Jeff Gordon’s decision to step away from full-time NASCAR Sprint Cup racing has resulted in a lot of discussion about aging drivers. We’re on the verge of a turnover as a number of drivers (Johnson, Stewart, Junior, Harvick among others) reach their forties. And what an appropriate topic for this week as I hit one of those milestone birthdays next week myself.
Slowing down is a part of aging. The print on menus shrinks, you wake up with aches and pains you can’t figure out where they came from, and you find that it takes you longer to recover from colds and injuries. Sprint Cup drivers are no different. In fact, it’s probably exacerbated because they subject their bodies to more physical punishment than your average human being.
But there are some advantages to aging. You’ve got more experience. And… well, I’m sure there are others.
So how does age affect a driver’s career? Let’s look at the numbers. (And while you’re at it, check out Eric Chemi’s blog – he took a different approach, but came up with mostly the same conclusions.)
What Do We Measure?
The challenge in questions like this is what to graph that actually makes some sense.
The first obvious thing to try is wins (or top 5s or top 10s) vs. age, right? I did this (at right) to look for obvious trends. (Note – you can click on any of these graphs and they should like to a full-size version so you can see details.)
This is pretty useless. Stewart won championships at ages 31, 34 and 40. All years where he won a respectable number of races; however, there are years where he won a lot of races and didn’t win the championship.
I also plotted Top 5s and Top 10s this way and it wasn’t any more enlightening.
So I had to re-think a little. What we’re interested in is whether a driver becomes a worse driver as he or she ages. This got me thinking about cumulative stats.
If you’re staying at the same level, you ought to add the same number of wins each year (on average, of course). So what if I plot the cumulative wins as a function of age. That turned out to yield some interesting information.
It’s always rewarding when you plot something and you realize you finally found the right thing to measure and graph. As a note, I did not include years at the end of a driver’s career where he (and they’ll all men here) didn’t run all the races that year. A number of drivers ran part-time at the ends of their careers, some for lower-tier teams and I didn’t think that would be a fair representation of their career to include those later years.
Let’s start by looking at stats for someone with a long career that spans a wide age range: Darrell Waltrip.
From top to bottom are cumulative wins, cumulative top 5s and cumulative top 10s. There are some subtle differences between the three graphs, but let’s talka bout what they have in common.
If you look at the later years, the graphs become essentially flat – which means there were no more wins, top 5s or top 10s. But the point at which they plateau changes. The wins flatten out first (no new wins after age 45), then the top 5s (only two more after age 50) and then top 10s (8 after age 50).
The areas where the slope of the graph is constant over a period of time I would characterize as consistent. They are adding to their record at the same rate. All three of Waltrip’s championships (shown in the highlighted regions) came during that period of time.
This would seem to suggest that this is a driver who reached a certain age and just couldn’t hack it after that – but there are some extenuating circumstances, namely a crash at age 43 and his transition from Hendrick to becoming a driver-owner shortly after. I’ve put a thumbnail of the graph to right – click to see it larger.
Just a warning that you have to be careful about the rationale.
A number of drivers have very similar looking graphs: Both Labonte brothers, Dale Jarrett, and Mark Martin. But in those cases, there were also extenuating circumstances in terms of changing to lower-tier teams (Bobby Labonte went from Gibbs to Petty, for example). So let’s look at the drivers who don’t follow this pattern.
Wow. You want to talk consistent? Here’s a man who (until the nightmares of the last two years) is almost one straight line from start to finish. The top 5s and top 10s are almost perfectly straight lines. The wins have a little more scatter – but that’s typical because the overall numbers are smaller. Jimmie Johnson’s graphs look very similar.
When we analyze graphs we like to talk about curvature. There’s no curvature here. If the graph curved up (i.e. looked like a saucer), that means the person was getting better. If the graph curves down (as it does when it plateaus), then the person is getting worse.
And now for one of the the interesting ones. It’s interesting in part because Jeff Gordon has driven for the same company his entire career, which eliminates the question of equipment from the analysis. Here’s the raw data for wins.
Again, it’s small to save space – click to get a larger version. This is really interesting. You can divide his career into specific segments – see how the slope changes in different ranges of years? My first attempt to explain this was to look at personal events like marriages and children. There might have been a correlation there, but them I looked at his crew chiefs.
That’s sort of interesting, huh? I didn’t make a line during Steve Letarte’s (I know, I spelled it wrong in the graph) tenure. There was a jump there, then it was pretty flat. But that’s a pretty convincing correlation, I think.
Gordon’s still very consistent when it come to the top 5s and top 10s.
Okay, But Can Older Drivers Compete Against Younger Ones?
I know. I got carried away with the data. I do that.
There’s been an influx of younger drivers – they start earlier and one might think that would lead the average age of the Sprint Cup Champion to be going down. Overall, though, it’s not. It’s going up. The most recent “Young” winner is Brad Keselowski – and he was 28 years old.
Don’t count the old folks out yet. Even at the advanced age of (gasp) 40-something, drivers like Tony Stewart (pre the last two years), Jimmie Johnson, and Matt Kenseth are remaining consistent with their performance when they started in the series.
Ever scarier, if you look at Kevin Harvick or Brad Keselowski’s graphs, they’re better than straight lines. These drivers are still improving (even as Harvick approaches 40 and Keselowski 30), which means we probably haven’t seen the best of them yet.