stat that actually addresses that gap — not just any stat.
The variable your model is missing (from Step 2):
—
The robot stat you want to change for Season 2:
Explain the connection — how does changing this stat address the problem you found?
Enter Your Day 8 Scatter Plot Equation
Equation: y =
x +
Current Speed Value (x)
Proposed New Speed Value (x)
Season 1 Actual Avg Goals / Match
Stat You Are Changing
Can't find your Day 8 equation? Use the class average equation on the board. Note it's an approximation.
Season 2 Hypothesis Builder
1
Stat Change
Fill equation fields above to auto-populate.
2
Equation Output
Fill equation fields above to auto-populate.
3
Model-Predicted Difference
Fill equation fields above to auto-populate.
4
Why does this stat change fix the problem?
Your stat change must connect to the missing variable you identified in Step 2.
Hypothesis Quality Spectrum

Not acceptable: "I will change Speed from 10 to 14 and score more goals." — no calculation.

Acceptable: "If I change Speed from 10 to 14, my equation y = 0.4x + 1.2 gives y = 6.8. Season 1 actual avg: 4.2 goals. Model-predicted difference: +2.6 goals/match."

Strong: Calculation present AND directly addresses the weakness you identified in Step 2.

What Would Prove You Wrong?
What Season 2 result would show your hypothesis was wrong?
Example: "If our error against high-Turning teams stays above 3 goals/match, the stat change did not fix the problem."
Your Season 2 Hypothesis
Hypothesis
Incomplete