Note:
Most posts will rely heavily on data. Blogger does not always allow
for easy integration of chart and tables into posts, so I've made data
available in a Google Spreadsheet called 2015 IndyCar Preseason Analysis - Finishing Group Model.
Last time, I tried to predict how many points the 2015 IndyCar champion would score during the season (excluding qualifying for the Indianapolis 500) based on champion points/race data. In this post, I use a different method to predict how many points the 2015 champion will score. Like before, data is limited to champions from 2012 through 2014, the years that the DW12 chassis has been used in IndyCar. Points for qualifying at the Indianapolis 500 and Iowa have been removed from champions’ end of year totals.
Last time, I tried to predict how many points the 2015 IndyCar champion would score during the season (excluding qualifying for the Indianapolis 500) based on champion points/race data. In this post, I use a different method to predict how many points the 2015 champion will score. Like before, data is limited to champions from 2012 through 2014, the years that the DW12 chassis has been used in IndyCar. Points for qualifying at the Indianapolis 500 and Iowa have been removed from champions’ end of year totals.
I started by collecting the champion’s finishes for 2012 through 2014 (see Chart 1 - Champion Finishes). I also collected the number of times each champion scored bonus points (by either winning the pole, leading a lap, or leading the most laps in a race). For 2015, IndyCar will award points for each of these achievements, so it seems worthwhile to include them in the model. I counted double points races (Indianapolis 500, Pocono, and Fontana, all highlighted in yellow) as 2 races, so the adjusted 2014 schedule had 21 races, even though the real 2014 schedule only had 18 races (double points were not awarded in 2012 or 2013). Since double points will be awarded at 2 races in 2015 (Indianapolis 500 and Sonoma), the 2015 schedule has 18 adjusted races.
Chart 1 - Champion Finishes
Chart 1 - Champion Finishes
| Year | 2014 | 2013 | 2012 |
| Champion | Power | Dixon | Hunter-Reay |
| Total Races | 21 | 19 | 15 |
| Race 1 | 1 | 5 | 3 |
| Race 2 | 2 | 2 | 12 |
| Race 3 | 5 | 11 | 6 |
| Race 4 | 8 | 18 | 2 |
| Race 5 | 8 | 14 | 27 |
| Race 6 | 8 | 4 | 7 |
| Race 7 | 1 | 4 | 21 |
| Race 8 | 2 | 23 | 1 |
| Race 9 | 2 | 6 | 1 |
| Race 10 | 14 | 16 | 1 |
| Race 11 | 11 | 1 | 7 |
| Race 12 | 10 | 1 | 24 |
| Race 13 | 10 | 1 | 18 |
| Race 14 | 14 | 7 | 1 |
| Race 15 | 9 | 15 | 4 |
| Race 16 | 3 | 19 | x |
| Race 17 | 6 | 1 | x |
| Race 18 | 1 | 2 | x |
| Race 19 | 10 | 5 | x |
| Race 20 | 9 | x | x |
| Race 21 | 9 | x | x |
For this model, I combined finishing positions into the finishing groups (FGs) listed below.
- Wins
- 2 & 3
- 6 to 8
- 9 to 12
- 13 to 17
- 18 to 24
- 25+
I determined the number of times that the champion finished in each FG and earned bonus points (Chart 2 - Champion Finishes/FG), then converted those numbers into percentages of the entire season (Chart 3 - % Champion Finishes/FG). For example, 2013 champion Scott Dixon finished 4th or 5th in four races, or in 21% of the races in 2013 (4 finishes divided by 19 races). I then calculated the average percent of the time the champion finished in each FG, and took an average of the percents from each season.
Chart 2 - Champion Finishes/FG
Chart 3 - % Champion Finishes/FG Chart 2 - Champion Finishes/FG
| Year | 2014 | 2013 | 2012 |
| 1 | 3 | 4 | 4 |
| 2 & 3 | 4 | 2 | 2 |
| 4&5 | 1 | 4 | 1 |
| 6 to 8 | 4 | 2 | 3 |
| 9 to 12 | 7 | 1 | 1 |
| 13 to 17 | 2 | 3 | 0 |
| 18 to 24 | 0 | 3 | 3 |
| 25+ | 0 | 0 | 1 |
| Poles | 4 | 2 | 0 |
| Most Lead | 4 | 3 | 2 |
| Races lead (excluding wins) | 8 | 4 | 2 |
| Year | 2014 | 2013 | 2012 | Average (DW12) |
| 1 | 14% | 21% | 27% | 21% |
| 2 & 3 | 19% | 11% | 13% | 14% |
| 4&5 | 5% | 21% | 7% | 11% |
| 6 to 8 | 19% | 11% | 20% | 17% |
| 9 to 12 | 33% | 5% | 7% | 15% |
| 13 to 17 | 10% | 16% | 0% | 8% |
| 18 to 24 | 0% | 16% | 20% | 12% |
| 25+ | 0% | 0% | 7% | 2% |
| Pole | 19% | 11% | 0% | 10% |
| Most Lead | 19% | 16% | 13% | 16% |
| Races lead (excl win) | 38% | 21% | 13% | 24% |
Next, I multiplied the average percent of the time the champion finished in each FG and scored bonus points by the 18 adjusted races there will be in 2015. This produced the number of predicted finishes for each FG (Chart 4 - Expected Results). From 2012 to 2014, the champion won an average of 21% of the races each season, suggesting that the 2015 champion will win 3.7 races in 2015. I rounded each predicted finish to the nearest whole number (with one exception) since it’s pretty difficult to win half of a race.
Chart 4 - Expected Results
| Predicted Finishes | Rounded Finishes | Points/Race | Total Points | |
| 1 | 3.7 | 4.0 | 51 | 204 |
| 2 & 3 | 2.6 | 3.0 | 37.5 | 112.5 |
| 4 & 5 | 1.9 | 2.0 | 31 | 62 |
| 6 to 8 | 3.0 | 3.0 | 26 | 78 |
| 9 to 12 | 2.7 | 3.0 | 19.75 | 59.25 |
| 13 to 17 | 1.5 | 1.0 | 15 | 15 |
| 18 to 24 | 2.1 | 2.0 | 12 | 24 |
| 25+ | 0.4 | 0.0 | 10 | 0 |
| Poles | 1.8 | 2.0 | 1 | 2 |
| Most Lead | 2.9 | 3.0 | 2 | 6 |
| Races lead (excl win) | 4.3 | 4.0 | 1 | 4 |
| Grand Total | - | - | - | 566.75 |
Then I determined the average number of points a finish in each FG was worth. A win is worth 50 points, and a race winner always gets a bonus point for leading a lap because the race winner always leads the last lap. Therefore, a win is worth 51 points. Finishing 2nd is worth 40 points, and finishing 3rd is worth 35 points. The average of 40 and 35 is 37.5 points per finish in the 2nd and 3rd place FG.
Finally, I multiplied the rounded finishes in each FG by the average points awarded to that FG. This model predicts that the 2015 IndyCar champion will score 566.75 points, excluding Indianapolis 500 qualifying.
I’ll end this post with the same disclaimer from the end of the post explaining the champion points/race model: this model assumes that there will be little performance difference between the Chevrolet and Honda engines and aero kits that will debut in 2015. If one manufacturer outperforms the other (especially if Honda is superior to Chevrolet), the champion will probably score more than 30.3 points/race.
If you want to get in touch, feel free to leave a comment or send an email to spottersstand@gmail.com. Follow the blog on Twitter (@spotstandblog) for reminders about new posts.
I’ll end this post with the same disclaimer from the end of the post explaining the champion points/race model: this model assumes that there will be little performance difference between the Chevrolet and Honda engines and aero kits that will debut in 2015. If one manufacturer outperforms the other (especially if Honda is superior to Chevrolet), the champion will probably score more than 30.3 points/race.
If you want to get in touch, feel free to leave a comment or send an email to spottersstand@gmail.com. Follow the blog on Twitter (@spotstandblog) for reminders about new posts.
No comments:
Post a Comment