Much has been made of the fact that NASCAR has only had seven distinct winners this year — and that three drivers (Harvick, Kyle Busch and Truex, Jr.) have won 14 out of 19 races this year. What exactly — if anything — does the small number of distinct winners mean?

NASCAR’s Big 3 are as dominant as ever. Why are we pretending anything else matters?

A recent Charlotte Observer headline thinks it means the season is basically over.  The article went on to suggest that the seven races before the playoffs “don’t really matter” and that we all already know that only Busch, Harvick and Truex, Jr. have a shot at winning the championship.

The gods of irony are having a good laugh: All of NASCAR’s tweaking of the playoff system to create continual excitement and properly weight winning produced this story — and a lot of other stories like it.

The Danger of Distinct Winners as a Figure of Merit

A Figure of Merit is an attempt to sum up a complicated situation in one number. Your grades are supposed to sum up an entire semester’s worth of performance. A restaurant rating on the front door sums up whether they follow all the health department laws.

There are some caveats about using Number of Distinct Winners that aren’t always being observed. So here’s the disclosure:

Item 1: Who won the race says nothing about how competitive a race it was.  If Harvick won five races by a nose at the finishing line, that isn’t necessarily ‘dominating’. Leading 80% of the laps (as he did at Atlanta) is domination. Remember all the conversations about whether Bowyer’s rain victory ought to have an asterisk by it? A list of winners doesn’t tell you whether the victor raced his way to the win, or whether he made a clever strategy move and lucked out with the weather.

Item 2: I’ve shown in a couple posts that plate races are much more random than other types of races. Plate tracks are the only type of track (with the exception below) where drivers who have little chance of winning on a weekly basis have a chance of winning. In one year, you might have all four of these races won by ‘dark-horse’ cars, which adds four additional distinct winners, or you might have these races won by The Usual Suspects. I have it on my list to check the data, but I’d be willing to bet that no driver who gets into the playoffs by winning a plate race and has never won at a non-plate track has been competitive in the playoffs.

Item 3: Road courses used have the same issue as plate tracks. Back in the day, it was common to have five to eight road course ringers who had a decent shot at winning. That’s all changed. Now, it’s mostly Cup regulars. Comparing numbers of distinct drivers today with the numbers from fifteen or twenty years ago isn’t exactly fair.

Item 3-1/2: There are still drivers in the recent era (Montoya and Allmendinger come to mind) whose experience makes them more likely to win on a road course, which would up the number of distinct winners. As with the plate-only winners, I suspect that these drivers aren’t actually competitive in the playoffs.

Item 4: As you’ll see below, we’ve only been running 36-race seasons since 2001. For this and item 3, you have to exercise a little caution in comparing numbers from today with numbers from the past.

So let’s look at the data and see if the following statements are true.

Seven is a Record-Low Number of Distinct Winners

There were only five distinct winners back in 1974, but going back that far is a bit of a cheat because the NASCAR world really was a different place. Steve Waid explained the circumstances that led to three drivers (Yarborough, Pearson and Petty) winning 27 out of 30 races that year. I recommend his article.

Let’s start by looking at more recent history: the number of distinct winners after 18 races over the last 23 years. Yes, I know we’ve had 19 races, but 18 is halfway and there’s a certain aesthetic appeal to symmetry.

• The average number of distinct winners over the 23 complete years is 10.4.
• The lowest number of distinct winners is 7, which happened in 2010 and in 1996.
• The highest number of distinct winners is 14, which happened in 2003.
• For the last four years, there have been 11 distinct winners.

A histogram (how many years we’ve had each of a particular number of distinct winners) gives us some additional information:

• It is overwhelmingly likely that, in a given year, there will be between 9 and 12 distinct winners after 18 races.
• Excel tells me that the mode (the most-frequently occurring number) of distinct winners is 11.
• We’ve only have 8, 13 and 14 distinct winners at this point of the year once.
• But we’ve had three times when there were only 7 winners. (Remember this: I’ll come back to it).

So 7 is a very low number of distinct winners, but this year is not the only time we’ve been in this situation. There still a second half of the season to run.

 

It’s Too Early to Tell Anything About the Season Yet

I’ve warned before about the danger of making pronouncements for the whole season based on a small subset of early data. But are we far enough along — and is the dominance of the top three so absolute — that we can start to draw conclusions about the 2018 season?

The number of distinct winners has increased after the halfway point every single year for the last 23 years. (This isn’t to say the streak might not be longer: I only had time to examine 23 years.)  The plot below shows the same data as above, but I added the number of distinct winners in the second half in green. The total length of the green + the yellow bars is the total number of distinct winners for a whole season.

• The smallest number of new winners in the second half was 1 (in 2015)
• The largest number of new winners in the second half was 7 (in 2013 and 2001)
• The average number of new winners is 3.57

Here’s the histogram of the number of drivers who win for the first time in the second half of the season.

• The mode (the most-occurring number) is 2
• In 13 out of 23 years, we got between 2 and 3 additional winners.
  • In 9 of 23 years, we got four or more additional winners
  • In 1 of 23 years, we got fewer than 2 additional winners

So it’s more likely that we’re on track for 9 or 10 distinct winners this year. Looking at the histogram for the whole season:

• The lowest number of total distinct winners is 11.
• The largest number of total distinct winners is 19 (in 2001)
• The average number of total distinct winners is 14

But remember that we didn’t run a 36-race season until 2001. If you exclude earlier races, you get a different-looking histogram

• The lowest number of total distinct winners is 12 (2008 and 2015).
• The largest number of total distinct winners is 19 (in 2001)
• The average number of total distinct winners is 14.9
• The most likely number of total distinct winners is 13

So while we can’t predict the number of unique winners, we can say that it will almost certainly be between 7 and 14, but will probably end up somewhere around 9-10. Either 9 or 10 would indeed be a record, not just for the era of 36-race seasons, but going back to 1995. (And possibly beyond, but I haven’t looked at that data.)

We Could Still Reach 13 or 14 Distinct Winners This Season

History says that is a possibility, but let’s see if we can learn anything from years in which there were a large number of first-time winners in the second half of the season. There were…

• 7 additional winners in 2001 and 2013
• 6 additional winners in 2011, 2010 and 2005.

Of those years, I’m going to focus in on 2010 because it’s the only other season in which we only had seven distinct winners at this point of the year. (So yes, 2018 is 2010 all over again.) The next chart is a pie graph that shows the winning drivers and number of wins at the mid-way point of the season.

Important note: I didn’t code the colors to the same driver in each chart because I wanted you to see how similar the patterns actually are. The colors on each chart correspond to the first winningest, second winningest, etc. driver. We’re interested in how the wins are split up, not who’s splitting them up at the moment. The top 3 winning drivers for each year are in dark blue, light blue and red

• In 2010, Johnson, Hamlin and Kurt Busch combined to win 12 out of 18 races
  • Johnson and Hamlin each had five wins
  • Busch had two
• In 2018, Harvick, Kyle Busch and Truex, Jr. combined to win 13 out of 18 races
  • Harvick and KyBu each have five wins
  • Truex, Jr. has three

(This makes me wonder if maybe we ought to be talking about the big two. Y’think if Truex, Jr. hadn’t won the championship they would be talking about him in the same breath?)

2010 had essentially the same set up as we have right now. So how did 2010 turn out?

• Johnson went on to win only 1 more race (for a total of 6)
• Hamlin won 3 more races, for a total of 8.
• Kurt Busch didn’t win any races in the second half of the season

And the Championship?

• Kurt Busch finished 11th
• Hamlin came in 2nd
• Johnson came in 1st.

Kurt Busch is thus our poster child for why we shouldn’t assume that the final race at Homestead will come down the Harvick, brother Kyle and Truex, Jr. Yes, I know Kurt Busch only won 2 races and Martin Truex, Jr. has won three, but look with the current playoff setup, even playoff points won’t save you if you have two disastrous races in a row. And it can happen, especially given that one segment has the new Charlotte Roval and another has Talladega. Harvick has been either top-10 or 39th this year. He hasn’t had two horrible runs in a row, but it could happen.

Things Are Unlikely To Change a Lot in the Second Half

Let’s look at what happened last year. In 2017, we had 11 distinct winners by the half-way point.

• Martin Truex, Jr. and Jimmie Johnson were tied at 3 races each
• Larson, Stenhouse and Keselowski were tied at 2 each.
• Kurt Busch, Ryan Newman, Joey Logano, Austin Dillon, Ryan Blaney and Kevin Harvick each had one win.

Now compare that to the end of the year. The hashed lines represent drivers who only won in the second half.

Do you notice an important name missing from the above list? Kyle Busch hadn’t won a race by the season midpoint last year. He went on to win five races and was the second-winningest driver of the season. Jimmie Johnson didn’t win another race after the half-way point.

But This Year is Different

This is the year of the Three Dominant Drivers, which wasn’t the case in 2017. As I mention above, you could argue against including Truex, Jr. Last year, he had 5 out of his 8 wins in the second half of the season, so maybe that’s why he’s being included. But there are definitely two dominant drivers as least.

And there’s another mitigating factor. Everyone was worried last year that Chevy would have an unfair advantage this year because they have a new car body. The opposite has happened: they’ve struggled mightily. Chevy’s only got one win this year and it’s at a plate track (see Item 2, above). There are fewer Ford drivers than Chevy drivers, so the wins the Chevy drivers aren’t getting are available to a smaller number of Ford drivers.

Not to deny that Harvick, Truex, Jr. and KyBu aren’t having great years. But we won’t know how good they are until the Chevy folks figure out what they’re missing. They will, and we might all be in for a surprise in the second half of the season.

Really?

The low number of distinct winners shouldn’t be any surprise if you know the numbers. Here’s the distribution of wins by manufacturer for the first 18 races.  Red represents Chevy wins, yellow represents Dodge, blue is Ford, orange is Toyota and green is for Pontiac.

There are only two other years in which we’ve had only seven distinct winners halfway through the season: 2010 and 1996. Notice what else they have in common on the graph above with 2018?

• Halfway through this season, the only race Chevy has won is a plate race. (See item 2, above)
• Halfway through the season in 2010, Ford had not won a single race.
• Halfway through the  1996 season, Pontiac had not won a race.

Ford went on to win four races in the second half of the 2010 season. Pontiac went on to win one race in the second half of the 1996 season.

 

So let’s be a little cautious. I’d argue that the three drivers dominating the series are able to do so because the competition is falling down on the job. That all changes if Chevy figures out their body problems. Truex won five races in the second half of 2017. I wouldn’t put it past Kyle Larson to pull off the same thing.

Bonus Graph

Do Drivers Who Haven’t Win During the Regular Season Win During the Playoffs?

Here’s a chart showing the number of distinct winners for the first 26 races in dark blue and the last 10 (which corresponds to the playoffs after 2004) in light blue.

• Between 1 and 4 drivers who haven’t won in the regular season will win during the playoffs.
• In 2016 and 2014 (and 1999) no new drivers won
• 2015, only one new driver won
• In seven years, there were two new drivers
• In six years, there were three new drivers
• In 2013 and 2001, four new drivers won

 

 

 

 

 

 

 

 

 

 

The Problem with ‘Number of Distinct Winners’

A figure of merit is a number that allows a person to characterize something, like a HVAC system or performance in a class or… or a racing season. Just like that ‘C’ you got in sophomore English because you had a panic attack during the final doesn’t reflect your overall performance in the class, using the number of distinct winners to characterize the entire racing season leads to some problems.

If you couldn’t beat them the past 19 weeks, what’s different now?

This season, too, has been something of an anomaly. Having so few different winners is not only uncommon; it’s never really happened before.