August 28, 20223 yr I'm not sure it matters whether you like or dislike Hurts at this point. It's clear he's the best chance we got at a successful season. I like Minshew but I didn't see enough this pre-season to overtake Hurts in my mind. The small improvements Hurts made may go a long way.
August 28, 20223 yr Author Just now, downundermike said: You are also wrong, you can not take the average of QB ratings and have that be the QB rating for that stretch of game. It is right here, these are the numbers from PFR, you know, the site you said I was arguing againts. Please tell us, if your method is correct, why is your answer 78.1 using the average of the 4 games, but PFR shows it correctly as 86.6. Because in that formula, his worst game counts less than the others. Think of it as a GPA, which a passer rating essentially is. Anything over 100 is an A, in the nineties is a B, in the 80s is a C, in the 60s a D, and anything below that is an F. Kurt had two Bs, a C and an F, which is a GPA of 1.75, or a D average. Which makes sense considering my ranking gave him a D. By making some games count more, which the formula did, his GPA was raised to a C, in the 80s. But average play didn't correlate to wins. He went 1-3. Which is below average. Similarly, in the last three games of the season, Hurts had two As and a B. Meanwhile, in the three games where you claimed he played better, he had two As and a C. That's the right answer. Passer rating is a formula that basically equates to letter grades in school. That's the way it was designed and the only correct way for it to be used.
August 28, 20223 yr For those who do not want to read, here are some examples in the gif form, showing what I have done to jsb today.
August 28, 20223 yr 10 minutes ago, jsb235 said: Because in that formula, his worst game counts less than the others. Think of it as a GPA, which a passer rating essentially is. Anything over 100 is an A, in the nineties is a B, in the 80s is a C, in the 60s a D, and anything below that is an F. Kurt had two Bs, a C and an F, which is a GPA of 1.75, or a D average. Which makes sense considering my ranking gave him a D. By making some games count more, which the formula did, his GPA was raised to a C, in the 80s. But average play didn't correlate to wins. He went 1-3. Which is below average. Similarly, in the last three games of the season, Hurts had two As and a B. Meanwhile, in the three games where you claimed he played better, he had two As and a C. That's the right answer. Passer rating is a formula that basically equates to letter grades in school. That's the way it was designed and the only correct way for it to be used. False. This is how the NFL calculates QB rating. You are 100% wrong. If you don't like it, have your lawyer call the NFL league office. Out of curiosity, is this when I am supposed to quit the board ?? I 100% call BS on you making anyone quit, I have absolutely wrecked you in here. If you had any pride left, you would deleted this thread and quit the board yourself.
August 28, 20223 yr Author You were right by the virtue of a rounding error. Three tenths of a point. Savor it. Maybe cut it into tiny pieces to make it last.
August 28, 20223 yr 5 minutes ago, jsb235 said: You were right by the virtue of a rounding error. Three tenths of a point. Savor it. Maybe cut it into tiny pieces to make it last. It does not matter how may times you say it, your rounding error excuse does not even apply. That is not how you calculate QB rating for a stretch of games. I have proven how the NFL actually works in the example below. If you do not like it, I have already given you the phone number for NFL legal in a previous post. Fact, using the method the NFL uses to calculate QB rating, it is not rounding, it is a 102.62 to 101.37 victory for me. GPA's, how you think a game should count less because of your feelings are irrelevant. Last time, please explain how you are 8.5 points off using your method vs how the NFL does it. Do not bring irrelevant topics into it, use this data and explain. 1 hour ago, downundermike said: Correct. Here is the example I gave last night that made him decide he was going to start a thread that was going to make me quit board. Kirk Cousins has played in 4 career playoff games, in those games, he has a combine QB rating of 86.6. 40 + 91.7 + 96.4 + 84 = 78.1. 87.1 / 4 = 78.1. ( This is your method @jsb235, if you could hold your response until I address the main post ) https://www.pro-football-reference.com/players/C/CousKi00/gamelog/post/
August 28, 20223 yr 39 minutes ago, EagleJoe8 said: You definitely got more in depth than I did, but it all seems to add up, (or not add up in JSB's case). When I wrote my reply earlier, it seemed as simple as tallying up a QB's ratings for each game, and seeing if the listed yearly rating equals the average rating doing it his way. Hurts' numbers were the closest, but the others seemed more off. Doing it his way should either match exact, or it doesn't, and his way didn't. Correct. I ran 14 QB's playoff game logs last night, obviously Kirk Cousins being the largest difference, but I found 9 other QB's that the difference was more than 5 points.
August 28, 20223 yr Author 1 minute ago, downundermike said: It does not matter how may times you say it, your rounding error excuse does not even apply. That is not how you calculate QB rating for a stretch of games. I have proven how the NFL actually works in the example below. If you do not like it, I have already given you the phone number for NFL legal in a previous post. Fact, using the method the NFL uses to calculate QB rating, it is not rounding, it is a 102.62 to 101.37 victory for me. GPA's, how you think a game should count less because of your feelings are irrelevant. Last time, please explain how you are 8.5 points off using your method vs how the NFL does it. Do not bring irrelevant topics into it, use this data and explain. Getting the most out of this three tenths of a point win, aren't we? Are you feeling a little dumb that you were right only on a technicality? Is it that unsatisfying?
August 28, 20223 yr 45 minutes ago, downundermike said: @HazletonEagle @EagleJoe8 @4for4EaglesNest @Swoop @Texas Eagle @LeanMeanGM @Bwestbrook36 @DeathByEagle @ToastJenkins @Rob331 I think I more than made my point, but you guys can judge. Your report is on Julian’s desk. The analytical team will bet back to you tomorrow.
August 28, 20223 yr Author Maybe you can change your name to three-tenthsmike in honor of this epic win. Seems like the only right thing to do.
August 28, 20223 yr 1 minute ago, jsb235 said: Getting the most out of this three tenths of a point win, aren't we? Are you feeling a little dumb that you were right only on a technicality? Is it that unsatisfying? It is a 102.62 to 101.37 victory for me. Math really is a struggle for you. 102.62 - 101.37 = 1.25
August 28, 20223 yr Author 1 minute ago, downundermike said: It is a 102.62 to 101.37 victory for me. Math really is a struggle for you. 102.62 - 101.37 = 1.25 Mmmm, savor it. The high point of your week/month/year probably.
August 28, 20223 yr Can we run a t-test on these groups of data? Im curious if Mike's incorrectly figured 3/10ths would still be significant. And if his actual 1.25 is statistically significant.
August 28, 20223 yr 1 minute ago, jsb235 said: Mmmm, savor it. The high point of your week/month/year probably. But you started this thread to own me and make me quit the board.
August 28, 20223 yr Author 2 minutes ago, downundermike said: But you started this thread to own me and make me quit the board. The three-tenthsmike nickname made it worth it.
August 28, 20223 yr 7 minutes ago, jsb235 said: The three-tenthsmike nickname made it worth it. Intellectually dishonest data. You have also proven you don’t understand how math and statistics actually work.
August 28, 20223 yr 51 minutes ago, jsb235 said: Getting the most out of this three tenths of a point win, aren't we? Are you feeling a little dumb that you were right only on a technicality? Is it that unsatisfying?
August 28, 20223 yr Did this get solved yet? IMO, it's perfectly fine to to compute the raw stats from a 3-game stretch to get a QB rating for that 3 game stretch. What I don't understand is why this is important to validating an opinion over what 3 game stretch in more impressive. Opinions will vary. Both opinions should be able to be supported with valid arguments.
August 28, 20223 yr Just now, brkmsn said: Did this get solved yet? IMO, it's perfectly fine to to compute the raw stats from a 3-game stretch to get a QB rating for that 3 game stretch. What I don't understand is why this is important to validating an opinion over what 3 game stretch in more impressive. Opinions will vary. Both opinions should be able to be supported with valid arguments. It was not an opinion argument, it was a fact argument. jsb stated as fact that his stretch had a higher QB rating, that is false. He then tried to argue how NFL QB rating is calculated, which your post says he was wrong.
August 28, 20223 yr 1 minute ago, brkmsn said: Did this get solved yet? IMO, it's perfectly fine to to compute the raw stats from a 3-game stretch to get a QB rating for that 3 game stretch. What I don't understand is why this is important to validating an opinion over what 3 game stretch in more impressive. Opinions will vary. Both opinions should be able to be supported with valid arguments. He used a specific NFL stat that has a formula. That's not an opinion at that point, but math. His math was wrong
Create an account or sign in to comment