Big Data approach to running

RayCun

Because there must be a Big Data approach to everything these days

http://www.pace-man.com/

Crowd based in UCD, they have taken the data from a lot of marathons and are using it to devise pacing plans (and eventually training plans). The idea being that they take 10,000 people who ran sub 4 in Dublin, find out where they went slower and faster, and give you a pace strategy based on that. (They slice the data finer than that, by age, gender, a few other things, but that's the basic idea)

I think their approach has some fundamental flaws but it's interesting to see them give it a go, and they will come out to your club/training group to talk about what they're doing.

TFBubendorfer

So the average runner is now the role model that we should all try to emulate?

That may well be the worst approach to marathon pacing I have come across yet.

RayCun

Yeah, that was my objection too

They exclude people who hit the wall, which they define as 33% slower in the second half. But that means they are including people who ran 25% slower. Their pace plan is a slow start, fast middle, and slow finish.

TFBubendorfer

RayCun wrote: »

They exclude people who hit the wall, which they define as 33% slower in the second half. But that means they are including people who ran 25% slower. Their pace plan is a slow start, fast middle, and slow finish.

So if I ran the first half in 2 hours I'd have have to run the second half slower than 2:40 for them to count that as hitting the wall. :eek:
I've hit the wall a few times, especially in my early marathons, but even my worst ever slowdown was still significantly better than that!


And that's not even mentioning the fact that checking your pace every 500 meters on a gadget is the worst possible way to run a race, never mind the actual pacing strategy.

Timmaay

I'll be honest I don't know the 1st thing about marathon pacing (and not sure I ever do haha), but definitely feel free to drop them a quick email pointing out what you think the flaws are with their methods.

RayCun

Timmaay wrote: »

I'll be honest I don't know the 1st thing about marathon pacing (and not sure I ever do haha), but definitely feel free to drop them a quick email pointing out what you think the flaws are with their methods.

They came out and spoke to the club. Objections were raised

I think there is some potential to the idea of using all the data that is there on strava to learn more about how people run, but it needs a lot of work to get from description to prescription.

and still ricky villa

This is like Dragon's Den
At first sight you think 'This is a great idea' until people start to point out the flaws.

Run and Jump

RayCun wrote: »

They came out and spoke to the club. Objections were raised

Like a lot of clubs are probably doing around this time of the year, we had our club Dublin Marathon info/prep-talk/get-together last night, with our coach being an old-school former DCM winner, sessions and targets being set, and all of us fired up for a summer of hard training and harder racing. If these well-meaning folk plan on going around the clubs of Ireland to talk about an average-runner DCM approach, I fear for their safety. They'll be ate out of it!

skyblue46

TFBubendorfer wrote: »

So if I ran the first half in 2 hours I'd have have to run the second half slower than 2:40 for them to count that as hitting the wall. :eek:
I've hit the wall a few times, especially in my early marathons, but even my worst ever slowdown was still significantly better than that!


And that's not even mentioning the fact that checking your pace every 500 meters on a gadget is the worst possible way to run a race, never mind the actual pacing strategy.

I'm wide open to learning here. I'm not long running but can readily admit to being a slave to the watch. I'm assuming, perhaps wrongly, that you can't train by the watch and race without it? What do you suggest as the ideal model?

Enduro

Being a bit of a data nerd I appreciate what the're attempting, but totally agree that more than likely their approach is doomed as a result of some very flawed starting assumptions. The vast majority of runners are poor pacers. That might sound elitist, but I don't care since I've spent a long running career taking full advantage of that to outrace the competition, often by simply pacing more effectively. There is unlikely to be much to be gained by basing a strategy on poor judgement on a big (data) scale.

skyblue46, I wouldn't discount that approach, but the ability to learn how to feel a pace and judge exactly what effort to deploy in a particular race is key. I'm not sure I have ever heard of an effective nailed on approach that can teach how someone can do that effectively. But the start-point is definitely the simplest thing... just to be aware that this is something to target and strive to attain and use every opportunity to consciously build your knowledge/feel.

TFBubendorfer

skyblue46 wrote: »

I'm wide open to learning here. I'm not long running but can readily admit to being a slave to the watch. I'm assuming, perhaps wrongly, that you can't train by the watch and race without it? What do you suggest as the ideal model?

As ever, there is more than one way to do it.

Ideally you have developed such a feeling for the pace that know what you're doing without the need to check yourself against a gadget. Few attain that level of self awareness.

The worst way is the one they are suggesting with their Big Data app, constantly having an eye on the gadget. If you check your pace every 500 meters you'll end up doing that 84 times in a marathon :eek: The mental energy you'd waste with that alone will have a negative impact on your race, it would become a real stress factor.

I find it helps to have a watch for a long race, just to re-assure yourself that you're on pace, and to have some readily available feedback. I find I check it reasonably often in the early miles (say, every mile or 2) but once you're settled in the effort you don't really need it any more.

But ideally you want to go by feel and not the watch/phone.

Halloween Jack

TFBubendorfer wrote: »

As ever, there is more than one way to do it.

Ideally you have developed such a feeling for the pace that know what you're doing without the need to check yourself against a gadget. Few attain that level of self awareness.

The worst way is the one they are suggesting with their Big Data app, constantly having an eye on the gadget. If you check your pace every 500 meters you'll end up doing that 84 times in a marathon :eek: The mental energy you'd waste with that alone will have a negative impact on your race, it would become a real stress factor.

I find it helps to have a watch for a long race, just to re-assure yourself that you're on pace, and to have some readily available feedback. I find I check it reasonably often in the early miles (say, every mile or 2) but once you're settled in the effort you don't really need it any more.

But ideally you want to go by feel and not the watch/phone.

What difference does it make, if you pace it properly and finish in the time you wanted, whether you paced by feel or by your watch? And I’d suggest it’s pretty easy to know by feel what roughly 7 mins per mile is for example, but can you effectively ‘feel’ the difference between 7 and 7:10? Cause over 26 miles or more that 10 makes a sizeable difference

Swashbuckler

skyblue46 wrote:

I'm wide open to learning here. I'm not long running but can readily admit to being a slave to the watch. I'm assuming, perhaps wrongly, that you can't train by the watch and race without it? What do you suggest as the ideal model?

I think you can to be honest. I think you can spend a lot of time in training being a slave to the watch to a certain degree as long as you constantly tune in to how you feel. Eventually over time you will become less reliant on the watch as you'll just know by how you feel. It's a hard hard thing to learn though. And most people (myself included) struggle with it.

TFBubendorfer

Halloween Jack wrote: »

What difference does it make, if you pace it properly and finish in the time you wanted, whether you paced by feel or by your watch? And I’d suggest it’s pretty easy to know by feel what roughly 7 mins per mile is for example, but can you effectively ‘feel’ the difference between 7 and 7:10? Cause over 26 miles or more that 10 makes a sizeable difference

You're looking at it from the wrong angle. Of course once you finished in a time that makes you happy it doesn't make a difference if you paced by feel or watch.

The point is that you're more likely to succeed if you have a good feeling for your effort.

As for your example, if I check my watch every mile or so for the first 5 miles, I can be very close to 7:00 pace if that's my target pace. After that I am tuned into the effort and I can indeed tell if it's 7:00 or 7:10 - the difference between the 2 is easily big enough to notice. After that, an occasional glance at the watch is more for reassurance than anything else.

Btw, that's only a minor side argument to the original point, namely the value of this "Big Data" approach. The far bigger concern is that trying to emulate the average marathon runner's pacing strategy is just wrong. Most runners are piss poor at pacing marathons, and that's not something you would want to copy deliberately.

Halloween Jack

TFBubendorfer wrote: »

You're looking at it from the wrong angle. Of course once you finished in a time that makes you happy it doesn't make a difference if you paced by feel or watch.

The point is that you're more likely to succeed if you have a good feeling for your effort.

As for your example, if I check my watch every mile or so for the first 5 miles, I can be very close to 7:00 pace if that's my target pace. After that I am tuned into the effort and I can indeed tell if it's 7:00 or 7:10 - the difference between the 2 is easily big enough to notice. After that, an occasional glance at the watch is more for reassurance than anything else.

Btw, that's only a minor side argument to the original point, namely the value of this "Big Data" approach. The far bigger concern is that trying to emulate the average marathon runner's pacing strategy is just wrong. Most runners are piss poor at pacing marathons, and that's not something you would want to copy deliberately.

Impressive if you can tell the difference for those narrow margins, I certainly couldn’t. I would agree that attempting to emulate a flawed pacing strategy is kinda futile.

skyblue46

Swashbuckler wrote: »

I think you can to be honest. I think you can spend a lot of time in training being a slave to the watch to a certain degree as long as you constantly tune in to how you feel. Eventually over time you will become less reliant on the watch as you'll just know by how you feel. It's a hard hard thing to learn though. And most people (myself included) struggle with it.

I'm talking more about sessions. I didn't make that clear, sorry. I find I can run in a way that feels very easy, easy, moderate etc but those paces would vary day to day. As for racing, I'll have to try one by feel.

Safiri

Halloween Jack wrote: »

What difference does it make, if you pace it properly and finish in the time you wanted, whether you paced by feel or by your watch? And I’d suggest it’s pretty easy to know by feel what roughly 7 mins per mile is for example, but can you effectively ‘feel’ the difference between 7 and 7:10? Cause over 26 miles or more that 10 makes a sizeable difference

It's not being able to tell the difference between 7 minute and 7:10 pace that is the goal of running without a watch, you are still a slave to pace if you think like that. It's about learning to feel an effort, not a pace although once you key in to effort for long enough; you become a lot better at judging that too as a result.

People can be so ingrained with a watch that even when it's taken off, they still try to think in terms of pace and not effort which is what you want to think about. That's the goal of running watchless and it helps in so many ways. Pace is a man made which is rigid and our bodies don't understand it. Our body ebbs and flows and is constantly changing from day to day whether that's from tiredness, stress, wind, outside temps, humidity and even down what you had for breakfast and a million other things. A watch or pace can't predict how you are feeling but effort can as it's your bodies feedback mechasism. Theres an old saying "listen to your body, don't be a blind an deaf tenant" which is hugely applicable to running.

It's a key skill in any endurance sport as nothing is going to be perfect in racing and training. There will be hills, headwinds and tailwinds etc. etc etc that all have an effect along with all the other things listed above which will determine how fast you go. Running by effort is by far the best measure of how to deal with these. Learning to run by effort will make you a better and more successful runner in both training and racing because you learn to listen to your body and when it's time to back off and when to push on in both training and racing and if you are training and racing better, that's has a hell of a double whammy effect on performance.

ger664

Halloween Jack wrote: »

Impressive if you can tell the difference for those narrow margins, I certainly couldn’t. I would agree that attempting to emulate a flawed pacing strategy is kinda futile.

I think how you feel at a given pace is very important in pacing a marathon. If your intended target is as TFB said is 7:00 min/mile. If it does not feel easy or reasonably comfortable at the start of a marathon then you wont sustain it throughout. You may have been training comfortable at this pace but on race day conditions or runner may have something amiss can mean readjusting your target on the day and not been a slave to the watch.

The broader point of getting data on how to pace a marathon here is ridiculous. Most runners in the field would be completing their first Marathon off a very low base. I would suspect that the norm would pretty large positive splits.

ultrapercy

Halloween Jack wrote: »

Impressive if you can tell the difference for those narrow margins, I certainly couldn’t. I would agree that attempting to emulate a flawed pacing strategy is kinda futile.

Its not that impressive(sorry Thomas) to gauge difference of seconds at race pace. 10 seconds might not feel much different at easy run pace like marathon pace plus 90 sec but Id notice even a couple of seconds on marathon pace or quicker. 10 seconds is like going up 2 gears in a car.

Halloween Jack

ultrapercy wrote: »

Its not that impressive(sorry Thomas) to gauge difference of seconds at race pace. 10 seconds might not feel much different at easy run pace like marathon pace plus 90 sec but Id notice even a couple of seconds on marathon pace or quicker. 10 seconds is like going up 2 gears in a car.

Maybe at the thinner end of the wedge than where I’m at

Halloween Jack

Safiri wrote: »

It's not being able to tell the difference between 7 minute and 7:10 pace that is the goal of running without a watch, you are still a slave to pace if you think like that. It's about learning to feel an effort, not a pace although once you key in to effort for long enough; you become a lot better at judging that too as a result.

People can be so ingrained with a watch that even when it's taken off, they still try to think in terms of pace and not effort which is what you want to think about. That's the goal of running watchless and it helps in so many ways. Pace is a man made which is rigid and our bodies don't understand it. Our body ebbs and flows and is constantly changing from day to day whether that's from tiredness, stress, wind, outside temps, humidity and even down what you had for breakfast and a million other things. A watch or pace can't predict how you are feeling but effort can as it's your bodies feedback mechasism. Theres an old saying "listen to your body, don't be a blind an deaf tenant" which is hugely applicable to running.

It's a key skill in any endurance sport as nothing is going to be perfect in racing and training. There will be hills, headwinds and tailwinds etc. etc etc that all have an effect along with all the other things listed above which will determine how fast you go. Running by effort is by far the best measure of how to deal with these. Learning to run by effort will make you a better and more successful runner in both training and racing because you learn to listen to your body and when it's time to back off and when to push on in both training and racing and if you are training and racing better, that's has a hell of a double whammy effect on performance.

Many thanks for the explanation

TFBubendorfer

ultrapercy wrote: »

Its not that impressive(sorry Thomas) to gauge difference of seconds at race pace. 10 seconds might not feel much different at easy run pace like marathon pace plus 90 sec but Id notice even a couple of seconds on marathon pace or quicker. 10 seconds is like going up 2 gears in a car.

I know it's not that impressive, I completely agree with you. 10 seconds per mile difference at race pace is quite a lot.

Safiri's point about feeling the effort rather than the pace is totally on the mark though, and I didn't feel the need to add anything in that regards.