electro's horse

He's probably looking somewhere that takes away rushing yards for sacks or something. To continue this "pedantic about data science" rant, not sure how you adjust for Bryce getting a coach specifically to craft an offense for him vs. Cam's offensive coordinator getting fired from Alabama for being too conservative and is also about to be found dead in a creek outside Columbia.

I'm going to humor you in case you're actually being sincere and are just ignorant. Also, you're not very good at tossing barbs. Please stop. I did not say random; I said arbitrary. That is a much different thing. You're comparing QBR in starts across multiple coaching staffs, seasons, offensive personnel, against defenses with wildly variable of levels of competency, and don't really have a lot of data points. Choosing the dividing line at pre and post benching is arbitrary, and it's not even consistent. He was benched in his first season too. Why not when Reich was fired? Or when Canales was hired? Or when they drafted XL? Also, what are potential lurking variables here? I would if I was grading this tell you that you're not really showing anything useful. If your argument is that his QBR has trended up on a per game average from "this point" on then congrats yes you're correct. This is called cherry picking data. I understand you're trying to make this quick and dirty, but QB rating by itself doesn't really paint a useful picture. A simple thing you could do it do a second line along the same time domains of what the defense ranked league wide at the time they played each other. I would try to use a cumulative stat from the same source, tho. So qb rating v scoring defense across these specific truncated set of games that have occurred across three seasons with multiple coaching staff. Some other suggestions for you 1. This is the wrong type of graph. You want a scatter point. 2. The range for QB rating improvement is like 15 points or whatever. The y axis is broken up into 25s. How can I quantify the line? 3. Quarterback rating goes to 158.3. Why does your line only go to 125? The only thing I would really take from this is that Bryce Young sucks in a way that QB rating doesn't pick up on. This is what you'd call a "problem with the data." Lastly, I would point out that QB rating as a formula was finalized in 1973. Offenses were so anemic back then it's referred to as the "dead ball era." Maybe not the best metric to measure QB success half a century later. like buddy i've taken masters level data science courses, and believe me I've done some shitty fuging work. my masters thesis ended up being about why my method of data collection didn't work well enough for me to defend my thesis. I wrote 100 pages just kinda shrugging and saying "Well ya know covid is so challenging etc." You're trying to tell people their eyes are lying to them and are posting a very bad graph to do so. Don't be upset when people not only argue with you but point out how bad your graph is.

No, his qb rating is trending up. You have two variables; qbr and single game performances over the parts of two seasons with an arbitrary break. you’ve chosen one way to quantify performance and showed a statistically insignificant change across an arbitrary time domain. and you’ve made it difficult to interpret because you don’t know how to make a graph. Like why is the y axis broken down by 25 with data points that variable this isn’t difficult

Okay yeah you’re trolling

Lmao you did this manually and still made it a bar chart. What’s the confidence interval nom this big guy? Do you know what button does that? powerful “information sciences is my passion” energy p

so 16% Would you hold your breath if his winning percentage reached the lofty heights of 27%

7 points is such a damn sucker bet

See if we follow this trend line in about 3 seasons he’ll have a qb rating over 120

Before the Penix pick six, the panthers had run 8 plays for seven yards while up 10-0. That includes a 5 play, 13 yard drive out of halftime where Bryce threw incomplete on third down. In that exactly what canales wanted?

Why isn’t this a labeled scatter plot. The y axis is useless. Post benching occurs across multiple seasons, players, etc I strongly suspect this isn’t statistically significant but it’s impossible to calculate because a child did this graph. Or an idiot asked chat gpt to do it for him and has never actually done a regression analysis

You’re talking about the 2022 SEC male athlete of the year put some respect on his name

Without looking, defenders requesting a trade from The bengals usually means they’re about to be tagged or extended.

Not in this defense. (also honestly I always think we drafted him for some reason)

Come on now.

I think it depends on how the bills game go. the nfc south is winnable, especially with the Bucs roster rotting on the side of the road.