06-23-2017, 01:30 AM
(This post was last modified: 06-23-2017, 11:33 AM by Sweetwater.)
[div align=\\\"center\\\"]What the f@%k is going on with CBs?[/div]
[div align=\\\"center\\\"]An in depth analysis of league cornerback statistics attributes, and archetypes.[/div]
Over the past weeks of the league, I was noticing that CBs seemed to be all over the board in terms of stats, and a plethora of builds that seemed to affect things. I was mostly brought to this analysis by having (what I thought was) a solid zone corner build, but somehow hemorrhaging yards to receivers. But why is this happening? Am I upgrading the wrong stats? Did I pick the wrong archetype? Am I just unlucky? Here's what I found:
To start this off, I'd like to tell you about my numbers and methodology. I started by pulling the raw game stats from the index, which gave me the amount of tackles, PDs, and INTs. From there I needed to establish some parameters. For the purpose of the charts, I only counted what I didn't have to question, meaning I only counted passes where the CB made the tackle, as you can't prove missed tackles and/or touchdown passes. So, Targets is the total amount of plays you had a DIRECT impact on. From there I needed to find out how many of your tackles were on passing plays (CB's stop runs too), and how many yards you let up on catches. So I went play by play and added everything up for all 18 games. I also pulled player attributes and TPE totals from their Player pages, did a bit of math, and came up with this first chart of raw data.
Ballhawk stats in this first chart are Agility/Intelligence/Speed/Hands, and Tackle stats are Strength/Tackle.
As you can see, definitely not what anyone would call definitive. But we can also see that teams tended to have a plan when it came to CBs. San Jose runs a zone, Orange County a more Man to man, Yellowknife strictly Man to man, and Colorado and Baltimore running a mix. After that, I decided to get a bit more in depth.
So, here are the CBs sorted by archetype.
Here we finally get a little bit of sense. Man to man archetype corners seem to do better overall than all the others, and Zone archetypes seem to do the worst. For me this raises more questions than it provides answers.
For instance, why is it that zone corners have higher Intelligence and Hands, both skills that the sim claims to improve ballhawking, have arguably the worst ballhawking stats? And why do they have such a higher yards per target average? I will delve even deeper to try to find some answers, but I have a serious question for the developers of the sim, is a zone defense just an incredibly difficult thing to simulate? Or, do you not understand how a zone defense works? When 2 of the top 5 corners in the league (Richard Sherman and Josh Norman) are zone corners, why are they underperforming in the simulation?
Time to go deeper.
Here are the Cbs sorted by total TPE (please forgive the flip of Lavelle and Hansen, it's late)
Here we see the total TPE means almost nothing. John Canton, the CB with the most TPE in the league is one of the worst performing corners, and H. Rivere, who hasn't ever updated their player, is statistically a much better corner. So, if total TPE doesn't really effect how well a corner performs, what does?
Here are the CBs sorted by total "Ballhawk" stats. Not the best metric for sorting, but these are the stats the developers tell us most effect how you perform as a cornerback. So, theoretically, the more points you have, the better you should do.
Again we see game performance all over the place, but also again, we see the players at the top performing worse than players below them on the chart. Obviously, this shows that there is definitely more to the stats than just "more is better", let's see if we can find out how.
Here are the CBs sorted first by total Tier 1 stats: Agility/Speed, then tiebreaker by Tier 2 stats: Intelligence/Hands
Now you can start to see that, as with almost everything in this game, speed appears to be an important factor in cornerback performance. It's not everything, but in general faster guys play better.
Here is the flip of that, CB's sorted by Tier 2 stats, which the developers have claimed increase your chances at PDs and INTs, with tiebreaker being speed
So it doesn't appear as if having more Intelligence or better Hands significantly improves your Ballhawking ability. This narrows down defensive plays to be more reliant on speed than anything else, and this next chart seems to confirm that.
Here are the CB's sorted by Speed
This all but confirms it. If we throw out J. Harris who is a statistical outlier due to his limited playtime, we see that 4 of the top 5 Ballhawks have below average Intelligence AND Hands.
But, the biggest oddity here is J. Williams, who despite being significantly below average in every stat except Intelligence, has the most interceptions in the league.
Finally here are some interesting player comparisons I looked at to try and understand the sim just a little better.
Here are two CBs with the same speed stats, but opposite Ballhawk and Tackle stats
So, Hands is preferable to Intelligence for Ballhawking right?
Or does Speed trump all of that?
Or is most of this just total bullshit and it's about 95% luck?
I'm going with the third option. But this journey has led me to believe that something is messed up about zone corners. They do worse in almost every game metric, despite being above average in almost every stat.
Some other interesting notes from my research:
Only two corners have given up over 100 yards in a game: Maddox and Delacour.
The Otters seem to use corners like linebackers, as their CB group has the most tackles on rushers compared to all other teams.
Strength and Tackle have seemingly no impact on corners, or at least have no noticeable correlation to in game performance.
While Speed and Hands seem to have the biggest impact on Ballhawking ability, nothing in the numbers really suggests what makes the biggest impact on yards per target. I invite any readers to pour over the numbers themselves and see if they can find something I missed.
Hopefully as the season goes on we will be able to see more of the impact our attributes have on our players. But, I would like to petition for zone corners to be able to change archetypes somehow, pending a post season analysis. But the numbers point to the Zone archetype being broken in some manner. If it's not the zone archetype, the numbers also point to speed being the most important stat to a CB, which would have made me distribute my TPE much differently than I currently have. I plan on getting the sim at some point here and running some tests to see how stats and defensive schemes really affect performance, but I think we can agree that the sim is not without its flaws.
I find it somewhat disheartening that my own player is struggling so much despite having everything going for him. John Canton has the most TPE of all corners and has at or above league average in every skill that impacts a corner. In fact, he is the only corner in the league not to have any skill below league average. In spite of all this he is still, statistically, the worst performing CB in the league. Something just doesn't add up in my book.
I'd like to think someone can get some use out of this analysis, and that I didn't just waste 3 hours going through every game and sorting data. But the moral of the story is: If you're making a CB, choose the Man to Man archetype, put as much as you can into speed, and everything else into hands, and apparently it'll be the best player on the field.
I'll be doing a followup to this piece before the playoffs to get some more data to input, and will be doing a film breakdown podcast in the coming week.
[div align=\\\"center\\\"]An in depth analysis of league cornerback statistics attributes, and archetypes.[/div]
Over the past weeks of the league, I was noticing that CBs seemed to be all over the board in terms of stats, and a plethora of builds that seemed to affect things. I was mostly brought to this analysis by having (what I thought was) a solid zone corner build, but somehow hemorrhaging yards to receivers. But why is this happening? Am I upgrading the wrong stats? Did I pick the wrong archetype? Am I just unlucky? Here's what I found:
To start this off, I'd like to tell you about my numbers and methodology. I started by pulling the raw game stats from the index, which gave me the amount of tackles, PDs, and INTs. From there I needed to establish some parameters. For the purpose of the charts, I only counted what I didn't have to question, meaning I only counted passes where the CB made the tackle, as you can't prove missed tackles and/or touchdown passes. So, Targets is the total amount of plays you had a DIRECT impact on. From there I needed to find out how many of your tackles were on passing plays (CB's stop runs too), and how many yards you let up on catches. So I went play by play and added everything up for all 18 games. I also pulled player attributes and TPE totals from their Player pages, did a bit of math, and came up with this first chart of raw data.
Ballhawk stats in this first chart are Agility/Intelligence/Speed/Hands, and Tackle stats are Strength/Tackle.
As you can see, definitely not what anyone would call definitive. But we can also see that teams tended to have a plan when it came to CBs. San Jose runs a zone, Orange County a more Man to man, Yellowknife strictly Man to man, and Colorado and Baltimore running a mix. After that, I decided to get a bit more in depth.
So, here are the CBs sorted by archetype.
Here we finally get a little bit of sense. Man to man archetype corners seem to do better overall than all the others, and Zone archetypes seem to do the worst. For me this raises more questions than it provides answers.
For instance, why is it that zone corners have higher Intelligence and Hands, both skills that the sim claims to improve ballhawking, have arguably the worst ballhawking stats? And why do they have such a higher yards per target average? I will delve even deeper to try to find some answers, but I have a serious question for the developers of the sim, is a zone defense just an incredibly difficult thing to simulate? Or, do you not understand how a zone defense works? When 2 of the top 5 corners in the league (Richard Sherman and Josh Norman) are zone corners, why are they underperforming in the simulation?
Time to go deeper.
Here are the Cbs sorted by total TPE (please forgive the flip of Lavelle and Hansen, it's late)
Here we see the total TPE means almost nothing. John Canton, the CB with the most TPE in the league is one of the worst performing corners, and H. Rivere, who hasn't ever updated their player, is statistically a much better corner. So, if total TPE doesn't really effect how well a corner performs, what does?
Here are the CBs sorted by total "Ballhawk" stats. Not the best metric for sorting, but these are the stats the developers tell us most effect how you perform as a cornerback. So, theoretically, the more points you have, the better you should do.
Again we see game performance all over the place, but also again, we see the players at the top performing worse than players below them on the chart. Obviously, this shows that there is definitely more to the stats than just "more is better", let's see if we can find out how.
Here are the CBs sorted first by total Tier 1 stats: Agility/Speed, then tiebreaker by Tier 2 stats: Intelligence/Hands
Now you can start to see that, as with almost everything in this game, speed appears to be an important factor in cornerback performance. It's not everything, but in general faster guys play better.
Here is the flip of that, CB's sorted by Tier 2 stats, which the developers have claimed increase your chances at PDs and INTs, with tiebreaker being speed
So it doesn't appear as if having more Intelligence or better Hands significantly improves your Ballhawking ability. This narrows down defensive plays to be more reliant on speed than anything else, and this next chart seems to confirm that.
Here are the CB's sorted by Speed
This all but confirms it. If we throw out J. Harris who is a statistical outlier due to his limited playtime, we see that 4 of the top 5 Ballhawks have below average Intelligence AND Hands.
But, the biggest oddity here is J. Williams, who despite being significantly below average in every stat except Intelligence, has the most interceptions in the league.
Finally here are some interesting player comparisons I looked at to try and understand the sim just a little better.
Here are two CBs with the same speed stats, but opposite Ballhawk and Tackle stats
So, Hands is preferable to Intelligence for Ballhawking right?
Or does Speed trump all of that?
Or is most of this just total bullshit and it's about 95% luck?
I'm going with the third option. But this journey has led me to believe that something is messed up about zone corners. They do worse in almost every game metric, despite being above average in almost every stat.
Some other interesting notes from my research:
Only two corners have given up over 100 yards in a game: Maddox and Delacour.
The Otters seem to use corners like linebackers, as their CB group has the most tackles on rushers compared to all other teams.
Strength and Tackle have seemingly no impact on corners, or at least have no noticeable correlation to in game performance.
While Speed and Hands seem to have the biggest impact on Ballhawking ability, nothing in the numbers really suggests what makes the biggest impact on yards per target. I invite any readers to pour over the numbers themselves and see if they can find something I missed.
Hopefully as the season goes on we will be able to see more of the impact our attributes have on our players. But, I would like to petition for zone corners to be able to change archetypes somehow, pending a post season analysis. But the numbers point to the Zone archetype being broken in some manner. If it's not the zone archetype, the numbers also point to speed being the most important stat to a CB, which would have made me distribute my TPE much differently than I currently have. I plan on getting the sim at some point here and running some tests to see how stats and defensive schemes really affect performance, but I think we can agree that the sim is not without its flaws.
I find it somewhat disheartening that my own player is struggling so much despite having everything going for him. John Canton has the most TPE of all corners and has at or above league average in every skill that impacts a corner. In fact, he is the only corner in the league not to have any skill below league average. In spite of all this he is still, statistically, the worst performing CB in the league. Something just doesn't add up in my book.
I'd like to think someone can get some use out of this analysis, and that I didn't just waste 3 hours going through every game and sorting data. But the moral of the story is: If you're making a CB, choose the Man to Man archetype, put as much as you can into speed, and everything else into hands, and apparently it'll be the best player on the field.
I'll be doing a followup to this piece before the playoffs to get some more data to input, and will be doing a film breakdown podcast in the coming week.
Code:
1325 Words