QB Sim Stats: Only AWR Matters

[Pogo27]

Kinne is the BEST example for me.

He doesn't break your rule of 81 AWR 60% completions.

But he does break the original theory you were trying to put forward, which is that ONLY awareness matters.

I mean, it's definitely odd that through your two samples all 81+ awr guys have 60+ completions, and all 80- awr guys are below...

However...
1) Completion percentage is not the only quarterback stat. In fact, in my online dynasty playing Syracuse, my quarterback is close to 80% completions. And I realize, we're talking about a non-simmed stat, but the point is...he's leading the nation by a huge margin in completion percentage, yet he'll never win the Heisman because he's averaging about 10 throws per game.
2) Through just two samples, you do have guys who come close to breaking your rule.
3) Linear or not, your rule still doesn't state all that much. I mean, 60% is exceptionally good, but 59% is good too. And if a guy with 80 AWR can do it, and your rule is not linear, then even if your rule holds true and guys sub-81 can never break 60%, it still doesn't mean that in some simmed season GJ Kinne could maybe hit 59%. Yes, your arbitrary, borderline meaningless rule of 81 AWR, 60% completions still holds true, but it absolutely proves that awareness is not the ONLY factor attributing to QB sim stats.


And obviously, awareness plays SOME role in determining simmed QB stats. I don't think anyone was ever questioning that.

[dan_457]

I've checked my stats from my 2018 season, and they match up well with what Polo has posted. It doesn't take a genius to figure out that AWR is the most important factor in completion percentage.

I'll give an example from my stats

QB Martin Scott - 90 OVR with 87 THP and 96 THA 75 AWR for the #8 Houston cougars

56% Completion
133.7 rating

QB Alphonso Brothers - 81 OVR 77 THP 81 THA 90 AWR for Central Michigan - 4th in the MAC

66% Completion
163.6 rating

You can call coincidence, or say there's not enough testing or whatever, but it's not rocket science and doesn't take an expert to notice a clear cut pattern.

[youALREADYknow]

Quote: Originally Posted by Pogo27

Kinne is the BEST example for me. He doesn't break your rule of 81 AWR 60% completions. But he does break the original theory you were trying to put forward, which is that ONLY awareness matters. I mean, it's definitely odd that through your two samples all 81+ awr guys have 60+ completions, and all 80- awr guys are below... However... 1) Completion percentage is not the only quarterback stat. In fact, in my online dynasty playing Syracuse, my quarterback is close to 80% completions. And I realize, we're talking about a non-simmed stat, but the point is...he's leading the nation by a huge margin in completion percentage, yet he'll never win the Heisman because he's averaging about 10 throws per game. 2) Through just two samples, you do have guys who come close to breaking your rule. 3) Linear or not, your rule still doesn't state all that much. I mean, 60% is exceptionally good, but 59% is good too. And if a guy with 80 AWR can do it, and your rule is not linear, then even if your rule holds true and guys sub-81 can never break 60%, it still doesn't mean that in some simmed season GJ Kinne could maybe hit 59%. Yes, your arbitrary, borderline meaningless rule of 81 AWR, 60% completions still holds true, but it absolutely proves that awareness is not the ONLY factor attributing to QB sim stats. And obviously, awareness plays SOME role in determining simmed QB stats. I don't think anyone was ever questioning that.

The word "only" in the thread title is your only complaint then. Change it to AWR matters the most and everything is fine.

The relationship does not have to be absolute and linear in order to obtain statistical significance. If that were the case then the pharmaceutical industry would never have a drug approved.

I'm sure the OP hasn't done a formal statistical analysis of the data because it has passed the basic smell test. Even if he did, based on what I'm looking at it appears that the results are significant enough to be useful. It's also poor programming logic to make AWR have that large of an impact in Completion Percentage while practically ignoring the QB Accuracy and Offensive Playbook/Style.

[Pogo27]

Quote: Originally Posted by youALREADYknow

The word "only" in the thread title is your only complaint then. Change it to AWR matters the most and everything is fine. The relationship does not have to be absolute and linear in order to obtain statistical significance. If that were the case then the pharmaceutical industry would never have a drug approved. I'm sure the OP hasn't done a formal statistical analysis of the data because it has passed the basic smell test. Even if he did, based on what I'm looking at it appears that the results are significant enough to be useful. It's also poor programming logic to make AWR have that large of an impact in Completion Percentage while practically ignoring the QB Accuracy and Offensive Playbook/Style.

By removing the word "only," the base theory does really say much. Again, I'm not arguing that awareness DOESN'T matter. And I don't think anyone is trying to argue that.

But even using this 81AWR/60%comp rule, AWR still doesn't inherently matter MOST. The only thing that matters is that if a QB has over 81 AWR, he will throw over 60%... if after enough sampling, this has been proven absolutely true. He's tested one field of quarterbacks just three times. And in those three sets, we've ALREADY seen players close to breaking the rule. It does imply that given enough testing, the rule easily could be broke.

And perhaps maybe the biggest problem is that no one is going out of their way to TRY breaking the rule--sometimes that's the most important part of any experiment. Force a team to start a QB who is terrible at everything but has 99 AWR, does he still throw for over 60%?

But then again, anyone who has messed around with the player editor already knows that AWR has a huge impact on a QB's skill level, just based on how much if factors into his OVR rating.


It may be that 81 AWR is the secret number at which point quarterbacks actually start calling audibles. If a quarterback with 81 AWR is calling audibles and a quarterback with 80 AWR isn't, regardless of whether or not the game is simmed (because I assume the sim takes this into account), then it stands to reason that the quarterback with 81 AWR is going to have significantly higher completion percentage. He is checking out of passes into runs when he reads the defense and knows they're pass committing and everyone will be overcovered. He's checking out of runs and into passes when he reads the defense and knows there are mismatches.

And if this is the case... if I we accept the 81AWR/60%comp (which I still am not settled that that is even proven), then the audible theory above is possible explanation (which when doing sim only games, is impossible to test). We could possibly do CPU vs CPU games testing QBs with 80 AWR vs 81 AWR only seeing whether one, none, or both audible. This would be part of the scientific method...reformulating a new question, hypothesis, etc.

I still argue that testing should be done to see whether someone with 99 AWR and basically no arm skills could complete 60% passes through a full season, in addition, whether a guy with 99 THA/THP could complete 60% passes with say 70 AWR. These are examples of trying to break the theory. Again, part of the scientific method.

[Palo20]

Quote: Originally Posted by Pogo27

Kinne is the BEST example for me. He doesn't break your rule of 81 AWR 60% completions. But he does break the original theory you were trying to put forward, which is that ONLY awareness matters. I mean, it's definitely odd that through your two samples all 81+ awr guys have 60+ completions, and all 80- awr guys are below... However... 1) Completion percentage is not the only quarterback stat. In fact, in my online dynasty playing Syracuse, my quarterback is close to 80% completions. And I realize, we're talking about a non-simmed stat, but the point is...he's leading the nation by a huge margin in completion percentage, yet he'll never win the Heisman because he's averaging about 10 throws per game. 2) Through just two samples, you do have guys who come close to breaking your rule. 3) Linear or not, your rule still doesn't state all that much. I mean, 60% is exceptionally good, but 59% is good too. And if a guy with 80 AWR can do it, and your rule is not linear, then even if your rule holds true and guys sub-81 can never break 60%, it still doesn't mean that in some simmed season GJ Kinne could maybe hit 59%. Yes, your arbitrary, borderline meaningless rule of 81 AWR, 60% completions still holds true, but it absolutely proves that awareness is not the ONLY factor attributing to QB sim stats. And obviously, awareness plays SOME role in determining simmed QB stats. I don't think anyone was ever questioning that.

1. If it's proven that 81 AWR is the key, then my original thread title still holds water. The title doesn't imply linear progression, just that AWR is the most important.
You may interpret differently, but AWR is all that matters if 81 is proven true. The 81 is the part of the AWR that matters.

2. I've only used Comp% and QB Rating, but if I charted TD/INT ratio, you'd see the same results. High AWR QBs get stuff like 32 TD and 4 INT. Low AWR QBs get 18 TD, 18 INT (for example).

3. I've specified this is ONLY sim stats. Played games are a completely different animal. And thats probably my biggest complaint about this: The game should play similarly when humans play and when it is simmed. If I decide to change every QB's AWR to compensate for the sim stats, then I'm sacrificing the on-field product when I actually play the game.

4. The drop off below 81 AWR is significant. When you look at the stats, there is a huge drop off between the "haves" and the "have nots." The top 15-20 QBs in the nation have very realistic stats compared to the top 15-20 in real life.

Then comes the huge drop off. It's embarrassing that over 80 QBs complete under 50% of their passes, especially when I pointed out some of the names that were there. Only a handful did that in real life last year.

5. I sorted by THA on one of my posts. Check it out. High 90s THAs completing less than 50%.

6. If i continue to simulate, and 95 THA QBs with low AWR NEVER complete a realistic % of passes, then how can we say THA has any value??

[Palo20]

Quote: Originally Posted by Pogo27

By removing the word "only," the base theory does really say much. Again, I'm not arguing that awareness DOESN'T matter. And I don't think anyone is trying to argue that. But even using this 81AWR/60%comp rule, AWR still doesn't inherently matter MOST. The only thing that matters is that if a QB has over 81 AWR, he will throw over 60%... if after enough sampling, this has been proven absolutely true. He's tested one field of quarterbacks just three times. And in those three sets, we've ALREADY seen players close to breaking the rule. It does imply that given enough testing, the rule easily could be broke. And perhaps maybe the biggest problem is that no one is going out of their way to TRY breaking the rule--sometimes that's the most important part of any experiment. Force a team to start a QB who is terrible at everything but has 99 AWR, does he still throw for over 60%? But then again, anyone who has messed around with the player editor already knows that AWR has a huge impact on a QB's skill level, just based on how much if factors into his OVR rating. It may be that 81 AWR is the secret number at which point quarterbacks actually start calling audibles. If a quarterback with 81 AWR is calling audibles and a quarterback with 80 AWR isn't, regardless of whether or not the game is simmed (because I assume the sim takes this into account), then it stands to reason that the quarterback with 81 AWR is going to have significantly higher completion percentage. He is checking out of passes into runs when he reads the defense and knows they're pass committing and everyone will be overcovered. He's checking out of runs and into passes when he reads the defense and knows there are mismatches. And if this is the case... if I we accept the 81AWR/60%comp (which I still am not settled that that is even proven), then the audible theory above is possible explanation (which when doing sim only games, is impossible to test). We could possibly do CPU vs CPU games testing QBs with 80 AWR vs 81 AWR only seeing whether one, none, or both audible. This would be part of the scientific method...reformulating a new question, hypothesis, etc. I still argue that testing should be done to see whether someone with 99 AWR and basically no arm skills could complete 60% passes through a full season, in addition, whether a guy with 99 THA/THP could complete 60% passes with say 70 AWR. These are examples of trying to break the theory. Again, part of the scientific method.

Could you go ahead and test that for me then? I'm busy for the next 8-10 hours.

You're also assuming that the sim engine runs the same way as the game engine. It definitely does not. Playing CPU vs. CPU games and simming games will get you COMPLETELY different results. It's always beent that way in EA games.

[Bootzilla]

Good work Palo. The future of Dynasty will really be screwed up if this pattern holds true. Hopefully EA is doing some major tweakage to the AWR and other issues with Dynasty progression.

And man, the "scientific" police sure came off as rude and borderline obnoxious, ha ha. Dude noticed something, ran a few tests on it and found his latter findings to be consistent with his earlier findings. Good enough for me.

Here's my scientific analysis. Walks like a duck. Quacks like a duck. It's a duck.

[Pogo27]

Quote: Originally Posted by Palo20

Could you go ahead and test that for me then? I'm busy for the next 8-10 hours. You're also assuming that the sim engine runs the same way as the game engine. It definitely does not. Playing CPU vs. CPU games and simming games will get you COMPLETELY different results. It's always beent that way in EA games.

I'm not interested in spending any amount of time simulating through seasons just to get results. Especially arbitrary ones that aren't that game breaking. Especially ones that talk about sim stats, as I'm playing all of my games.

And I'm not making an assumption that the in-game engine and sim engine work the same. I'm merely making the same assumption you are, in that a players attributes affect how he performs. If it can be proven that QBs with 81 or better awareness will audible at least some of the time, while QBs with 80 and below absolutely never audible (I'm not talking about the no-huddle look to the sideline audible--that's not a QB audible)...if that can be proven as an absolutely hard and fast rule, then I think it's a safe assumption to translate over to the sim engine.

And your results have proven that players have come at least close to breaking your rule through just 3 samples (and you haven't even given complete results for the 3 samples you took).

You also haven't REALLY gone out of your way to attempt breaking the rule.

It's not just whether a 95 THA with low AWR consistently has bad completion percentage. It's does 99 THA/THP with 70 AWR (or even 80, since according to your rule, the break is at 81) throw below 60% every time over numerous, numerous samples? Does 50 THA/THP with 99 (or even 81, since again, that's your rule) consistently throw above 60% every time over numerous, numerous, numerous samples? If you can't show these two things, then even your very simplistic, and as I said, borderline meaningless, rule is disproved.


Ultimately, perhaps what bothers me the most, is that there seem to be an abundance of posters on this forum that are looking to prove something. They post a miniscule amount of data to back their claim, have a limited understanding of the scientific knowledge, usually at first don't know how to really test the theory and try proving their own theory wrong, and then once it has been explained to them how to truly test their theory and attempt proving themselves wrong, they have limited to no interest in attempting to do it. They created the thread because they were interested in proving some random stat of arguably questionable significance and coming back and proving that their original idea was wrong would make their hole thread pointless--or so they think.


I hope you understand that I'm not trying to prove you wrong. If anything, I'm trying to prove you right, which you've failed to convincingly do so far. Not that you've posted any data that's contrary to your theory, just that you're really failed to truly flesh this theory out and really try testing it and breaking it.

You're aligning QB's AWR rating and completion percentage. And thus far, again with what most people who understand the scientific method would call a laughable amount of data, it has seemed to hold true. And again, the limited data that you've brought so far is no doubt interesting. But so far, I don't really see anything significant proven.

A quarterback with less than 81 awareness SHOULDN'T complete 70% of his passes. At 80 awareness, I figure he should be misreading defenses at least 20% of the time, and you're just not going to complete the pass every time you read the defense correctly. Moreover, QBs with higher awareness should have higher completion percentages despite weaker, less accurate arms, because they're reading the defense correctly and throwing to the most open WR a much higher percentage of the time.

And if it's not linear, or anywhere near linear (which it's not, as you've stated numerous times), then that inherently PROVES without question that AWR is not the ONLY factor that matters. All you've proven (or think you've proven but lack sufficient data and testing to really stake a claim at PROVEN) is that awareness may be the only factor that determines whether or not a QB can throw for 60% of better.