Isn’t that often times the irony of analytics? The most fervent devotees fancy themselves as capable of seeing the game in a highly objective fashion and yet some of the metrics they lean on are based off of a foundation of subjectivity.
I fully understand what you are saying. But take a closer look at the names and notice how “bottom heavy” the top runningbacks are on the list.
I think this is a case where someone put together some faulty data. Calling it “analytics” doesn’t suddenly make faulty data correct. My gut instinct is that the model is built by someone who doesn’t know football and only tries to graph a mathematical interpretation of something that you actually need an understanding of football to have a grasp on. And if the model featured in this article is what was used, its only a couple of years old and was created by guys who are from another country and have no clue what American Football really is. Sometimes that can be a plus. In this case I think it created a faulty model.
Analytics are such a much more natural fit in most aspects of baseball. The various methodologies for analyzing fielding though tend to suffer from the same issues that football and basketball do at times.
Well, Monty fared better than JW according to this, but he’s also to the negative. You would’t think a guy who breaks a lot of tackles would be one of those (supposedly) underperforming.
I wouldn’t call it subjectivity…I mean they’re trying to model something that is more complex than what they capture in the model. I’d call it incomplete rather than subjective. Its not like judging who is the most handsome RB – that would be subjective. The things they are putting in the model - e.g. down and distance, formation, tracking data, all make sense and are objective measures. These objective measures are just not capable of capturing something useful right now. it will improve over time. By how much? That’s the key question.
Jamall was a straight ahead battering ram. Which was great on goal line but not great for long runs obviously. I am curious to see how the run game progresses.
I am truly impressed with Holmes and crew. The stacking of solid draft picks and free agents has really started to change this team. KC is literally the only team on schedule that I think should beat the Lions everyone else I expect the Lions to be favored. Crazy
The names are a red flag, but only part of the story. Go read how they come up with the data. Its not based on football. Its just an exercise in GPS data at the time the ball was handed off.
Here is another year of the data (2020). Notice the line of expectation, which basically says that Derrick Henry is just getting the yards he was supposed to get. Nothing more. This was during his 2k yard season. When I tell you Derrick Henry is only getting the yards he is supposed to get and that’s it…does that pass the smell test for you?
I will put this in a separate post. I love it when “analytics” crushes conventional wisdom and gets us outside of our comfort zone. But we also have to be able to see when we aren’t getting what we think we are getting out of the exercise. There is a difference between data and information. Data is just that…data. Information is what we do with the data. In this case the data itself is useful, but the way its being used and interpreted is flawed.