|
|
Post by afboltfan on Oct 10, 2022 18:30:04 GMT -7
Well there you go... Analytics was not a factor, just a dumb decision.
|
|
|
|
Post by cthommes on Oct 10, 2022 18:40:32 GMT -7
Ah, these last several posts make more sense to me -- in that it makes more sense to question the approach as opposed to the singular decision (which resulted in a win). I think the reason we need Staley (or whomever) and not computers to make all the decisions is simple, at least to my understanding -- the NFL doesn't allow for computers to be used in game in that capacity. I'd read somewhere that coaches who are analytics-centric need to have that stuff mapped out in advance. I.e. computers can't be used in real time to look up or calculate potential outcomes of particular situations. This topic is fascinating to me, as I'm a data driven individual. I am neither all-in in favor of nor against such an approach, but do love to try to understand it. Some interesting tidbits I've found, if anyone else is similarly interested (although it seems I'm alone, so talking to myself): theathletic.com/3674898/2022/10/09/la-chargers-analysis-week-5/"The win probability math varied significantly among different models on this decision. Ben Baldwin’s model had the decision inherently giving the Chargers 11.7 percent in added win probability. ESPN’s model had the decision adding 5.2 percent in win probability. Next Gen Stats’ model — which had the yards to gain at 1.7 versus just one — had the Chargers losing 0.2 percent in win probability with the decision." live.numberfire.com/nfl/8124After 4th down failure, Cleveland had a 55% chance to win. One big-picture thing I think you are missing as you consider the situation. The analytics are by definition a factor of what data is given to the almighty computer. There is simply no way that the guy doing the data-input can put in every single variable ahead of time to address the particular situation.Brissett being the QB, rather than Mahomes, is a good example. You would have to have info such as "what is the opposing QBs completion percentage". Or would you plug in "what is his QB rating by quarter?" Or "what is his percentage in the last 2 minutes on potential game winning drives"? And then you would juxtapose that against similar questions for the Defense that you have. Which, of course, would not be possible to have beforehand. And then, even if you had all this info input beforehand, the computer would not know "Player X is clearly winded". Or that "Player Y has been getting owned today - perhaps because he was distracted all week at practice by the death of his grandmother". Etc, etc, etc. I was a baseball coach for many years, and remember the birth of analytics. It is a useful tool. But only to INFORM decisions, not to MAKE them. Cheers. These are all very salient points, my obviously-not-a-fan-from-OC friend. And I think they are also demonstrated. For example, Staley had indicated that the decision to go on 4th down (absent a negative play on 3rd) was already made. In particular, citing a favorable match up with MW vs some random rookie substitute corner whose name I can't be bothered to look up. And also his mention that he was confident in the O performance, while D had moments but inconsistency. So correct, it's not (and cannot be) purely robotic. |
|
|
|
Post by joemcrugby on Oct 10, 2022 19:26:21 GMT -7
|
|
|
|
Post by ltocbolts on Oct 10, 2022 21:52:30 GMT -7
One big-picture thing I think you are missing as you consider the situation. The analytics are by definition a factor of what data is given to the almighty computer. There is simply no way that the guy doing the data-input can put in every single variable ahead of time to address the particular situation.Brissett being the QB, rather than Mahomes, is a good example. You would have to have info such as "what is the opposing QBs completion percentage". Or would you plug in "what is his QB rating by quarter?" Or "what is his percentage in the last 2 minutes on potential game winning drives"? And then you would juxtapose that against similar questions for the Defense that you have. Which, of course, would not be possible to have beforehand. And then, even if you had all this info input beforehand, the computer would not know "Player X is clearly winded". Or that "Player Y has been getting owned today - perhaps because he was distracted all week at practice by the death of his grandmother". Etc, etc, etc. I was a baseball coach for many years, and remember the birth of analytics. It is a useful tool. But only to INFORM decisions, not to MAKE them. Cheers. These are all very salient points, my obviously-not-a-fan-from-OC friend. And I think they are also demonstrated. For example, Staley had indicated that the decision to go on 4th down (absent a negative play on 3rd) was already made. In particular, citing a favorable match up with MW vs some random rookie substitute corner whose name I can't be bothered to look up. And also his mention that he was confident in the O performance, while D had moments but inconsistency. So correct, it's not (and cannot be) purely robotic. I hope our friend Freak sees this and gives you the coveted post-of-the-day. Because some forget that there are no Charger fans in OC, and it is important to remind them so they not be misled.
|
|
|
|
Post by ltocbolts on Oct 10, 2022 21:53:22 GMT -7
Hi Joe! That is a great rundown and shows just how good DJ is.
|
|
|
|
Post by chargerfreak on Oct 11, 2022 5:17:22 GMT -7
These are all very salient points, my obviously-not-a-fan-from-OC friend. And I think they are also demonstrated. For example, Staley had indicated that the decision to go on 4th down (absent a negative play on 3rd) was already made. In particular, citing a favorable match up with MW vs some random rookie substitute corner whose name I can't be bothered to look up. And also his mention that he was confident in the O performance, while D had moments but inconsistency. So correct, it's not (and cannot be) purely robotic. I hope our friend Freak sees this and gives you the coveted post-of-the-day. Because some forget that there are no Charger fans in OC, and it is important to remind them so they not be misled. cthommes gets nominated for post of the day by bid of an Alpaca.
|
|
|
|
Post by chargerfreak on Oct 11, 2022 5:17:58 GMT -7
I was promised there would be NO MATH. This thread is filled with it. Thread Reported.
My head is spinning. This is why I have Math Advisors on the Royal Court. I can't be bothered by such trivial matters.
"The QB is dependent variable X." ?? What the living hell is THAT crap ?!?!?!?!?!?!?!
|
|
|
|
Post by joemcrugby on Oct 11, 2022 16:48:00 GMT -7
|
|
|
|
Post by afboltfan on Oct 11, 2022 17:04:15 GMT -7
No comment... 
|
|
|
|
Post by moekid on Oct 11, 2022 18:08:32 GMT -7
I was promised there would be NO MATH. This thread is filled with it. Thread Reported. My head is spinning. This is why I have Math Advisors on the Royal Court. I can't be bothered by such trivial matters. "The QB is dependent variable X." ?? What the living hell is THAT crap ?!?!?!?!?!?!?! PoTD |
|
