Moneyball?

 Hello loyal readers,

I've been discussing this article on Row2K on coxswain evaluation and using analytics to decide seat racing between coxswains. 

It's an interesting read.

Let me get this out of the way first. I appreciate the effort. I always appreciate it when a coach or a program looks at the status quo and asks, "Can we do this better?" That is a question that is not asked enough. Too often programs simply recycle what's been done to death because, "This is how we've always done it," or, "This is how I was taught..." and it's boring. I've long pushed for more accurate evaluations when it comes to coxswains - better training as well - likening most coaches choosing coxswains to that of building a high performance sports car only to decide that the type of engine doesn't matter, or purchasing a vehicle based on whether or not it comes with wheels and tires. 

With that being said, I find this article to be nothing more than fluff. It lacks substance, and the arguments or approach pushed can be blown apart rather easily.

The theory behind Moneyball, without getting into the nitty gritty, is to evaluate talent based on things that matter, otherwise known as, what contributes to winning. It was an attempt to get away from the nonsense that baseball scouts had been evaluating talent by - swing, pop, baseball body, etc. - and get into what truly mattered: runners on base, scoring runs, winning. What truly contributed to a players value? 

Now, in the movie - my favorite movie - Jonah Hill's character - who does not exist in real life, but is instead a mix of several real life individuals - says to Brad Pitt playing Billy Beane, "It's about getting it down to a number."

And that is where I find this article. Let's take coxswain evaluation and bring it down to a single number. And let's identify the one aspect of coxing that the author believes is the singular aspect that defines success or failure.

Steering.

Everyone is going to have their own opinion on what it takes to be a successful coxswain, and this individual chose steering. Fair enough. I completely disagree, because what I value in a coxswain cannot be taught whereas steering, race planning and strategy, etc. can all be - and should be - taught. I evaluate based on personality. 

The author talks about losing a race over 2000 meters by a bowball, and using this argument, determines that steering truly matters.

Here's the first of many problems I have with this.

If you lose by a bowball, there is no way you will be able to prove to me that the coxswain was responsible. You might take the GPS - as they did - and show me that the course deviated from the "ideal" but my response would be, "Show me the power curves of each rower, as well as the pick up/drop in rate and speed from the boat for the entire length of the race." Any deviation from any rower at any part of their stroke, and you'd suddenly have a serious variable on your hands that you'd have to explain away. 

It'd be a chicken and the egg situation. Did the steering contribute to the deviation in power application, or did the deviation in power application impact the steering? You'd have to line up the course GPS with the power curves. The only way to know is to then trace the course and curves the entire length and see which deviated first, and by how much, and how that impacted the run of the boat. 

I can get into the maths, and science, but I won't because this would then become a lengthy counter argument. 

I'll tell you the issue that I have with this both from the specifics of the examples given, and generally.

Specifics: MAC rows out of the LA Harbor. Salt water. It's highly suspect water. Balboa Creek where they showed some testing is also salt water and subject to tidal impact. It's a very straight course, but it empties into the ocean and is not protected from the impact of tidal pull. The harbor has boats moving in and out of it. 

Generally: Wind, weather, water conditions and temps. rigging, experience in the boat, rates, etc. There are a myriad of factors that go into impacting hull speed and the course. Athletes are a major factor. 

These are not factors that can be easily dismissed - as done in the article - or limited to a number. You can, but your formula begins to resemble a long, "If/then" statement with fewer numbers and more questions, or arbitrary numbers. 

Wind speed from low to high: 1-5

Water temp: 1-5

Athlete experience: 1-5

You see that suddenly we've got some issues that must be accounted for. You can't dismiss them because if you do, then a coxswain who drove the boat with the tide is now faster, and the coxswain against the tide is slow. Afternoon evaluations when the water has had time to warm up in the California sun vs. the temp of the water in the morning. Salt water in the harbor vs. fresh water at Lake Harsha. 

The purpose behind Moneyball is to eliminate this. The purpose behind most evaluations is, or should be to eliminate or account for as many variables as possible so that what is being tested is truly what or specifically whom is impacting hull speed.

That's why seat racing is a pain. Conditions must be the same. Lineups must be the same. Courses must be the same. 

I appreciate what this person is trying to do. I always appreciate looking at the status quo and asking if you can apply science and make it better.

But the science and reasoning that I'm seeing here is the same to me as those health "experts" who are on a company payroll telling you that the product the company is putting out is the best product on the market. It's the same "science" behind the blogger, "Foodbabe" who was sponsored and paid to endorse or dismiss products. This isn't scientific. It's arbitrary numbers that are dressed up with formulas designed to look like science.

A 2K course lane is roughly 12' in width. That accounts for about a foot or so on either side of the blade tips which are roughly 9.5-10.5 in length (overall length of 12') depending on inboard and general length. 

The deviation a coxswain is going to make within a racing lane is minimal. The steering done by an experienced coxswain in a race lane is going to be negligible because there's nowhere to go. An inexperienced coxswain may steer all over the place but at a selected camp, you're not finding a novice.

Again, anything under a length, and there's factors that must be considered and that will have at least as equal or more say in the loss. 

And finally, we come to the question of head racing. Moneyball, or the Moneyball theory applies across the board. This example only applies to a straight course. It cannot account for the turns seen in most head racing situations. Yes, there is an "ideal" course at any head race, but simply sticking to the ideal course doesn't guarantee a victory, while going outside of that course doesn't guarantee a loss.

This is an interesting idea. I like the thinking behind it. But the variables were not accounted for. The reasoning is weak and easily disproved or questioned without a legit response, or proof of concept - winning a race doesn't prove the concept.

Again, there are many factors that go into winning and losing a race. Trying to narrow it down to a singular reason ignores too many questions and ultimately leaves you with more questions than answers.

An "A" for thinking. A massive failure for execution and reasoning.