Sliders and QB acc for short, medium, long

[capa]

I am wondering if once again the "fix" for high QB completion % is due to the ratings they have for short, medium, and long passes. Is there any correlation between the ratings and overall completion %...and should there be?

For example, if a QB "averages" around 62% completion rating over the course of a season, should his short, medium, and long attribute ratings mathematically tie into that?

Let's say QB A avgs 62% over the course of a year. How did we know what the proper rating for each of the pass categories should be? Should they average out to 62%?

An example would be in this case the total base numbers to work with to gain an average of 62 would be 186 (3x62). Then somehow divide that up into the 3 categories: short = 74, medium = 60, long = 52 (just an example and would vary by different QBs). Something along those ranges anyway.

Just thinking out loud here, but I do think once again ratings screw up the game play and cannot be truly corrected through slider settings.

And yes, I know it screws up incoming draft classes...but if you only play 1 or 2 seasons it might hold up.

Thoughts?

C

[JoshC1977]

The problem is that Completion percentage, by itself, is not a reflection of how accurate a QB throws. This is analogous to ERA in baseball, where it is a reflection of the pitcher's skill, his ballpark, and his defense. This is why Fielding Independent Pitching (FIP) is a much better indicator of pitcher performance.

Now, taking this analogy to football, you need a something similar to measure completion percentage. You have to factor in WR drops, DB deflections, interceptions, tips at the LOS, etc. But for each of these, is it a reflection of a bad pass by the QB, or the other player? Not only that, but a completed pass can actually be poorly thrown. So, at the end of the day, we are subjectively evaluating every throw a QB makes and making a judgment call on whether it was an accurate throw or not.

I am aware that Football Outsiders has a lot of advanced metrics, do they have one that might help better evaluate QB accuracy than raw completion percentage?

[D81SKINS]

Quote: Originally Posted by capa

I am wondering if once again the "fix" for high QB completion % is due to the ratings they have for short, medium, and long passes. Is there any correlation between the ratings and overall completion %...and should there be? For example, if a QB "averages" around 62% completion rating over the course of a season, should his short, medium, and long attribute ratings mathematically tie into that? Let's say QB A avgs 62% over the course of a year. How did we know what the proper rating for each of the pass categories should be? Should they average out to 62%? An example would be in this case the total base numbers to work with to gain an average of 62 would be 186 (3x62). Then somehow divide that up into the 3 categories: short = 74, medium = 60, long = 52 (just an example and would vary by different QBs). Something along those ranges anyway. Just thinking out loud here, but I do think once again ratings screw up the game play and cannot be truly corrected through slider settings. And yes, I know it screws up incoming draft classes...but if you only play 1 or 2 seasons it might hold up. Thoughts? C

I was wondering the same thing, the qbs completion % is mainly based on these 3 ratings(sac, mac, dac)well considering they are not scrambling, having the pass deflected, or WR drops. Obviously the %'s would be lower calculating those into it.

It would be a pretty simple test to edit these 3 ratings to avg. What the qbs actual comp % is irl and see how it plays out over a hand full of games.

[D81SKINS]

Quote: Originally Posted by JoshC1977

The problem is that Completion percentage, by itself, is not a reflection of how accurate a QB throws. This is analogous to ERA in baseball, where it is a reflection of the pitcher's skill, his ballpark, and his defense. This is why Fielding Independent Pitching (FIP) is a much better indicator of pitcher performance. Now, taking this analogy to football, you need a something similar to measure completion percentage. You have to factor in WR drops, DB deflections, interceptions, tips at the LOS, etc. But for each of these, is it a reflection of a bad pass by the QB, or the other player? Not only that, but a completed pass can actually be poorly thrown. So, at the end of the day, we are subjectively evaluating every throw a QB makes and making a judgment call on whether it was an accurate throw or not. I am aware that Football Outsiders has a lot of advanced metrics, do they have one that might help better evaluate QB accuracy than raw completion percentage?

Football outsiders has some new metrics updated 12/29/14. It gives us a good idea of more than just comp %'s.

http://www.footballoutsiders.com/stats/qb

[D81SKINS]

http://www.footballoutsiders.com/stats/qb
If you click on the qbs name it will show you his game chart stats,
ie. P. Manning 2013 season: Rnk: 10 Short: 55% MID: 26% Deep: 13% Bomb: 6%.
I'm assuming they don't have the 2014 season calculated yet.

This could help in pinpointing what each qb should be rated in those areas.

[D81SKINS]

Looks like another case of EA over rating players all for the benefit of the casual gamer. "E" everyone can pick up the controls and have a fairly successful passing game.

Wonder what P. Manning would end up with % wise if we rated him off his real life game charts? I doubt we would be able to go as low as Sac: 55 Mac: 26 Dac: 13 and how could we incorporate bomb %'s into it?

[capa]

Quote: Originally Posted by D81SKINS

Looks like another case of EA over rating players all for the benefit of the casual gamer. "E" everyone can pick up the controls and have a fairly successful passing game. Wonder what P. Manning would end up with % wise if we rated him off his real life game charts? I doubt we would be able to go as low as Sac: 55 Mac: 26 Dac: 13 and how could we incorporate bomb %'s into it?

I think you would just have to combine long and bombs into one %. Interesting stuff, but yeah...hard to tell how it would translate to the game. Probably not well. I am sure there is some synergy and co-dependence between the AI engine and the default player ratings that EA puts out. Once we tinker with that who knows how it will play?

C

[D81SKINS]

Quote: Originally Posted by capa

I think you would just have to combine long and bombs into one %. Interesting stuff, but yeah...hard to tell how it would translate to the game. Probably not well. I am sure there is some synergy and co-dependence between the AI engine and the default player ratings that EA puts out. Once we tinker with that who knows how it will play? C

The game chart stats are based on 100% of the qbs throws and not his comp %. I was reading it wrong.
So P. Manning completed 55% of his total passes short, mid: 26, deep: 13 and bomb: 6