Get Faster Without Running a PR: Goal Setting with Statistics in Mind

With new year resolutions still on the mind, I thought it would be important to explore how to set running goals for the new year. Running goals come in a couple of different flavors. You have the super sweet donut goal that might be the pursuit of a new personal best/personal record (PB/PR) or an equally shallow goal that might involve impressing other people. This goal can be highly motivating and often times it can become the main thing you think about when you visualize success.

Now, these goals can be useful. Life without donuts is not a life worth living. But the problem is that a donut is fat + simple carbohydrates that tastes amazing when you eat them but quickly burn away leaving you hungry 10-seconds later. Doughnut PRs taste amazing when they happen, but at best the joy is short lived and you become hungry for the next PR goal just days (or hours) after. At worst, a PR goal can haunt you and tempt you to run every workout faster and every long run harder, draining you of the joy running is supposed to give you as the pursuit of another PR takes control. Many a runner has become haunted by the PR-That-Never-Was, a race time that seems forever out of reach no matter how hard they try to reach it.

After donut goals, we have what I'll call the steak dinner goal. A steak dinner goal is one that leaves you satisfied and full for a long time. Where the donut goal's joy is short lived, the steak goal's satisfaction stays with you. Afterwards, you enjoy a glass of wine and share some wonderful conversation with others. The evening continues on, long into the night where you make memories that last a lifetime. A steak goal can sometimes be a new PR, but it is more often than not a goal that goes deeper than a race result. A steak goal has more to do with you as a person and an athlete. A steak goal, like the steak itself, takes a long time to create. The steak has to come from a cow that is raised (hopefully) in a grassy field, where ranchers tend to it daily and invested time in it so the cow grows. The steak is then butchered and cooked by a patient hand in order to ensure that the steak is neither too rare nor too well-done. Then it is served with other dishes and consumed slowly and with pleasure. The steak dinner goal is very different from the donut goal. This goal has to do with self-improvement. It may involve overcoming a personal obstacle, a personal failure, fixing a constant problem, showing bravery, or risking something sure for something sunsure but more valuable.

What are some donut goals? Well, here are a few I often hear.

Donut Goals: 1) Run under 20-minutes for a 5k, 2) qualify for the Boston Marathon, 3) place in my age-group, 4) just finish the race, 5) run a new PR.

Problems with the time goal: This is a common donut goal that I often have in my head. Not shockingly, my brain wants lots of donut goals. There are just a couple problems with this donut goal. For one, the goal is useless. There is no functional difference between 20:01 and 19:59. A better goal is to USE the time goal to create a steak dinner goal.

A) Work to improve a weakness: Maybe you really hate tempo runs but you love track workouts. Or perhaps you're really a track workout person but hate long runs and tempo runs. Whichever one you dislike is probably the one where your weaknesses lie. Put more effort and focus into these types of workouts and improve this aspect of your training.

B) Aim to further a strength: Are you really good at tempo runs? Extend them a bit or speed them up and push this strength to a new level. Do you really love track workouts? Add a few extra reps to the workouts and increase your volume of work.

C) Improve your weekly training schedule: What aspect of your training needs to be improved the most? Frequency? Maybe you had 5 aerobic/endurance exercise sessions per week during your last training cycle. Commit to 6 for your next training cycle to improve your race time. Maybe you ran your long runs too fast. Slow down for a whole training cycle and see what it does. Maybe you've done the same workouts for the last two years? The law of diminishing returns suggests that you won't get much out of those sessions moving forward, so maybe find some new workouts to do to create a new stimulus for adaptation.

D) Practice being a better competitor and racer: Maybe you constantly go out too fast and get passed by a lot of people at the end of the race. Or maybe you try to stay with a group but usually end up getting dropped by them sometime during the race. These are polar opposite problems, but they both need experimentation and courage to improve. If you go out too fast usually, then be brave and go out slower next time EVEN IF that means you runs slower than you want to. Be brave. If you get dropped by a group of runners and can't stay with them, then be brave and allow yourself to try and fail. Commit to either stay with the group and finish strong or stay with the group and blow up completely at the end and have to start walking. Be brave!

without trying to run a new PR all of the time. Basically, the goal is to get faster without trying to set a PR. Get faster without running a PR? "But that's crazy," you must be thinking.

Is it though?

Most people approach a race by thinking about their recent training (or lack thereof), and their PR in that distance.

(this might sound familiar.)Tell me if this sounds familiar. You have just finished a race, and you've run a PR. You're ecstatic and excited, and you're already looking toward the next race where opportunity awaits for another PR. If you just ran a 5k, you might think: could I run 15 seconds faster? What about just 10 seconds faster? That can't be too hard, right? If you just ran a marathon PR, you're probably thinking to yourself: 2 minutes faster is only 4 1/2 seconds per mile faster. Surely I can do that!

Then you get up to the start line of your next race, prepared to suffer and run a new PR, and you inevitably fail. You feel like the race "didn't go as planned" or you just "used the race as a workout" or you "just tempod the race." All common covers for a disappointing run. Or maybe you say that you were just "hoping to finish." Another cover for a disappointing run. But if that all you were really hoping to do, then you would't care that much about your pace (right when you cross the line, or during training). You wouldn't wear a watch. You wouldn't run workouts. This is a defence mechanism to protect against disappointment. But, this type of thinking is wrong. It is a mental fallacy. You can't hope to run close to your PR every race because that's not how life or statistics work.

We are hopeful creatures, which is great when establishing goals and setting sights high for the future. And goals are necessary to progress and keep from stagnating. Without goals, we just atrophy, both physically and mentally, and things begin to deteriorate.

Goals don't even have to be about running faster. In fact, I try to stay away from only having time-based running goals. When I was in high school, I was obsessed, as any 800-meter runner would be at 17, with breaking 2-minutes. But race after race, I tried to break 2-minutes, and I failed. I developed a very negative relationship with running and races, and I became very anxious whenever I toed the line.

You see, I failed to realize (because I was 17) that I am not a clock or a robot. I could not just "program" myself to run sub-2. Only, after seven or eight 2:00.XX races (including one where my electronically timed finish was 2:00.00 with lots of agony afterwards), did I run my first 1:59.

But that race when I first broke 2-minutes was different from the others. I focused on racing that day and trying to beat people. It was only after I shifted my focus away from something I couldn't control (time) to something I could control (competing against people) that I ran my fastest time. In fact, during my last marathon, I specifically had no time goal. Instead my goal was to run conservatively enough that I felt strong through the last 6 miles and could compete against other runners. I was not in great shape going into that race, but I only ran 1% slower than my PR (partly because I did not focus too much on time).

The problem with the perpetual PR seeking (or at least PR matching) is that it ignores fundamental facts of statistics, which is to say: life. While this may sound crazy, when you run a PR or have some other outlying performance (like fastest time in 5 years) then you have to expect your running times to be slower the next time you run that race. This is because of a fun statistical phenomena called 'regression to the mean.'

If you took stats in high school or college you may remember this. Basically, any outlying performances (in running these are amazingly fast or depressingly slow race times) are just that: outliers. Which means that the next race after a recent PR or fast time is more likely than not going to be slower than what you just ran. Similarly, if you had an uncommonly slow race, then your next one is likely to be faster.

Steve Magness, head coach at the University of Houston, tracked the PR progression in women 5k times and found the same pattern. A runner would run races around the same time give or take 5 or 10 seconds (near the mean) and then she would have a huge jump and PR and run 20 or 30 seconds faster than she had ever run before. What happened during her next race? She regressed back to the mean and ran a slower time. The mean might now be faster than before, but she did not run close to her PR very often. Most of the time it was quite a bit off. The runners would continue this pattern until the next big jump of 15 to 30 seconds.

So, where does this leave us regular runners? Well, for one, establishing and defining a successful race should NOT focus on just your PR. A better method would be to find your mean time for the race distance and then make your goal to improve your mean over the course of a year or two.

Let's use me as an example.

2014 - 16:50* (1st place, alone), 17:05

2015 - 17:05, 17:25* (1st place, alone), 17:14, 16:56

2016 - 16:50 PR, 17:05, 16:41 PR, 17:26**, 17:20**

2017 - 17:14, 18:09 (1st place, alone), 18:07 (1st place, >90F, alone), 17:15, 18:01**

2018 - 18:03 (1st place, alone), 17:56 (1st place, >90F, 30 mph wind, alone), 17:30**

* Converted the time because of hilly course (for every 10 ft of climbing you subtract 1.74 seconds)

** Converted from a cross-country race

So, now that I have all of my data, I can sit down and think about how to set goals and how to define a successful 5k race. For one, I need to calculate my mean race time. For this calculation, I am going to exclude some but not all outlying performances.

First, I am going to exclude races where I ran alone, which means that it wasn't much of a race and it was more like a hard tempo run. Racing against other people makes me run faster, but when I'm alone in a race then I'm far less motivated to push when it gets uncomfortable.

Second, as mentioned earlier, I needed to equalize some of my races with hills in them by making a best-guess approximation of what my time would have been on a flat course. This way, I can compare flat course runs with hilly course runs.

Third, I converted some of the cross-country races that I ran into 5Ks by correcting for hills and then taking the 5K split from that race.

And fourth, I converted the times to decimal numbers. So here are my race times expressed as decimal numbers. 16.83, 17.08, 17.08, 17.23, 16.93, 16.83, 17.08, 16.68, 17.43, 17.66, 17.23, 17.25, 18.01, 17.50.

My mean comes out to 17.16 or 17:12 or an average pace of 5:28/mile. Now that is my overall mean or average over the course of the last 4 years. It makes more sense, however, to look at my average over that last year or two and aim to improve that. My race times over the past two years give me an average 5k time of 17:28 or 5:37/mile.

So, how do I use this for goal setting and how do I know when I am successful? What if I run an 18:00 5k? Is that a successful race or should I feel let down by it? Well, remember the main problem: regression to the mean. For me, based on these numbers, I should not feel bad when I run an 18:00 5k because that is a slow, outlying performance that will likely happen through the course of the next year of racing. It only becomes a problem and disappointment is it happens more than it should, based on the normal distribution of race times I can expect next year. Let's take a look.