Just over a month ago, I posted a piece on this site regarding TrainAsOne, and their AI based training plans. If you missed it, the original post can be found here. A month on from setting up my training plan, I thought I would post an update about how it is going.

I am still undecided on which autumn half marathon I’m going to enter myself into, but I set the training plan up for the Leicester half, which is due to take place on October 14. When I submitted my conditions to allow the system to compute my training plan, there were 18 weeks until race day. I completed the compulsory perceived effort, 6 minute assessment and 2 mile assessment runs and then expected to see a variety of speed sessions over the coming weeks.

For information and from my experience, aside from the three assessment runs mentioned above, TrainAsOne will build your training plan using four different run types. There is the economy run, threshold, interval and repetition. Economy runs are essentially easy running, intervals are generally 3 minutes long each and repetitions are 30 seconds long. Threshold workouts are slightly more variable, with intervals ranging from 6-12 minutes long each.

Anyway, following on from my assessment runs, I expected to see a variety of these within my training plan. Not so much intervals and reps, but definitely threshold runs, one of the key workouts for half marathon training. What I was presented with instead was a diary full of economy runs. Three runs per week, every one of them an economy run, for as long as the plan would allow me to see.

In truth, for beginners, this is what you would expect to see. One of the most important parts of any training plan is building the base. The beginning of the plan should involve miles and miles of easy running, before some threshold and speed sessions as the race gets closer. However, I am not a running beginner, this will be my eighth half marathon, and my second of the year. I already have a good base to start from and therefore should really be incorporating speed sessions in from early on in the training plan.

For the past month then, I have been doing nothing but economy runs. This is partly due to wanting to stick to the plan to see how it pans out and partly due to not wanting to push it too much due to a couple of niggling injuries that I’ve been carrying for a while. Today, we are 12 weeks away from race day. As part of the free subscription, the TrainAsOne system will let you see your scheduled workouts for the next five weeks. Looking ahead, I have more economy runs scheduled, with the first speed session ( a threshold workout) scheduled in for August 22, just over seven weeks away from the race. Not only that, but over the next three weeks, according to the schedule, I need to complete another 6 minute assessment, another 2 mile assessment and another perceived effort run. I’m struggling to see the point.

As I discussed in my previous post, I have used TAO before and seen some impressive results, but I believe that was due to me not knowing beforehand what was required to make myself a faster runner. TrainAsOne claim that they offer “award winning training plans for beginner to elite.” I would argue that this system is more for beginner runners. If you are an elite runner, advanced, or even intermediate (where I would place myself) you already know the type of speed sessions required to make yourself faster at each race distance. You already know how to manage your training so that you avoid overtraining. You do not need an AI system to suggest to you when it is time to push and when it is best to back off.

My aim was to continue using TAO until the speed sessions kicked in, and then make a decision from there as to how much benefit I was getting from it. Writing this piece has made me think otherwise. I feel like I would be wasting valuable training time, particularly if I proceeded with the upcoming assessment runs. I pushed it a little bit towards the end of this week and completed a nine mile run, my longest non-stop run since March, and I came through unscathed, without any pains. I therefore think it’s time to get in some speed work, and ditch the TAO training schedule.

If you are interested in TrainAsOne’s AI training plans, I have included a link below. I would suggest however, that it is more suited to beginners, or runners who do not already have speed sessions in their schedule. TAO offer a free subscription (which is the one that I have been using) and a paid subscription which costs £6 per month. The paid subscription offers premium benefits such as heart rate based workouts, advanced analysis and analytics, workouts adjusted for elevation and weather and the ability to see more of your plan into future weeks.


Categories: Blog


Chris Rees · April 28, 2019 at 08:37

Any update on this? As a coach, I cannot stress the need for time building base. Would be interested to know how you did and whether you stuck with the program or did indeed ditch it

    Paul Price · April 28, 2019 at 19:54

    I ditched it at the time, although I have tried to pick it up again since as I do like the idea of what it’s trying to achieve. For me though, there are a few too many issues. The benchmarking runs are too frequent, and provide inaccurate results, the algorithm that prevents you from overtraining is too conservative and the speed sessions aren’t varied enough. I tried a perceived effort run recently where it asked me to run at my natural easy pace. This is 09:00 – 09:30, which is what I ran at. Subsequently, my economy runs have been scheduled for 11:30 pace, and yet threshold is supposedly 07:05. Again, I really like the idea, but it doesn’t quite work for me

Leave a Reply

Your email address will not be published. Required fields are marked *