Wednesday, October 30, 2013

It’s Not Really a Tie: Jimmie Johnson Still Has the Edge over Matt Kenseth

Matt Kenseth had a career day this past Sunday at Martinsville, leading the most laps and finishing second to Jeff Gordon. Somebody who did not have a career day was Jimmie Johnson, whose fifth-place finish could actually be considered a disappointing result, given how well he’s typically performed there.

As a result of Kenseth’s outperformance over Johnson, the two of them now have exactly the same number of points in the standings (2294), with Kenseth holding the tiebreaker.

According to our calculations, however, it’s Jimmie Johnson who still has the higher chance of winning the championship. In fact, even by finishing fifth, Johnson’s title chances increased this weekend, to 53% from 52%.

Kenseth, now at 41%, made the biggest move, jumping from 33% last week. The extra 8% he gained came at the expense of Kyle Busch and Kevin Harvick, who both finished outside the top-5. The two of them combined only have a 5% title chance, much lower than their 14% last week. If you are Harvick, for example, sixth-place doesn’t cut it when your main competitors all beat you.

Jeff Gordon, with his win, is still barely hanging on to title hopes, at 1%. He will have to effectively win out the final three races to make it happen. Even though he is third in the points standings, he is fifth in our probabilities, because he has a lower chance of winning multiple races, compared to Harvick and Busch.

The expected points total for the champion now hovers around 2410, which would beat the 2403 that Tony Stewart earned two years ago.

Taking 2410 as the assumed points total needed, that would mean Johnson or Kenseth would have to earn 116 points in the final three races, or 39 points per race, which is a top-5 average finish.

For Jeff Gordon, he would need 143 more points to reach 2410, or 48 points per race, which is only doable by winning and leading the most laps in the final three races. This is unlikely to happen. Gordon fans should be less concerned about his own success, but rather hoping for Kenseth and Johnson to suffer problems. In that case, the total points necessary to win the title might come down to 2390. This would reduce the number of points Gordon needs to 123, which is 41 points per race, or a third-place average finish.

Check in next week, after Texas, for the latest probabilities.

Thursday, October 24, 2013

After Talladega, Jimmie Johnson Takes a 52% - 32% Lead Over Matt Kenseth

The big story from Talladega this past Sunday was that none of our Chase contenders crashed out – instead, they took 12 of the top 20 spots.

Unfortunately for now-former points leader Matt Kenseth, he finished in that that 20th place.

Jimmie Johnson, however, led the most laps and finished 13th, putting him into the official points lead with 4 races to go.

According to our forecasting models, that gives Johnson a solid 20% edge in winning the title:

At 52.4%, Johnson becomes the first driver since the start of this year’s Chase to have over 50% shot of winning. It’s now Johnson versus everybody else.

Kenseth is hanging in there at 32.6%, which still provides him a reasonable opportunity to win. But the shift in momentum matches our initial forecast: all along, we expected Johnson to eventually overtake Kenseth, and now we’ve reached that point.

Kyle Busch (8.0%) and Kevin Harvick (6.4%) have low title chances. Their path to winning involves flat out dominating the final four races, finishing in the top 3 at worst.

The rest of the field amounts to a rounding error – nothing short of a miracle will result in anyone else walking out of this year’s Chase with the title.

Finally, our expected points total for the champion is still hovering right around 2400, right where we ended up the previous two years.

Check in next week, after Martinsville, for updated forecasts.

Monday, October 21, 2013

How Did My Talladega Picks Do?

Let's review how my picks from last week's Talladega preview video turned out:

Talladega Favorites
1) Dale Earnhardt, Jr. (led 38 laps and finished second)
2) Jeff Gordon (led 3 laps and finished 14th)
3) Matt Kenseth (led 32 laps and finished 20th)
4) Denny Hamlin (finished 38th, after an engine problem)
5) Jimmie Johnson (led the most laps, 47, and finished 13th)
6) Brad Keselowski (finished 29th)

Like I said last week, Talladega is the most difficult track to predict winners, because of the randomness of the finishes. At least we can take some consolation that Earnhardt, Johnson, and Gordon were the three drivers who led the most laps on Sunday.

Thursday, October 17, 2013

Chase Midpoint: 89% Chance of Johnson or Kenseth Winning Title

Another Chase race – another pair of top-5 finishes for Jimmie Johnson and Matt Kenseth. By consistently finishing up front, our two leaders continue to squeeze out every other driver from any chance at a title.

The probability of Johnson (47%) or Kenseth (42%) winning the Chase is now 89%, up from 85% last week. They have not shown any signs of weakness in the past five races.

Even Johnson’s late bobble and missed opportunity to win Charlotte did not cost him much, according to these probabilities. If Johnson had won the race, he’d have a 48-41 lead over Kenseth rather than his current 47-42 lead. If he stays consistent throughout the Chase, his Charlotte mistake shouldn’t matter.

What about everyone else? Kevin Harvick (6%) and Kyle Busch (4%) have a very small chance of sneaking in, while the rest of the field is basically done – the bottom nine drivers have a combined 1% chance. The chart above does not display the lowest five drivers, because according to the model, they have no way of winning the title.

Johnson and Kenseth are so far out in front, that Jeff Gordon’s fourth place rank is actually closer in points to tenth place Carl Edwards than it is to the lead. Gordon is 36 points behind first and only 31 points ahead of tenth.

If any statistical “miracles” were going to happen, this weekend’s race at Talladega would be the place. The racing there is practically random, and the results give anybody a chance. This will be the biggest opportunity to shake up the field – and the best chance for anybody not named Johnson or Kenseth to make a move forward.

The only way for the field to make up ground is if Johnson and Kenseth both crash, but it’s more likely only one of them will, leaving the other to grab a giant points lead. That said, the most likely outcome at Talladega this weekend is less about the field catching up to the two leaders, but rather about one of them breaking away from the other.

As our two leaders continue to finish up front each week, the expected points for the champion continues to rise. Our model’s championship expectation is now above 2400. At this rate, they would break Tony Stewart’s 2403 total from two years ago:

As always, check in after the race next week for updated numbers.

Sunday, October 13, 2013

Reviewing my Charlotte Picks

In last week's BSports Video, I gave my picks for yesterday's race at Charlotte. Here's how those picks did:

Charlotte Favorites

  • Kasey Kahne (led 138 laps and finished second - got passed for the lead with 9 laps to go)
  • Matt Kenseth (led 1 lap and finished third)
  • Jimmie Johnson (led 130 laps and finished fourth - lost the lead on final pitstop)

Threats for First Charlotte Win

  • Kyle Busch (led 4 laps and finished fifth after an early pit road issue)
  • Greg Biffle (finished 16th)
  • Brian Vickers (finished 25th)

I would say 2-3-4-5 is not bad, with that group leading 273 of the 334 laps.

On to Talladega next weekend, this one will be much harder to pick. Maybe I'll just go with the "Dave Theory".

Wednesday, October 9, 2013

And then there were two: 85% Chance of Johnson or Kenseth Winning the Title

The race at Kansas finally provided an opening for the rest of the Chase field to catch up to our top 3 drivers: Jimmie Johnson (6th), Matt Kenseth (11th), and Kyle Busch (34th) all finished outside the top 5.

Kevin Harvick took advantage of this opportunity. By winning the race, Harvick's championship hopes rose to 9%.

Kyle Busch was the big loser, crashing down to 4.6%, well below his 21% last week.

Despite having several problems on Sunday, Johnson’s 6th place finish pushed his title chances to 46.7%, and has given him a reasonable lead in our predictive model:

Even though Kenseth officially leads by 3 points in the standings, our model gives Johnson the edge, because the numbers suggest Johnson has a higher chance of outperforming Kenseth in the next six races.

Kenseth’s 11th place finish at Kansas, his worst of the Chase, still boosted his title chances ever so slightly, from 37.6% up to 37.8%.

The main problem for the rest of the field: our two top drivers, Kenseth and Johnson, actually increased their chances of winning the title to 85%, up from 75% last week.

The other nine drivers only have the faintest of hopes: 1.0% for Jeff Gordon, and 0.8% for everybody else combined.

If you look closely at the table, you will see Kasey Kahne is missing. That’s because the model gives him absolutely no hope at all of winning the title. In 50,000 championship simulations, Kahne wins the title zero times. The other drivers on the list still have a miracle shot, like Ryan Newman at .002% (1 title in 50,000 simulations) or Carl Edwards at .022% (11 titles out of 50,000).

Finally, we still expect the points total for this year’s champion to be right around 2400 points, just like the last two champions:

Check in next week, after Charlotte, for updated forecasts.

Sunday, October 6, 2013

Video of Kansas Preview + Reviewing Picks

In last week's latest BSports Video, I discussed the most recent championship forecast probabilities, and previewed the race at Kansas.

How did my picks do?

Kansas Favorites

  • Jimmie Johnson (led 6 laps and finished 6th after a few problems, including entering the pits right after a caution flag came out)
  • Matt Kenseth (led 21 laps and finished 11th after a few problems, including a pit road speeding penalty)
  • Greg Biffle (led 0 laps and finished 13th - not a great bad day)

Kansas Dark-Horses

  • Aric Almirola (raced into the top 5 and finished 10th)
  • Martin Truex, Jr. (finished 19th but was in the top 5 earlier)
  • Ricky Stenhouse, Jr. (raced in the top 5, was running well but hit the wall and finished 30th)

Tuesday, October 1, 2013

Kenseth + Johnson + Busch = 96%. Everybody else 4%

For the third consecutive week, the three best drivers in the Chase dominated.  Jimmie Johnson had a convincing win at Dover, while Matt Kenseth (7th) and Kyle Busch (5th) both led many laps and finished strong.

This trio continues to pull away from the field, and has increased the chances of the title going to one of them.

Here’s the latest view:

Johnson’s win closed the gap on Kenseth in our probability table.  According to the prediction model, Johnson now has a 37.1% chance of winning the title, right behind Kenseth at 37.6%. Kyle Busch has been hanging steady in the 20-22% range the last couple weeks, and remains there at 21%

Other than Kevin Harvick at 2.7%, everybody else has a title chance under half a percent. In fact, the bottom ten drivers only have a combined 4.4% chance of winning the title. Unless we see a major turn of events in the next few weeks, the Chase is down to three drivers.

Here is some perspective on the disparity between the top three and everybody else: fourth place in the standings is Kevin Harvick, 39 points behind the leader. Yet Harvick is also 39 points ahead of Kasey Kahne, who is dead last in the Chase.

Think about that: fourth place in this Chase is just as close to last place as it is to the leader.

It’s not simply that these three drivers have an ever-increasing points lead: it’s also that they were the three most likely drivers to win the Chase in the first place. These are not surprise drivers who are having a lucky run. If they were on a short hot streak, we might expect better drivers to catch up from behind. But as it looks right now, the three best drivers are consistently delivering the best results, and increasing their lead on the field.

Finally, we see our expected championship points total is hovering right around 2400, exactly where the last two champions finished:

Check in next week, after Kansas, for updated forecasts.