The presentation is a summary of a half-year research. It shows how a careful allocation of funds possessed by team may influence its final position at the end of a season. The model predicted the whole podium of 2011 F1 season.
What Makes The Formula One Champion. Regression Analysis
1. What Makes
The Formula One
Champion?
Regression Analysis
Wojciech Latocha
2. WHAT IS FORMULA ONE?
• the highest class of single-seater auto racing
• the season is compounded of a series of races, known
as Grand Prixes
• races are held on purpose-built circuits and public roads
• results of each race are added to define two annual World
Champions:
• individual
• constructor
3. THE IMPORTANCE OF THE F1
ANALYSIS
• Social factors:
• Worldwide popularity
• each F1 Grand Prix attracts a global audience of 600 million
people
• 2001 F1 season: 54 billion TV viewers, 200 countries
• Economic factors:
• Perfect example for Snob Effect and Veblen Good concept
• Formula One revenues are equal to GDP of Mongolia or
Kyrgyzstan and almost two times higher than GDP of Faroe
Islands
4. LITERATURE REVIEW
• Numerous general nature book publications and articles:
• Explaining rules
• Presenting history
• Describing former World Champions
• Liu Xiao: Formula One’s Financial Crisis, 2009
• A few scientific resources:
• Eichenberger and Stadelmann, Who Is The Best Formula 1 Driver? An
Economic Approach to Evaluating Talent, 2009
• Mastromarco and Runkel, Rule changes and competitive balance in
Formula One motor racing, 2009
5. DATA
• The official website of the Formula One
• Official websites of Formula One teams
• Official websites of drivers
• Formula Money
8. SIMPLE DATA SERIES
• Was the individual world champion racing for the team? –
a dummy variable
• Number of years that a main engineer spent with a team
9. RATIOS
• Percentage spread between drivers' points
• First driver salary / First driver podiums
• First driver salary / First driver positions awarded with points (excluding
podiums)
• Second driver salary / Second driver podiums
• Second driver salary / Second driver positions awarded with points
(excluding podiums)
• Teams' testing laps / The most active tester's laps
• Pit-stop spread
10. REASONS OF THE DISTINCTION
• Changes in regulations
• Forecasting disability of some simple series (no ability of
their value's control)
• Driver’s transfers
• Number of competing team’s changes
11. SUMMARY OF STATISTICS
Standard
Variable Mean
Deviation
final position 6.018182 3.223128
was the individual world champion racing for the team? 0.090909 0.290129
percentage spread between drivers' points 0.692449 0.341747
number of years main engineer spent with the team 2.927273 2.417946
teams' testing laps to the most active tester's laps 0.69628 0.270442
pit stop spread 0.187378 0.193195
1st salary/podiums 4.589197 6.190581
1st salary/points 2.642526 3.218502
2nd salary/podiums 3.351749 5.292386
2nd salary/points 1.857905 2.539029
12. RESULTS OF THE REGRESSION
final position Coefficient Standard Error t P>t
was the individual world champion racing for the team? -1.86251 1.232371 -1.51 0.138
percentage spread between drivers' points 2.287651 .9805094 2.33 0.024
number of years main engineer spent with the team -0.27952 .138966 2.01 0.05
teams' testing laps to the most active tester's laps -1.88072 1.584256 -1.19 0.241
pit stop spread 2.643283 1.948494 1.36 0.182
1st salary/podiums -0.07556 .0665366 -1.14 0.262
1st salary/points -0.1668 .1104518 -1.51 0.138
2nd salary/podiums 0.195983 .0838879 2.34 0.024
2nd salary/points -0.43651 .1611728 -2.71 0.01
constant 7.177511 1.498906 4.79 0
13. DISCUSSION
• 7.18 - an initial position of any team competing in The Formula One
• 1.86 - a position improvement due to employing a former individual World
Champion
• 0.23 - a result depreciation caused by each 10% of a difference between
points collected by each of two drivers
• 0.28 - a position increase through prolonging team’s cooperation with its
main engineer
• 0.44 - a position improvement due to each $1 million paid to the second
driver for scoring a position points-awarded
• 0.2 - a position drop caused by each $1 million paid to the second driver
for scoring a podium
14. FORECAST FOR 2011
Virgin-Cosworth
HRT-Cosworth
Lotus-Renault
Williams-Cosworth
STR-Ferrari
Sauber-Ferrari
Force India-Mercedes
Renault
Mercedes
Ferrari
McLaren-Mercedes
RBR-Renault
0 2 4 6 8 10 12
actual final position predicted position
15. CONCLUSION
• The model was able to predict an exact final position of one out
of twelve teams
• Numerical values returned by the equation indicated four out of
twelve positions in the final standing, including:
• The World Champion
• All three teams on the podium
• The forecast indicated a fierce battle for a few positions – what
was proved on the track
• The model underestimated a result of the first seven teams and
overestimated the result of the others