I apologize in advance, but I gotta drag this one out again and further flog the dead horse!

What I've struggled with in understanding this still bothers me from time to time, and that is I do not know how the game handles all ratings for all positions.

That brings up a further related suggestion Id like to make -
In the manual at some point, it would be awesome if we had a matrix/chart, with the rows being = POSITION, and the columns being each ATTRIBUTE.

So for instance>

STR AG ARM ....
QB Affects how far can throw Affects ability to dodge QB key Attribute for getting a ball to a receiver. Higher Arm means more completions. Factors more on long throws.
RB Affects ability to break tackles Affects ability to dodge, catch
FB Affects blocking....
...

This would make referencing for us users very simple, and makes sure we know exactly how the attributes are used.
I made up some of the values above, bc I still don't know how they are all used, and this is why I feel I can't always make an acccurate Custom Overall Rating. It's important to me to make an acurate one, for the simple fact of my own proper understanding of the way the game engine works, if nothing else. Of course, I would also like it to be an accurate reflection of just how good players are in comparison to one another.

For instance -
I thought "ARM" for QB meant how far he could throw the ball, whereas Accuracy means just that, how accurate. So, I wanted to have a strong arum QB who was not very accurate, kind of like a Josh Allen-type(2018 Josh Allen, anyway). So, I gave him an ARM of 91, with Accuracy of 78.
Then, I created a QB who I wanted to be very accurate, but not necessarily a really strong arm. Perhaps someone like Chad Pennington, but a little bit stronger arm -
ARM: 78, ACC: 91

I misunderstood, you see, how important ARM actually is as ar as the way the game engine uses that rating. Once I located a forum post where Ratings were more speciically discussed, I came across QB and ARM, and read this stmt in particular - "Higher Arm means more completions." I DID NOT understand that beforehand, but it makes sense now.
I thought my 2nd QB in the example above would be a better QB, as Accuraccy is typically more important in the NFL than having a really strong arm.
However, after having played most of a season, the QB1 in the example, the strong-arm QB, is obviously the better player. Having come across that more speciic description of how ARM is used, and how Accuracy "factors more on short throws."(another specific way the game engine handles this rating that is not intuitive, at least not to me)

So tying this all toghether, I would now place a greater weight on ARM vs ACC in the Custom Overall Rating forumulas than I had previously. This way, QB's will be more acccurately represented with their Overall Rating, so I can create the players I want. Previously, for instance, in the Custom OVerall Rating formula screen, I had assigned a weight o ACC = 50, ARM a 40. Now I might swap them and have ARM at 50, or perhaps lower ACC to 40 so they are at least equal.

Ultimately, I guess my requests are these-

1) Create, for manual or at least here on forum, a Matrix comprised of POSITION x RATING
2) Review the default Custom Overall Rating formula, make sure it accurately reflects how important each rating truly is for each position to ensure the overall custom rating is as accurate a relection of the player's overall ability as possible

Brooks, I know your usual response is that advanced users don't rely on the overall rating, they know which Rating's are important for them, etc.
I dispute this, though, as it's not really a matter of personal preference in many cases - I WANTED Accuracy to apply to a QB's ability to place the ball for both short and long passes, and ARM to only cover ability to fire the ball in but more importantly, how far he can throw the ball. What I wanted did not matter, bc Arm apparently covers much more than what I at first thought, and is more important to the game's code/engine that what I assumed. I would not have known this, as I did not code the thing! I only knew after re-examining the specific notes located in this post - viewtopic.php?f=256&t=27634
It makes me wonder - are there still other POSITION x RATING relationships I don't understand? For instance, could QB x STR affect how far a QB can throw the ball? Seems intuitive to me, but again, I didn't create the game! So I have no way of knowing. Perhaps all STR affects for a QB is breaking tackles. But I wouldn't be 100% sure until we have something like that matrix idea I listed above.
We can't know which Ratng's we should be looking at if we don't understand all the relationships. Say I don't care about overall rating - with Guards, I lookat their STR and RUN BLOCKING, as I'm a run heavy team. Little do I know, though, that I should also be considering AGI, as that "Affects ability to adjust in blocking." This is not necessarily intuitive. A solid Custom Overall Rating default ensures that, even if I as a User don't understand the importance AGI plays in a blocking player, they player's overall rating will factor AGI in appropriately and spit out an Overall Rating I can then be sure to rely on when determing my depth chart.

okay I really feel like I made a good argument this time! Brooks?

Bottom line is I don't know exact numbers to give you. All of those expected attributes affect different aspects of the engine. Personally I think Arm should be higher than Acc because it affects a few more items, but if people set up a short passing game then absolutely they'll want a higher Acc QB. So the default OVR in the game is close enough for me from a 10,000 foot level in evaluating talent. After that it's up to people to evaluate ratings and choose the players that work for them regardless of OVR.

We've already posted elsewhere the attributes that are important for positions. And the default custom OVR has that matrix applied. If you want to say weights should be +- 10 points here or there I can't really argue that, which is why it's custom and you are free to adjust it to your liking.

This wound up being REALLY LONG, even for me, so I apologize in advance.
I humbly request that you do read through it Brooks as much as your patience allows, as this is as important to me as any other matter I’ve written about here on the forum. I feel it really touches on some of the core aspects of the game, as I attempt to gain as thorough an understanding of those key areas as possible.
If you need to, skim the first part but I am esp interested in your thoughts about what ive presented here in the “POSITION SKILL Weight in Custom Ovr Rating calc-“ section.

QUICK SUMMARY-
1) ONLY BROOKS OPINION MATTERS HOW RATINGS SHOULD BE VALUED BC HE CODED GAME –
a. “Personally I think Arm should be higher than Acc because it affects a few more items, but if people set up a short passing game then absolutely they'll want a higher Acc QB” – that’s all that matters, so you just made it obvs ARM should have a higher weight than Acc; sure short passing values ACC but teams are gonna wanna throw it deep SOMETIMES; this is this info I want. Are you sure we have it all? Recently you said “ “he(qb) has to pass other checks (intelligence is one) to know when to scramble” -not sure this is clearly stated in “GUIDES”(im telling yah the Matrix idea could be great! Make things 100% clear and comprehensive)

2) LETS DISCUSS/REVIEW WHAT ALL POSITION SKILL RATINGS DO IN FACT AFFECT
SUMMARY END*******************************************

Rating weights are not a matter of opinion; how I value arm strength in real life vs how you or a hundred other scouts is irrelevant bc it's subjective, there is no definitive answer.
But in our game, we can see the “Matrix”, as it were, we do have a definitive answer about which is more important, Arm vs Accuracy, bc it is coded a certain way.
I understand there's still some subjectivity and taste - I may value quicker more agile blockers while you may like stronger behemoths.
But understanding how each and every rating affects each and every position, and to what extent, is not something governed by my opinion-
Only one opinion here matter - BROOKS' (and his team's, presumably) OPINION, as game designer!
If I want an effective QB, fpr example, I better give him a sufficient ARM score.
Montana didn't have the biggest Arm as far as throwing the ball 80 yards, but he got by with unparalleled accuracy.
Based on the way the game engine is coded, which I understand through both trial and error and the way it is described in the documentation, I cant just give him all ACC with no ARM. I do that and I won't see Joe Montana like results. This is exactly what im talking about - I need a better understanding of how the game engine/the code, the way you wrote it, uses each rating for each position so I can create the players who will perform, or be reasonably expected to perform, how I want them in THIS game. It makes no sense if I want a Mike Vick type QB and give him high SP if I dont also know this key rating that I would otherwise have had no idea about, and based on my opinion, it wouldnt matter-but the game dictates he needs proper INT to know when and when not to scramble to minimize injury risk and maximize yards/carry. This is what I mean. If I dont know your coding dictates INT affects his scrambling, I wouldnt place any weight for it in my Custom Overall Attribute calculation - and thus my calc would be objectively WRONG! There is personal preference and then there is right and wrong, and real world intuition and opinion don’t always translate. Do I Disagree that INT is important for a scrambler to judge when to take off, or AGILITY making a player a better blocker? No, I totally agree, but I wouldn’t have known to give them their proper weight in formulating my Overall Ratings bc there are arguments for just about every attribute affecting every position - its not my knowledge of my own personal perspectives but my understanding of how your perspectives drove u to design the game. I could say "hey STR is important for wideouts bc it represents hand strength which allows them better grip on the football" but if I used this personal logic and applied weight with STR for WR, my overall would be inaccurate compared to one that did not apply this weight in calculation. Whats the harm?

Well Brooks, I know you are always saying advanced coaches shouldnt even have much use for OVR, instead knowing which key attributes they value for the type of system they are running/results they are seeking. One WR might have a higher OVR than another bc of higher SP but lower hands, but in a West Coast system that values say Hands and agility more so than straight line speed, this slower WR might actually be of greater value to the west coast team – the advanced coach therefore looks at Hands and agility before SP, despite what the OVR indicates as to who the “better” WR is.

This is all well and good, but it isn’t always so simple, OVR def has its importance and place when things arent so clear cut and obvious; ex-
QB1 ACC 85, ARM 80
QB2 ACC 80 ARM 85
Who is better, ie, who should start on my team? In general, just by glancing, its difficult to say. Yes, perhaps a team that says they want to throw the long ball more would value the bigger arm, and the west coast team takes QB1 bc of greater ACC; but hold on. Looking closely at the exact way the ratings are described, and based on my own testing, ARM is more than just simple arm strength.
"Higher Arm means more completions" says the guide. I don’t think you could make this definitive a stmt about “Arm Strength” in real life. QB2 will still be accurate w an 80, and i have found that ARM is at least as important as Accuracy in this game(as oppose to real life- plenty of mediocre arms with great Accuracy are very successful-think of Drew Brees, or even Chad Pennington for a more extreme example-whereas NFL history is littered w guys w big guns without the accuracy to match who flamed out-Ryan Leaf anyone?), and probably more important. My exp w the game would lead me to take QB2 in either case, and a properly weighted Custom OVR rating would reflect this.


POSITION SKILL Weight in Custom Ovr Rating calc-
Where I still feel the greatest disconnect lies in the weight of Player Position ratings in the OVR calculations - it seems to have changed quite a bit from how it used to be weighted, and based on my understanding of the game engine, I think you had it right before. I have brought up this specific issue before and it still bugs me a lot -
As it currently stands, with the weights from the default Custom Overall Rating, it takes a difference of something like 6-8 points in Position Rating to change one point of Overall Rating; important attributes like Accuracy change overall rating with a mere 2 points of difference.
THAT ratio def feels off to me; but again, ('m more than willing to accept that I'm wrong, and Position Rating just doesnt have the impact on player performance that both-
1) I thought it did based on observations from previous versions of the game/overall rating
2) I feel Overall rating SHOULD have; I think it was a minor stroke of genius to include this oft ignored aspect to a player. Again, what I think SHOULD be doesn’t matter, so im after what is, so some of this last part of the post is confirmation that the way we have things really is how they should be.
Almost as important as anything, I'm fairly certain in older versions of the game, POSITION SKILL DID have a greater affect/higher weight in regards to OVR. I think it was more around variance of 4 pts of SKILL rating for every 1 point of OVR. This felt right.

Here's the other discussion, and probably the most important part of this whole massive post -
Some guys, as has been well-documented, JUST KNOW HOW TO PLAY FOOTBALL!! They dont score great in the Combine measurables don't blow u away with their 40 yard time, but they are just gamers and get it done at their position. Emmitt Smith has to be one of the greatest examples of this - not particularly fast or quick, nor strong, or ridiculously athletic, he was the anti-Barry Sanders in a way. Despite the lack of any obvious signs of greatest, when he got on the field, he had an innate sense of how to play RB, to find the soft spots in the D, the ability to make split-second decisions to change directions ever so slightly to get every inch of every run, and how to fall forwards always, and he took this innate ability all the way to the record books by rushing for more yards than anyone EVER! So, as an exercise, lets use the tools at our disposal in this game to create an accurate recreation of Mr. E. Smith. Speed mediocre, STR decent but nothing in the 90s, looking closely at the ratings descriptions again from the "GUIDE" thread, looks like "INT" is our best bet -
"Helps ballcarriers find holes, react to defenders movements."
That actually looks pretty great, we might even max that out to 100. Endurance would also be 100 surely for a guy who had 1000 yard seasons well into his 30s, and who kept going strong into the 4th quarter and into the 30+ carries range, game after game well after the defense was exhausted. There are two other possible tools I can think of that could reflect "hidden" talent, that which cant be described by the typical measurables - size, which was certainly vital for him in that he used his low center of gravity to avoid hits and prolong his career, as well as hide behind those giant offensive linemen, but from my understanding, size does not have any bearing on the game engine(be great if we could get size to be incorporated in some respect going forward), and Position Skill, which again I think you max out to 99 or 100 to represent his general skill of being a good Running Back(also a good receiver out of the backfield and blitz pickup, though these may be covered by Hands and Pass Blocking, respectively, so its probably therefore mostly representative of the ability to run, to gain positive yards running the football). With this, we have a pretty good chance of creating an RB who maybe has SP 85, AG 88, STR 75, good but not great attributes, who could go on to rush for 18000 yds+. How could he have an OVR anywhere near that of a Barry Sanders, his annual rival for best RB in the league during the 90s? Well, INT only counts 10, END only 1(Brooks, thinking of Emmitt Smith, should END perhaps have more than a 1 weight? Does the engine take into account things like the defense being gassed in the 4th quarter, while the RB who can keep on chuggin? This is CRUCIAL for guys like Emmitt, and maybe even more so bruisers like Derrick Henry, at least it is in the real NFL - how it is programmed?).
If SKILL remains at only 10 weight, then guys like Smith who get by more on ability/skill than physical superiority might wind up not even starting. This raises another important question – is this a failing in the OVR rating’s ability to properly represent a player’s ability, or a shortcoming in the game’s ability to feature player’s like Smith, who had more skill than size, speed or strength? If POSITION SKILL really does only affect the game’s results so that a 10 weight is correct(meaning, for instance, Speed is 5X more important than a RB’s Position Skill, which also leads me to conclude that you would have to flub Emmitt’s SPEED rating to get appropriate results; as if you left his SP at say a more realistic 85, even if u put his Position Skill at 100, he would simply not be as good as his real world counterpart) , then we are limited in the different kinds of player’s we can create, and an accurate simulation of Emmitt Smith as things currently stand is impossible. Boost the importance of Position Skill to the game engine calculations (which would then of course necessitate a corresponding raise in the weight of that rating in the OVR calc), and some new interesting possibilities arise, and I can’t see this being anything but great for the overall well-being of the game.

Sure, feel free to show your Custom Overall calculations and I can take a look at adjusting the ones I put in. That was part of the goal of having that custom formula... people can tailor it to fit whatever views they have of rating, or whatever player universe they're working with. If you're importing from somewhere else, or creating mods, the formula may not work for those scenarios. You can also take the formula and smooth everyone out so they're closer together, or use an 80 point scale, or whatever. But certainly people can share how they've tweaked it, you guys can discuss, and let me know if you think the defaults in the game need to be updated.

I don't think we ever intended Skill to override physical attributes... it was more of a way to spread out ability across various positions so you couldn't just move WR to RB and vice versa without some penalty for not learning the position. Having a 99 speed RB with 50 skill and a 80 speed RB with 99 skill I'd still take the fast guy all day every day.