Hopefully, Madden Uses Data Like This

[KBLover]

Here's the broken (for offense) and missed (for defense) tackles rates for last season.

http://www.footballoutsiders.com/stat-analysis/2013/broken-tackles-2012

http://www.footballoutsiders.com/stat-analysis/2013/broken-tackles-2012-defense

At the very least, use it for ratings and representing how guys play. Guys like Foster produce if nothing else for stamina and consistency, not stiff-arming and spinning and juking their way down the field.

Hopefully, quick guys are quick and not having sky high "moves" ratings. These are guys that you shouldn't get a lot of tackle attempts against because it's hard to get in position to attack them. They use vision to find daylight - not get swarmed and use a ton of "moves" to escape (those kinds of runs are few and far between) as if it's the new "Pro-Tak moment".

Hopefully, the TAK/POW ratings will be represented by data as well and that elite TAK actually looks and feels like it. Not seeing 90+ TAK guys "bounce" off HBs to emulate "elusiveness". ("Elude" doesn't not mean take contact and bounce off it...)

Hopefully not all DBs will get crap TAK ratings. However, hopefully STR actually matters so that DBs tackle guys they typically do (and the most physical stand out over the cover guys - likewise for wide outs, separate the pure speed merchants vs the powerful WR vs the dangerous speed-power WRs)

Hopefully, the "Run Free" doesn't turn this all into more of mess than it is now. Hopefully, it's just a way for the ball carriers to display their running abilities (power or quickness or both if so gifted) within reason and with fewer animation limits while defenders that are excellent also can still make plays. There's elusive and quick and powerful ball carriers in the NFL, yet Bruce Carter had a 100% success rate and Willis is there again with his 98%. There's DBs that also were up there.

I know I'm probably just wishing on a star (and until I see it, I won't believe it), but here's a hope that data like this and from other sources are what get used to make these decisions, not just jacking up/dropping ratings for the sake of "sexy moves" or "balance" or "popularity" or whatever else that doesn't relate to the actual on-the-field/on-tape performance of the player.

[juggalotusx]

Re: Hopefully, Madden Uses Data Like This

Im sure they wont look at things like this when making roster adjustments,couple clicks here a couple clicks there we got a new rating.

[Big FN Deal]

Re: Hopefully, Madden Uses Data Like This

Reading this thread instantly made me remember this:

"The NFL has been tracking player movement during Thursday night games this season, with the goal of amassing detailed information that could give the league, its partners and its fans a better understanding of the physics and strategy of football, CBS Sports reported recently. According to a developer at Madden studio EA Tiburon, the data "would potentially revolutionize" the way the team comes up with player ratings, perennially one of the game's most controversial elements."

http://www.polygon.com/2012/12/28/3811280/nfl-player-movement-data-madden-ratings

And hoping it transitions to some form of this:

"I'm also really proud of our partnership with Synergy Sports Technologies. They are a great partner and I feel privileged that we have exclusive access to the same data that all 30 NBA teams use for scouting teams and players. This is the data they use in order to prepare for upcoming games, finding the advantages or insights needed to try and gain the advantage on their opponent. I want NBA LIVE to be relevant to the NBA fan throughout the entire NBA season, providing data and content updates that keep your gaming experience fresh and in lockstep with everything that's going on in the real world NBA. I want to remove subjectivity from how we rate players and have that data provided by Synergy Sports to completely drive how the game plays. I'm really excited about being able to update our game within an hour of something happening in the NBA. We're going to be making an enormous investment into this connected experience."

http://easports.com/nba-live/news-updates-gameplay/article/nba-live-14-xbox-one-ps-4-e3

We can only hope this is the next step for player ratings in all sports games at some point, using real world data to program virtual athletes.

[kmart2180]

Re: Hopefully, Madden Uses Data Like This

I cannot stand when Patrick Willis gets a broken tackle animation from a little back (Don't like anyone to break a tackle from him really) but when someone like Charles or Spiller breaks there tackle it really pisses me off.

I wouldn't mind if they used there speed and juke abilities to get by Willis, but the way the game allows the defender to attempt a tackle and then the player jukes out of it is insane and should not happen. IMO if Willis or Bowman get a hold of you 99% of the time it should result in a tackle or at least a holdup for a assisted tackle.

[juggalotusx]

Re: Hopefully, Madden Uses Data Like This

I cannot stand when Patrick Willis gets a broken tackle animation from a little back (Don't like anyone to break a tackle from him really) but when someone like Charles or Spiller breaks there tackle it really pisses me off.

I wouldn't mind if they used there speed and juke abilities to get by Willis, but the way the game allows the defender to attempt a tackle and then the player jukes out of it is insane and should not happen. IMO if Willis or Bowman get a hold of you 99% of the time it should result in a tackle or at least a holdup for a assisted tackle.

if you watch when players try to do a "gang tackle" now in 13 only one player will actually grab the player and tackle them, the rest just fall around the player never actually pulling the player down. Thats why assists wont read its because there not allowed to tackle, kinda a first come first serve tackle system this year. Looks like the same thing for madden 25 gang tackles no longer exist.

[KBLover]

Re: Hopefully, Madden Uses Data Like This

I wouldn't mind if they used there speed and juke abilities to get by Willis, but the way the game allows the defender to attempt a tackle and then the player jukes out of it is insane and should not happen. IMO if Willis or Bowman get a hold of you 99% of the time it should result in a tackle or at least a holdup for a assisted tackle.

Exactly - the game just either can not accommodate that type of running logic (since everything seems predicated on breaking or disrupting tackle attempts, not avoiding tackle attempts altogether) or it wouldn't allow for the "combos" or other button pressing stuff they want to do.

And it's not just your opinion. If you look at the charts, guys who spend time recording events agree that Willis is 98% (literally) successful when he gets the chance to take down a ball carrier. Bowman didn't register on the top or bottom, but a fail rate of 3 to 10% means he's at LEAST 90% successful.

Tackles just are NOT broken in the NFL as much as the highlight reels want to make it look like. The worst person on the defense among starters was Ed Reed - and he still succeeded more than 3 out of 4 times.

And the fact the same guys who scored well on not missing tackles were noted as perhaps not having the range to get in "tackling situations" (a lot like how some fielders in baseball don't make errors because they can't get close enough to the ball to screw up) just shows how ball carriers try to avoid tackles, not break contact constantly.

if you watch when players try to do a "gang tackle" now in 13 only one player will actually grab the player and tackle them, the rest just fall around the player never actually pulling the player down. Thats why assists wont read its because there not allowed to tackle, kinda a first come first serve tackle system this year. Looks like the same thing for madden 25 gang tackles no longer exist.

This definitely happens, but what also happens is when the engine recognizes "the act of being tackled". I've seen one of my LBs fighting with a ball carrier and then another finishes everything off (not just falls down but actually hits the engagement and ends the play). That should be an assist for both (like it was in M12, actually). Same for when two (or more) defenders contact the ball carrier nearly simultaneously and the ball carrier goes down. That can happen in M13, but only one tackle is recorded.

Instead, assists often happen in some sort of weird event where the ball carrier is still considered live but "in the process of being tackled" and then if he's actually finished off, an assist often occurs. But it seems to be when the ball carrier is on another body (so is not technically down by contact, but is not actually up and "running free". So the game is about to record assists for whoever comes and puts the ball carrier on the ground.

I have seen a rare few assists outside of those situations, but I can't isolate the events that made it occur. I don't know if it's literally exactly simultaneous (i.e. both hit down to the millisecond instead of "close enough to the naked eye" like a human scorer would do it), or another odd alignment of the digital stars at work.

[juggalotusx]

Re: Hopefully, Madden Uses Data Like This

Exactly - the game just either can not accommodate that type of running logic (since everything seems predicated on breaking or disrupting tackle attempts, not avoiding tackle attempts altogether) or it wouldn't allow for the "combos" or other button pressing stuff they want to do.

And it's not just your opinion. If you look at the charts, guys who spend time recording events agree that Willis is 98% (literally) successful when he gets the chance to take down a ball carrier. Bowman didn't register on the top or bottom, but a fail rate of 3 to 10% means he's at LEAST 90% successful.

Tackles just are NOT broken in the NFL as much as the highlight reels want to make it look like. The worst person on the defense among starters was Ed Reed - and he still succeeded more than 3 out of 4 times.

And the fact the same guys who scored well on not missing tackles were noted as perhaps not having the range to get in "tackling situations" (a lot like how some fielders in baseball don't make errors because they can't get close enough to the ball to screw up) just shows how ball carriers try to avoid tackles, not break contact constantly.



This definitely happens, but what also happens is when the engine recognizes "the act of being tackled". I've seen one of my LBs fighting with a ball carrier and then another finishes everything off (not just falls down but actually hits the engagement and ends the play). That should be an assist for both (like it was in M12, actually). Same for when two (or more) defenders contact the ball carrier nearly simultaneously and the ball carrier goes down. That can happen in M13, but only one tackle is recorded.

Instead, assists often happen in some sort of weird event where the ball carrier is still considered live but "in the process of being tackled" and then if he's actually finished off, an assist often occurs. But it seems to be when the ball carrier is on another body (so is not technically down by contact, but is not actually up and "running free". So the game is about to record assists for whoever comes and puts the ball carrier on the ground.

I have seen a rare few assists outside of those situations, but I can't isolate the events that made it occur. I don't know if it's literally exactly simultaneous (i.e. both hit down to the millisecond instead of "close enough to the naked eye" like a human scorer would do it), or another odd alignment of the digital stars at work.

I am 100% sure you think of these things more than 90% of the EA Madden staff. I am almost positive they dont think this hard about the game at all..

[HeavyHitter55]

Re: Hopefully, Madden Uses Data Like This

Here's the broken (for offense) and missed (for defense) tackles rates for last season.

http://www.footballoutsiders.com/stat-analysis/2013/broken-tackles-2012

http://www.footballoutsiders.com/stat-analysis/2013/broken-tackles-2012-defense

At the very least, use it for ratings and representing how guys play. Guys like Foster produce if nothing else for stamina and consistency, not stiff-arming and spinning and juking their way down the field.

Hopefully, quick guys are quick and not having sky high "moves" ratings. These are guys that you shouldn't get a lot of tackle attempts against because it's hard to get in position to attack them. They use vision to find daylight - not get swarmed and use a ton of "moves" to escape (those kinds of runs are few and far between) as if it's the new "Pro-Tak moment".

Hopefully, the TAK/POW ratings will be represented by data as well and that elite TAK actually looks and feels like it. Not seeing 90+ TAK guys "bounce" off HBs to emulate "elusiveness". ("Elude" doesn't not mean take contact and bounce off it...)

Hopefully not all DBs will get crap TAK ratings. However, hopefully STR actually matters so that DBs tackle guys they typically do (and the most physical stand out over the cover guys - likewise for wide outs, separate the pure speed merchants vs the powerful WR vs the dangerous speed-power WRs)

Hopefully, the "Run Free" doesn't turn this all into more of mess than it is now. Hopefully, it's just a way for the ball carriers to display their running abilities (power or quickness or both if so gifted) within reason and with fewer animation limits while defenders that are excellent also can still make plays. There's elusive and quick and powerful ball carriers in the NFL, yet Bruce Carter had a 100% success rate and Willis is there again with his 98%. There's DBs that also were up there.

I know I'm probably just wishing on a star (and until I see it, I won't believe it), but here's a hope that data like this and from other sources are what get used to make these decisions, not just jacking up/dropping ratings for the sake of "sexy moves" or "balance" or "popularity" or whatever else that doesn't relate to the actual on-the-field/on-tape performance of the player.

With the sharing capabilities, this can be fixed. It's just going to take an effort on the part of those who want it to be properly reflected.

[infemous]

Re: Hopefully, Madden Uses Data Like This

Again, this is entirely dependent on Madden devs knowing what each rating and each slider actually does.

When I discovered that Elusiveness is more important than any 'move' when it comes to breaking tackles etc. I felt robbed. I used to always lower my shoulder with Steven Jackson and his 97 TRK and 83 STR, trying to plow through a loaded box, only to see him brought down.

I do a quick flick of the right stick in a loaded box with Steven Grant and see him break 2 tackles and hit daylight for a touchdown. The 'elusive' backs are better at trucking and slipping tackles than the guys with the actual attributes for it.

Until EA rectify this and either remove the elusive rating, or define it and manifest it in a realistic manner, then the game will not be able to work properly.

Likewise, the way that fumbles are more based on the player who is doing the hitting's ability than it is the ball carrier's makes the Carry rating worthless. It doesn't matter if Steven Jackson has 97 truck, 83 strength and 95 carry, because if Patrick Willis is hitting him and he isn't bracing for the hit, that ball is coming out. Meanwhile, I can juke Willis into sliding off my RB with a 60 STR 90 ELU 5,9 185lbs RB...

I am going to continue to implore and request this, but EA, you HAVE to tell us what each rating does and means for each individual position.

If you tell us, we can then edit the rosters to make the game play properly, OR we can give you feedback in ways to address how you not only rate, but how each player plays. We can and are willing to do so because you obviously aren't.

Then we want to know what each slider does too.

This is the LEAST you can do EA. The very least.

[KBLover]

Re: Hopefully, Madden Uses Data Like This

Until EA rectify this and either remove the elusive rating, or define it and manifest it in a realistic manner, then the game will not be able to work properly.

Absolutely - and I think it should just be gone.

After all, an "elusive" back is one that is good at jukes and spins and stutter steps to throw off pursuit. Then they typically have good explosiveness to gain the most separation from that moment's hesitation by the defender.

It really illustrates the issue with the engine, IE or not. It still can not resolve non-contact situations with the subtlety of the real game. It still doesn't deal with the fraction of a seconds here and there that can make all the difference. Instead, you have to use a "move" and "fake out" half the defense or guys who have no business buying a "fake" but just reaching out and tackling.

IE might make contact situations resolve better, but what about avoidance/absence of contact?

EA has to rely on a crutch like ELU and the move ratings in general instead of letting the HB do his own thing by his own changing of direction and control of his speed and letting AGI and ACC help elusive backs while STR and Size and enough AGI to avoid head-on/direct contact (even AP doesn't just run right into people all the time, he gives them a shoulder to hit, deflects contact just enough to keep his leverage at the advantage, etc) to help power backs.

When I play, I avoid using the "move" buttons both on offense and defense, and it's possible to achieve everything but a spin. Even on pass rushing, the "moves" should be how the defender moves against how the blocker moves with BSH, STR, Size and AGI making the difference for the defender and blocking technique plus the rest for the blocker. Pass rushing got more fun for me once I figured a way to move and "fight" (such that it is) without having to rely on the "move" working or not, especially with my slider for pass rushing - thing is - that's just the way it should work and "moves" should trigger if the player knows them and when the situation plus L-stick movement puts the player in position to use it.

[tarantism]

Re: Hopefully, Madden Uses Data Like This

Great thread as usual, KB. The ratings structure is one of the many things in Madden that need a complete overhaul and a fresh approach.