Crankworx and Optimization

CRANKWORX AND OPTIMIZATION.

 

Yeaaah baaaaaabbbyyyy!!! I got a podium at the first race of the year! 

The first race is intense. We spend five months in our caves trying to get stronger, faster, and tougher, hoping to somehow outpace the rest of the field. If we’ve done our jobs right, we come into to the first race with the confidence that we prepared the absolute best we could and gave 100% at every opportunity. It’s not until the weekend is over that we get to honestly evaluate how the off-season went. Luckily, I think mine went pretty damn good.

I can’t explain how awesome it feels to not only have all of your hard work pay off but to also have so many of the hard decisions you made turn out to be the right ones. It’s been fascinating to look back on how I’ve felt after I get a good result. Interestingly enough, this is the least confused I’ve ever been haha! Crankworx went great, but the more important thing is that I’ve made such a huge leap forward in the way I go about improving.

Finding a viable target to improve upon is something that requires constant re-evaluation, self-reflection, creativity, and open-mindedness. Finding the right process to achieve that improvement requires acute attention to detail, an empirical mindset and accepting “good enough.”

I’m excited to share my thought process for this year and to give you a little bit of the theory behind why Jarrett and I do what we do.

 

THE TARGET VARIABLES

In every system there are bottlenecks. One or more elements that put a cap on the performance of everything else. At the end of last year, we identified a few areas that, if improved, would increase the potential of the system exponentially. They were:

  1. Race Week Endurance
  2. Technical and Wet Riding
  3. My Attitude at the Races

Knowing these were the areas I needed to work on, I wanted to start from scratch. I began with the following questions:

  • What if my race week endurance isn’t due to my training off the bike?
  • What if technical and wet riding isn’t due to my training on the bike?
  • What if my attitude at the races isn’t due to my attitude at all?

Like I talked about in a previous post, so many times, we come at a problem from the most ‘logical’ point of view and most of the time that ‘logical’ point of view is not the most ‘logical’ at all. It’s just the way we want to solve it.

 

TARGET 1: RACE WEEK ENDURANCE

I know my training off the bike is great. Jarrett is one of the most thorough, passionate, and smart people I have ever met. I have no doubt that what we do is better than everyone else. He is one of the reasons that made me want to come at this from a different angle. If I know my off the bike training is amazing, and I do; maybe I’m not doing everything I can on the bike. He is just one tool in my tool belt, and it’s not very efficient to use one tool for everything.

We talked about what we could give up, what needed to stay, what we could supplement, what we could replace, what I needed, and what he needed. Interestingly enough, what we came up with wasn’t more, but less. It was the bare minimum and pretty much the opposite of everyone else because we knew what we wanted, there was a focus. There was a goal for everything, and once that goal was hit, we would re-asses and see if it needed to be raised or maintained. We tried to incorporate as many problems into the plan as we could. Addressing some issues I’ve had from the beginning; like not being able to get into as good of a flow as I did on my motorcycle, reducing the variance of my times from timed practice to finals, not over-riding during a race run, etc., etc. These are challenges that everyone faces at some point; the important distinction was to ask why I was having these problems and how we were going to fix them

I won’t get into all of the details, but since January I’ve done over 400 runs and every single one of those runs was working toward a specific goal.

 

TARGET 2: TECHNICAL AND WET RIDING

This is another one that I decided to re-evaluate after last year.

Even though I live in California, I’ve spent most of my time riding in different parts of the world. Including some very very wet ones. I realized that I needed more work than I thought after the first three races of last year. Lourdes ended with a crash in qualifying over the one 50 foot stretch of wet roots on the whole track. Fort William was the same. Only that section was about 25 feet. Leogang, a track that I usually enjoy, was the worst race of the year for me. I struggled with the technical bit and it ruined the weekend for me.

After this, I decided to go to Morzine, France for a few weeks and ride the gnarliest stuff I could find, especially if it was wet. I found the most difficult off-camber root sections I could and pushed up to hit them 20-30 times a day. I also did a ton of runs while I was there and realized that I need to ride a lot. A whole lot. Comfort and speed on the bike are things that, for me, come with seat time. Downhill bike seat time. It was important for me to realize and accept that.

This point in the year was actually an incredible tipping point for me. I came away saying three things: If I could be this happy at the races, I would do amazing. If I could feel this good on my bike at the beginning of the year, I would do amazing. If my bike could feel this good all the time, I would do amazing. Ironically, or not, these three things were the start of the targets, and the questions I asked going into the 2017 season. I rode a lot with Luca and Greg, and I stayed with Marcelo. I just had such an incredible time. Here were three top 5 riders doing things completely differently. Having so much fun. Trying super hard. Enjoying life. It was such an eye-opening experience for me based on the bubble I had been in. I wanted to feel like that all the time.

Greg helped me set up my bike and after that, I think I maybe did one or two clicks of adjustment for the rest of the year. It gave me an inkling that this was a bigger issue than I thought. No matter how much I practiced, I couldn’t even begin to do some of the lines Marcelo was able to do, (and still can’t haha!) his bike just seemed to stick. It wasn’t until I got on my bike this year that I realized how much better my bike could have been. I was able to do things that were impossible before. It played to my weaknesses instead of my strengths. We’ve done some pretty special stuff with the bike this year. Dave and I have put a ton of hours into making sure our machine is ready for the first round and even though we still have work to do, this bike does things that I didn’t even know were possible.

The most straight forward answer for not being good at riding wet tracks is to get better at riding wet tracks. For me, that wasn’t the answer at all. It was everything but that. At Crankworx I was never nervous or scared about the track being wet, I never lost any confidence or speed. I felt amazing and felt like I could ride it like it was dry, despite it being one of the wettest races I had done in more than nine months.

 

TARGET 3: ATTITUDE AT THE RACES

One of my most valuable assets is also one of my biggest weaknesses. I love being analytical and thinking deeply, but sometimes it’s detrimental. The biggest takeaway from 2016 was to not analyze my race runs as I was racing. It’s all fine and good to look back and think about what could have gone better after the run is done, but doing it during the run is like starting to train on the plane. There is no more work to be done, no more gains to be made, we’re in the thick of it. If you’re not ready by now, then you’re not ready.

Realizing I needed an environment that promotes being less analytical was also a takeaway from Morzine. These were the hardest changes to make. I had to make adjustments to how I thought on a fundamental level, in addition to taking a huge risk by looking for a new team. The latter gave me a new respect and perspective on what I do, why I do it and the people I surround myself with. I’ve talked about it before, but it is an experience I will never forget. It taught me so many great lessons about myself and life in general. It also taught me that if you want something you have to go for it. No matter what people say.

 

THE OPTIMIZATION PROCESS

During the process, we sometimes get caught up with making everything better without keeping our target in mind. I have to hit these numbers in the gym, or I need to be able to produce these watts or ride like this person. Without stopping to think “What problem is this going to solve?” These pursuits could be secondary to our primary target, and if we go too far, it can actually hurt the performance of the target, if not directly, then indirectly in the form of opportunity cost.

Improvement is a logarithmic function. When we start a new activity, one hour of time may give us a 10% improvement. However, as we get better and better, that same hour may only yield 1% or .1% or even .001% depending on what level we are at. If we treat all of our attributes the same and allocate an equal number of hours to developing each one, we can be sure that we are not being efficient with our time.

Unfortunately, there aren’t an unlimited amount of hours in the day, so we have to figure out how much time to allocate to each aspect. The better we are at something the less time it needs. In this respect, it’s just as important to figure out our strengths as it is to figure out our weaknesses. Once we have an idea of each end of the spectrum, it is essential to isolate them as best we can. Taking myself as an example: While I was in Queenstown, out of the 250 runs I did, about 15 were done on the jump trail. That’s six percent. 94% of the time I was riding something else because my jumping skills don’t need to be improved (at least not until everything else catches up) they only need to be maintained. Any extra time spent over that is time wasted.

The interesting thing about great targets is they complement each other. The further Jarrett and I got, the more we started to notice how interconnected all of the targets we set were. Race week fitness was helped by run repeatability, which also helped comfort of the bike, which helped comfort on steep and wet tracks, which was also helped by being on a bike that suited me more, that was caused by being on a new team, which helped my environment at the races, which gave me more confidence, which helped with riding wet tracks, etc., etc.

 

THE PROCESS

So how do we optimize something? The goal of the process is to find a setting, or parameter value, that gives us the best possible outcome for our target.

In reality, there are three main ways we try to optimize things:

  • Brute force
  • Hill Climbing (or Gradient Decent, looking for the lowest point)
  • Simulated Annealing

Let’s take setting the compression adjustment on a fork as an example.

Brute force would be trying every possible setting. The underlying idea is that we have no idea what the best setting is, where to start or why we would start in a certain spot. This can sometimes be feasible if our dimensionality is low, but can it quickly get out of hand. A fork with 10 clicks would only take 10 runs to try every combination, but if we add in fork rebound and shock compression and rebound, then we have 10^4 combinations or 10,000 runs of adjustment. In the real world, wanting to use a brute force approach is also a good indicator that we don’t have a clear goal or good understanding of the underlying mechanism. Something we want to avoid

Hill climbing is the process of only looking for improvement. If we begin at our starting point and go a click to the right, we don’t find any improvement, so we reset and go a click to the left. That offers an improvement, so we keep going. Eventually, we get to the Local Maximum, say 5 clicks, and go one step past to try to keep finding improvement. We find that we didn’t increase our target variable, so we declare the optimal value at 5 clicks. The downside to this approach is while it’s fast, you also run the risk of not finding the Global Maximum. It’s greedy. This is the approach we use most. We pick an arbitrary starting point (middle of the suspension dials) and hill climb our way to some maximum value. We never know if we have truly found the maximum because we haven’t tried every combination, but we have found a maximum, and that is usually good enough for us.

 

 

 

 

 

 

Simulated Annealing is the same as hill climbing, but once we find our maximum, we save that value and start again at a random spot. For example, our first random start begins at the very right and we end up there. The second one starts at the Global Minimum and we have an equal chance of ending at the right or the Local Maximum. We do a third random start somewhere in-between the Local Minimum and Global Maximum, but this time we end up finding the Global Maximum. This isn’t a foolproof situation, either. If we decide on 3 random starts and they all start to the right of the Local Maximum, then we still at best only find the Local Maximum. Each restart increases the probability of finding the Global Maximum at the cost of increasing our optimization time

 

Application

It’s important to understand these search algorithms because then we can make informed decisions on the tradeoffs between time and thoroughness. We can also use domain knowledge to help guide us to the optimum value. To give you an example: Just a few weeks ago Dave and I used a form of simulated annealing to find a setting for the front end. We were looking for a combination of increased front ride height and increased entry steering. By the end of the day, we found something that seemed to work really well. I pushed that setup as hard as I could and I got a 1:28. We went somewhere else the next day and realized a few of the implications of what we did. Day three we started off in a totally different point and began to work on the same problem. I got a 1:33, then a 1:30, then a 1:29, and finally a 1:26. We found another maximum that was higher than what we had before, but we had to start in a valley and hill climb our way out of it

I think it’s a bit easier to think of searching in this way when we have a question like “How many clicks of compression do I need?” It gets a bit more difficult when we ask things like “How many watts do I need to produce over 4 minutes for downhill?” or “Why do I not race as good as I practice?” It’s not that these questions are harder to answer, it’s the fact that were not used to distilling them down into a single target.

Race week endurance was a good example because there were multiple times where we were maintaining a few parameters, but after we took away our bottleneck, we realized those needed to be improved again. It’s a never ending process. Hill climb our way up hand strength, maintain watts, hands strength is good, realize watts could go up, hill climb watts, maintain mobility, etc, etc. A typical argument to this is, if you’re going to keep improving everything anyway, why ever stop to maintain? Two things: First we don’t know if we will ever have to improve that variable again. When I first started my leg speed was horrendous, we worked on it and I haven’t had to touch it in 6 years. Second, like I was saying before, we only have a limited amount of time. We don’t have the luxury to guess that increased leg speed will yield better results, we need to maintain until a later time where we actually need that leg speed for a specific purpose.

 

Conclusion

I went a bit into the theory of improvement here, but it’s important to understand and remember that all we are really doing is setting goals and achieving them. As we get better and better, our time becomes more and more valuable, and it becomes more and more important to be smart about what we are doing. The biggest takeaway from this off-season for me is to never get stuck doing or listening to what has worked for others just because. It’s so easy to look at the people winning, or with a lot of experience, or listen to what we are “supposed” to do and somehow forget that we have different life experiences, different skillsets, and different strengths and weaknesses. We are different people and we can never be scared to do things differently because of some social pressure.

Improvement is about being dynamic and analyzing one’s self as an individual. It’s about never thinking you have figured it out. It’s about taking risks and being unconventional, and above all, it’s about never doubting what you truly need.

I know this season will be my best season yet because I know I took every opportunity to try my hardest, be better, and do what was best for me!

See you soon!

 

Leave a Reply

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