If that corner has this thing called str and tkl that 99% of the userbase think is only for D-line u win
Forum > Goal Line Blitz > Judging a CBs performance
deez11010
offline
offline
Originally posted by jampysos
i say recallowed/targets
If a zone defense is called ever, this wont work. Apparently GLB counts a target if a pass is thrown anywhere in the zone the CB is supposed to be covering, even if theres 3 receivers there and he's in position to cover 2 but the pass is thrown to the 1 on the edge of his zone away from him. He'll still get a targeted and reception allowed.
Even if you ran 100% man coverage, it still will be skewed because if a Defender is covering his man underneath and the pass goes to the DBs over the top but the underneath defender jumps up and knocks the ball down, the DB will get a +1 targeted and +0 reception allowed. Even though he did nothing and may have even given up a play.
GLB DB stats are a joke. Only way to see who is better is to watch film on all of them and compare same plays.
i say recallowed/targets
If a zone defense is called ever, this wont work. Apparently GLB counts a target if a pass is thrown anywhere in the zone the CB is supposed to be covering, even if theres 3 receivers there and he's in position to cover 2 but the pass is thrown to the 1 on the edge of his zone away from him. He'll still get a targeted and reception allowed.
Even if you ran 100% man coverage, it still will be skewed because if a Defender is covering his man underneath and the pass goes to the DBs over the top but the underneath defender jumps up and knocks the ball down, the DB will get a +1 targeted and +0 reception allowed. Even though he did nothing and may have even given up a play.
GLB DB stats are a joke. Only way to see who is better is to watch film on all of them and compare same plays.
tautology
offline
offline
Originally posted by deez11010
GLB DB stats are a joke. Only way to see who is better is to watch film on all of them and compare same plays.
Well, not a "joke," they are meaningful....but watching film is really the best way to figure it out.
Originally posted by jdbolick
(Deflections + KLs + INTs) / Targets.
If you are looking for a statistical formula, this is a good one. I might double up the INTs though.
Originally posted by Dpride59
Watch film - you may see a dot dominate like this, in a wl title game too
One man wrecking crew single handedly dominating MPHD
[Demian Maia highlights]
Yep, probably the best CB in WL last season, haven't really watched this season.
GLB DB stats are a joke. Only way to see who is better is to watch film on all of them and compare same plays.
Well, not a "joke," they are meaningful....but watching film is really the best way to figure it out.
Originally posted by jdbolick
(Deflections + KLs + INTs) / Targets.
If you are looking for a statistical formula, this is a good one. I might double up the INTs though.
Originally posted by Dpride59
Watch film - you may see a dot dominate like this, in a wl title game too
One man wrecking crew single handedly dominating MPHD
[Demian Maia highlights]
Yep, probably the best CB in WL last season, haven't really watched this season.
Edited by tautology on Jan 6, 2013 18:25:39
Golan
offline
offline
Targets/Rec allowed have been broken forever. There's just no good way to judge how many of each came in zone or single coverage or double coverage or indeed if the QB made a bad throw you didn't have to defend. It will slap a targeted bump on you anyway. Not to mention, you can get a rec allowed without actually being counted as targeted to further mess up the stat line.
Even PD's are somewhat broken because you'll often see a dot get the first and most important deflection, then some secondary defender (often a zone defender) gets a meaningless PD after, and the last dot is then credited with the PD.
If you actually watch games, the best CB's will give up a catch about 50% of the time in 1vs1 against quality opponents. A CB with 90 jumping in front of a WR with 50 doesn't really seem to matter much if both dots are moving when the throw occurs.
Even in the WL game where Dpride posted replays from, the dot he is raving about didn't have an amazing stat line: (http://goallineblitz.com/game/game.pl?game_id=2209368 ) 13 targets, 7 rec allowed, 1 KL, 4 PD, 8 tackles + 1 missed. Clearly the dot played a great game if you watch the replays though.
Even PD's are somewhat broken because you'll often see a dot get the first and most important deflection, then some secondary defender (often a zone defender) gets a meaningless PD after, and the last dot is then credited with the PD.
If you actually watch games, the best CB's will give up a catch about 50% of the time in 1vs1 against quality opponents. A CB with 90 jumping in front of a WR with 50 doesn't really seem to matter much if both dots are moving when the throw occurs.
Even in the WL game where Dpride posted replays from, the dot he is raving about didn't have an amazing stat line: (http://goallineblitz.com/game/game.pl?game_id=2209368 ) 13 targets, 7 rec allowed, 1 KL, 4 PD, 8 tackles + 1 missed. Clearly the dot played a great game if you watch the replays though.
Originally posted by Diamond Spade
u have to look at film which yello doesnt do since its too time consuming.
You would be laughing at me if you knew how much film I actually DO watch.
My game, every game until its a blow out and then some.
And then usually dozens of plays on the opposing team for the next game.
And thats not even counting R&D journeys into World League replays.
u have to look at film which yello doesnt do since its too time consuming.
You would be laughing at me if you knew how much film I actually DO watch.
My game, every game until its a blow out and then some.
And then usually dozens of plays on the opposing team for the next game.
And thats not even counting R&D journeys into World League replays.
Edited by yello1 on Jan 6, 2013 21:21:59
Originally posted by tautology
watching film is really the best way to figure it out..
Well this is what brings me here.
Watching film there is a CB on my team that drives me insane.
He just gets beat sometimes. My other CBs generally do not get beat other than by a fake. They lose the catch-deflect contest, but they are there tight with their man usually.
And yet numbers wise, the annoying CB makes a boatload of pass deflections as well.
watching film is really the best way to figure it out..
Well this is what brings me here.
Watching film there is a CB on my team that drives me insane.
He just gets beat sometimes. My other CBs generally do not get beat other than by a fake. They lose the catch-deflect contest, but they are there tight with their man usually.
And yet numbers wise, the annoying CB makes a boatload of pass deflections as well.
Originally posted by Icy Warrior
Just curious, you talking about my CB? Because he is in decline now, you know.
Nope. Your guy is good.
And the guy I really want to bottle and sell is Lemon Lime. Man he makes KLs actually work.
Just curious, you talking about my CB? Because he is in decline now, you know.
Nope. Your guy is good.
And the guy I really want to bottle and sell is Lemon Lime. Man he makes KLs actually work.
Dpride59
offline
offline
Originally posted by yello1
Well this is what brings me here.
Watching film there is a CB on my team that drives me insane.
He just gets beat sometimes. My other CBs generally do not get beat other than by a fake. They lose the catch-deflect contest, but they are there tight with their man usually.
And yet numbers wise, the annoying CB makes a boatload of pass deflections as well.
He sucks .. Watch the replays I linked if your dot is moving like that you're good - if he's stuck in quicksand move on
Well this is what brings me here.
Watching film there is a CB on my team that drives me insane.
He just gets beat sometimes. My other CBs generally do not get beat other than by a fake. They lose the catch-deflect contest, but they are there tight with their man usually.
And yet numbers wise, the annoying CB makes a boatload of pass deflections as well.
He sucks .. Watch the replays I linked if your dot is moving like that you're good - if he's stuck in quicksand move on
Originally posted by jdbolick
(Deflections + KLs + INTs) / Targets. It's not perfect, but it's the only way I can think of to measure the impact of the CB himself without so much noise from bad throws and such.
I essentially agree with this. That said, I do think Targets:RecAll is more useful than you suggested in this thread, but ONLY when comparing CBs within a conference, and with a grain of salt. Even so, that probably doesn't tell you much that the ratio you suggested (though INTs is largely dependent on what level of the game we're talking about) doesn't already provide you with.
Actually, I think you could make an argument that (Deflections + KLs + INTs):RecAll might works slightly better because it reduces the factor of QB error and basically asks, if the QB gets the ball in the right area, what percentage of the time does the CB make a play, versus what percentage of the time does the WR make the catch. Incomplete passes where neither player touches the ball get discounted.
In any case, you have to look at the big picture. And replays are important.
EDIT: It occurs to me that the ratio I suggested, though probably good to discount QB error, introduces WR error (e.g., drops) as more of a factor. So maybe (Deflections + KLs + INTs) : (RecAll + Drops), wherein drops are drops by the WR on plays where the CB in question is "targeted".
It'd take way too much time to compute that with the stats formatted as they are, though.
(Deflections + KLs + INTs) / Targets. It's not perfect, but it's the only way I can think of to measure the impact of the CB himself without so much noise from bad throws and such.
I essentially agree with this. That said, I do think Targets:RecAll is more useful than you suggested in this thread, but ONLY when comparing CBs within a conference, and with a grain of salt. Even so, that probably doesn't tell you much that the ratio you suggested (though INTs is largely dependent on what level of the game we're talking about) doesn't already provide you with.
Actually, I think you could make an argument that (Deflections + KLs + INTs):RecAll might works slightly better because it reduces the factor of QB error and basically asks, if the QB gets the ball in the right area, what percentage of the time does the CB make a play, versus what percentage of the time does the WR make the catch. Incomplete passes where neither player touches the ball get discounted.
In any case, you have to look at the big picture. And replays are important.
EDIT: It occurs to me that the ratio I suggested, though probably good to discount QB error, introduces WR error (e.g., drops) as more of a factor. So maybe (Deflections + KLs + INTs) : (RecAll + Drops), wherein drops are drops by the WR on plays where the CB in question is "targeted".
It'd take way too much time to compute that with the stats formatted as they are, though.
Edited by Warhawk on Jan 7, 2013 07:11:00
Edited by Warhawk on Jan 7, 2013 07:10:17
Edited by Warhawk on Jan 7, 2013 07:09:20
Edited by Warhawk on Jan 7, 2013 07:09:08
Edited by Warhawk on Jan 7, 2013 07:05:53
jdbolick
offline
offline
Originally posted by Warhawk
That said, I do think Targets:RecAll is more useful than you suggested in this thread, but ONLY when comparing CBs within a conference, and with a grain of salt.
No, it really doesn't. It doesn't even work well on the same team. The reality is that quantitative analysis of football is incredibly complicated and difficult, whether in a browser simulation or the actual games themselves. It's such an inter-related game where the actions of one player affect another that it's generally pretty misleading to assign value to discrete events. I enjoy the work of K.C. Joyner, PFF, Pro Football Outsiders, etc but as someone who watches an obscene amount of games, their numerical conclusions always ring false to me. According to PFF, Chris Gamble was one of the best corners in the league this season before his injury, but most Panthers fans would think that absurd.
That said, I do think Targets:RecAll is more useful than you suggested in this thread, but ONLY when comparing CBs within a conference, and with a grain of salt.
No, it really doesn't. It doesn't even work well on the same team. The reality is that quantitative analysis of football is incredibly complicated and difficult, whether in a browser simulation or the actual games themselves. It's such an inter-related game where the actions of one player affect another that it's generally pretty misleading to assign value to discrete events. I enjoy the work of K.C. Joyner, PFF, Pro Football Outsiders, etc but as someone who watches an obscene amount of games, their numerical conclusions always ring false to me. According to PFF, Chris Gamble was one of the best corners in the league this season before his injury, but most Panthers fans would think that absurd.
You are not logged in. Please log in if you want to post a reply.
