Tom Brady: The Game Following a Bad Performance

[ivanvamp]

One of the memes going around is that Tom Brady doesn't have two bad games in a row. Is that really true? Let's make this scale (this isn't a "Tom Brady" scale, which is higher than most QBs...it's an "objective" scale that applies to all QBs):

QB Rating - Category
115+ - excellent
100-114.9 - very good
85-99.9 - good
75-84.9 - average
65-74.9 - poor
<65 - terrible

Tom Brady, in his entire career, has played in 181 games (regular season + post season). Here's the number of games he's played in each of those categories:

excellent: 41 (22.7%)
very good: 41 (22.7%)
good: 26 (14.4%)
average: 29 (16.0%)
poor: 19 (10.5%)
terrible: 25 (13.8%)

Now, let's focus on just the terrible games, since what we're looking for is how Brady does when he comes off a terrible game. Last week's AFCCG qualifies, by QB rating, as a "terrible" game (57.5). So let's zero in on those 25 terrible games. How has Brady done in the games following those 25 terrible games? Well, in 3 of them, he didn't have a "next" game (obviously he hasn't had a "next" game yet after this most recent one, once came in the 2009 wild card game against Baltimore so that ended the season, and the other was his short appearance in his 2000 debut). So there were really 22 terrible games where he then played a next game.

Of those 22, here's where his next games ended up on this scale:

excellent: 4 (18.2%)
very good: 3 (13.6%)
good: 4 (18.2%)
average: 7 (31.8%)
poor: 2 (9.1%)
terrible: 2 (9.1%)

Here were the actual QB ratings in his games following a terrible performance:

149.0
124.4
116.4
115.2
108.8
105.8
101.5
93.4
93.4
91.6
87.1
84.6
84.2
82.6
79.7
79.6
79.5
75.4
70.4
68.5
63.6
57.6

11 of them were over an 85 rating, and 11 of them were under. The average QB rating for Brady following a terrible game is 91.5.

On 3 of these 22 occasions Brady's terrible performance was followed up by a playoff game. Here was Brady's QB rating in the playoffs following a terrible performance (either in a previous playoff game or in a week 16 game):

116.4 (2005 - WC win over Jax, 28-3)
79.5 (2006 - AFCCG loss to Ind, 38-34)
70.4 (2001 - Divisional win over Oak, 16-13)

So here's the history: Brady is a great quarterback. He normally plays very well. His career playoff QB rating is significantly lower than his career regular season QB rating (87.6 vs. 96.4). And he really has been about 50/50 in terms of following up a terrible performance over the course of his career.

If you expand this study to look at how he does following up a game where he is anything less than "good" (i.e., he has an "average", "poor", or "terrible" game), the numbers are more encouraging. Following up any of these three levels, he has done this in his next game (65 occasions):

excellent: 14 (21.5%)
very good: 15 (23.1%)
good: 9 (13.8%)
average: 14 (21.5%)
poor: 6 (9.2%)
terrible: 7 (10.8%)

So he follows up a less than good game with a good, very good, or excellent game 58.5% of the time. So he's a better than average bet to play pretty well. But it isn't nearly the slam-dunk that most Pats fans think, unfortunately.

 

[PATRIOTSFANINPA]

I would like to see what Brady's performance is coming off a bad performance followed by a week off to prepare for the next game?

A HOF QB given some extra time to prepare has to be a decent improvement 2 weeks later.

 

[Ring 6]

One of the memes going around is that Tom Brady doesn't have two bad games in a row. Is that really true? Let's make this scale (this isn't a "Tom Brady" scale, which is higher than most QBs...it's an "objective" scale that applies to all QBs):

QB Rating - Category
115+ - excellent
100-114.9 - very good
85-99.9 - good
75-84.9 - average
65-74.9 - poor
<65 - terrible

Tom Brady, in his entire career, has played in 181 games (regular season + post season). Here's the number of games he's played in each of those categories:

excellent: 41 (22.7%)
very good: 41 (22.7%)
good: 26 (14.4%)
average: 29 (16.0%)
poor: 19 (10.5%)
terrible: 25 (13.8%)

Now, let's focus on just the terrible games, since what we're looking for is how Brady does when he comes off a terrible game. Last week's AFCCG qualifies, by QB rating, as a "terrible" game (57.5). So let's zero in on those 25 terrible games. How has Brady done in the games following those 25 terrible games? Well, in 3 of them, he didn't have a "next" game (obviously he hasn't had a "next" game yet after this most recent one, once came in the 2009 wild card game against Baltimore so that ended the season, and the other was his short appearance in his 2000 debut). So there were really 22 terrible games where he then played a next game.

Of those 22, here's where his next games ended up on this scale:

excellent: 4 (18.2%)
very good: 3 (13.6%)
good: 4 (18.2%)
average: 7 (31.8%)
poor: 2 (9.1%)
terrible: 2 (9.1%)

Here were the actual QB ratings in his games following a terrible performance:

149.0
124.4
116.4
115.2
108.8
105.8
101.5
93.4
93.4
91.6
87.1
84.6
84.2
82.6
79.7
79.6
79.5
75.4
70.4
68.5
63.6
57.6

11 of them were over an 85 rating, and 11 of them were under. The average QB rating for Brady following a terrible game is 91.5.

On 3 of these 22 occasions Brady's terrible performance was followed up by a playoff game. Here was Brady's QB rating in the playoffs following a terrible performance (either in a previous playoff game or in a week 16 game):

116.4 (2005 - WC win over Jax, 28-3)
79.5 (2006 - AFCCG loss to Ind, 38-34)
70.4 (2001 - Divisional win over Oak, 16-13)

So here's the history: Brady is a great quarterback. He normally plays very well. His career playoff QB rating is significantly lower than his career regular season QB rating (87.6 vs. 96.4). And he really has been about 50/50 in terms of following up a terrible performance over the course of his career.

If you expand this study to look at how he does following up a game where he is anything less than "good" (i.e., he has an "average", "poor", or "terrible" game), the numbers are more encouraging. Following up any of these three levels, he has done this in his next game (65 occasions):

excellent: 14 (21.5%)
very good: 15 (23.1%)
good: 9 (13.8%)
average: 14 (21.5%)
poor: 6 (9.2%)
terrible: 7 (10.8%)

So he follows up a less than good game with a good, very good, or excellent game 58.5% of the time. So he's a better than average bet to play pretty well. But it isn't nearly the slam-dunk that most Pats fans think, unfortunately.

Nice woirk, but it must be noted that passer rating is a horrible scale of actual QB play. But I recognize its the only number available to you for the purpose you are trying to use it.

 

[BlueThunder]

One has to wonder how his injured non-throwing shoulder affects his performance.

It has been reported that he requested his QB trainer to help him out and analyze his throwing mechanics.

 

[everlong]

Nice work. There's still a lot of variables. Who was the opponent the next week? If you're coming off of a bad performance and you get a top defense you might only put up a 85 QBR but you played really well. When in the career did it happen? One of the examples is Oakland back in 01 when he was basically a rookie and it was snowing like crazy.

I use QBR more than any other stat but unless you watch the actual games and get a feel for how the QB actually played it's not complete.

 

[TruthSeeker]

Stats-smats, It's pretty simple. If Brady plays like he did against the Broncos, we win, if he plays like he did against the Ravens, we lose.

 

[FCB02062]

more importantly, how does Brady perform in leap years????

 

[ivanvamp]

I would like to see what Brady's performance is coming off a bad performance followed by a week off to prepare for the next game?

A HOF QB given some extra time to prepare has to be a decent improvement 2 weeks later.

Brady has only had two "terrible" games which were followed by a bye week before his next game. Here they were:

2001 - week 16, he went 17-29 (58.6%), 198 yds, 1 td, 2 int, 62.1 rating against the Panthers. They earned a bye for the playoffs as the #2 seed, then hosted the Raiders in the Snow Bowl game. For that game, Brady had this line: 32-52 (61.5%), 312 yds, 0 td, 1 int, 70.4 rating.

2002 - week 6, he put up this line in a 28-10 loss to the Packers: 24-44 (54.5%), 183 yds, 1 td, 3 int, 44.0 rating. They had a bye for week 7, then they lost to Denver in week 8, 24-16, with Brady putting up this line: 15-29 (51.7%), 130 yds, 1 td, 0 int, 75.4 rating.

In 2007 he had a "bad" game in the AFCCG against SD in a 21-12 win: 22-33 (66.7%), 209 yds, 2 td, 3 int, 66.4 rating. They had the bye then played the Giants in the SB, where Brady put up this line: 29-48 (60.4%), 266 yds, 1 td, 0 int, 82.5 rating.

Here are all of Brady's stat lines the week after a bye. I'll designate which games followed anything less than a "good" performance, according to my rubric (marked with an *):

2001 - Divisional Round - 32-52 (61.5%), 312 yds, 0 td, 1 int, 70.4 rating - beat Oak 16-13*

2001 - Super Bowl - 16-27 (59.3%), 145 yds, 1 td, 0 int, 86.2 rating - beat StL 20-17

2002 - Week 8 - 15-29 (51.7%), 130 yds, 1 td, 0 int, 75.4 rating - lost to Den 24-16*

2003 - Week 10 - 15-34 (44.1%), 212 yds, 0 td, 0 int, 64.8 rating - beat Dal 12-0

2003 - Divisional Round - 21-41 (51.2%), 201 yds, 1 td, 0 int, 73.3 rating - beat Ten 17-14

2003 - Super Bowl - 32-48 (66.7%), 354 yds, 3 td, 1 int, 100.5 rating - beat Car 32-29*

2004 - Week 3 - 17-30 (56.7%), 298 yds, 2 td, 0 int, 112.9 rating - beat Buf 31-17*

2004 - Super Bowl - 23-33 (69.7%), 236 yds, 2 td, 0 int, 110.2 rating - beat Phi 24-21

2005 - Week 7 - 14-21 (66.7%), 199 yds, 1 td, 0 int, 113.0 rating - beat Buf 21-16*

2006 - Week 6 - 18-27 (66.7%), 195 yds, 2 td, 0 int, 112.4 rating - beat Buf 28-6

2007 - Week 10 - 31-39 (79.5%), 373 yds, 5 td, 0 int, 146.1 rating - beat Buf 56-10

2007 - Divisional Round - 26-28 (92.9%), 262 yds, 3 td, 0 int, 141.4 rating - beat Jax 31-20

2007 - Super Bowl - 29-48 (60.4%), 266 yds, 1 td, 0 int, 82.5 rating - lost to NYG 17-14*

2008 - N/A

2009 - Week 8 - 25-37 (67.6%), 332 yds, 1 td, 1 int, 93.5 rating - beat Mia 27-17

2010 - Week 5 - 27-44 (61.4%), 292 yds, 1 td, 2 int, 69.5 rating - beat Bal 23-20

2010 - Divisional Round - 29-45 (64.4%), 299 yds, 2 td, 1 int, 89.0 rating - lost to NYJ 28-21

2011 - Week 7 - 24-35 (68.6%), 198 yds, 2 td, 0 int, 101.8 rating - lost to Pit 25-17

2011 - Divisional Round - 26-34 (76.5%), 363 yds, 6 td, 1 int, 137.6 rating - beat Den 45-10

That's 18 games following a bye. Of these 18, here's where they would fit in my rubric:

excellent: 3 (16.7%)
very good: 6 (33.3%)
good: 3 (16.7%)
average: 2 (11.1%)
poor: 3 (16.7%)
terrible: 1 (5.6%)

So 12 of the 18 (66.7%) were "good" or better.

Of these 18, just 6 of them followed a performance that was "average" or worse. Here were the QB ratings in those 6 instances: 70.4, 75.4, 100.5, 112.9, 113.0, and 82.5.

So again, a bit of a mixed bag. He's never been hideous in that situation, but he's never been insanely brilliant either. He's either been below average or very good. 50-50 proposition, really.

Wish I had better news to share!

 

[ivanvamp]

Nice work. There's still a lot of variables. Who was the opponent the next week? If you're coming off of a bad performance and you get a top defense you might only put up a 85 QBR but you played really well. When in the career did it happen? One of the examples is Oakland back in 01 when he was basically a rookie and it was snowing like crazy.

I use QBR more than any other stat but unless you watch the actual games and get a feel for how the QB actually played it's not complete.

No doubt. And a QB can play really well but still finish with just a so-so QB rating. For example, let's say Brady hits 90% of his completions and gets the team down to the 1 every time, but they run the ball in from there. He won't have the TD on his stat line which really would have boosted his QB rating.

So to you, and to Andy, I agree, it's not a crystal clear picture. But it really is about all I have available, unfortunately. Baseball stat geeks have done a ridiculously great job of breaking down everything that happens on a field and the stats are incredible. There is something called "linear weights" that quantifies everything. I wish there was something like that for football but, to my knowledge, there is not. So this is all I have to go with. Sorry!

 

[SHOWTIME15]

I would like to see what Brady's performance is coming off a bad performance followed by a week off to prepare for the next game?

A HOF QB given some extra time to prepare has to be a decent improvement 2 weeks later.

Brady played bad in the 2007 AGCCG, he played bad in the super bowl 2 weeks later. I do remember him sitting out the first week of practice prior to the super bowl, probably with that ankle sprain.

 

[oldrover]

Nice work on compiling the info. Food for thought.

FYI... for the Rams Super Bowl, there was no bye week because of 9/11's effect on the season.

 

[BigHoss28]

Stats-smats, It's pretty simple. If Brady plays like he did against the Broncos, we win, if he plays like he did against the Ravens, we lose.

Disagree... The Ravens are just as good as the Giants in my opinion.

 

[Theplayer]

all depends on the defense he is playing...shouldn't that be obvious? I think my head will explode if I hear Brady doesn't have 2 bad games in a row 1 more time.

 

[ALP]

thanks for the analysis

 

[Seacoast Fan]

Brady has only had two "terrible" games which were followed by a bye week before his next game. Here they were:

2001 - week 16, he went 17-29 (58.6%), 198 yds, 1 td, 2 int, 62.1 rating against the Panthers. They earned a bye for the playoffs as the #2 seed, then hosted the Raiders in the Snow Bowl game. For that game, Brady had this line: 32-52 (61.5%), 312 yds, 0 td, 1 int, 70.4 rating.

2002 - week 6, he put up this line in a 28-10 loss to the Packers: 24-44 (54.5%), 183 yds, 1 td, 3 int, 44.0 rating. They had a bye for week 7, then they lost to Denver in week 8, 24-16, with Brady putting up this line: 15-29 (51.7%), 130 yds, 1 td, 0 int, 75.4 rating.

In 2007 he had a "bad" game in the AFCCG against SD in a 21-12 win: 22-33 (66.7%), 209 yds, 2 td, 3 int, 66.4 rating. They had the bye then played the Giants in the SB, where Brady put up this line: 29-48 (60.4%), 266 yds, 1 td, 0 int, 82.5 rating.

Here are all of Brady's stat lines the week after a bye. I'll designate which games followed anything less than a "good" performance, according to my rubric (marked with an *):

2001 - Divisional Round - 32-52 (61.5%), 312 yds, 0 td, 1 int, 70.4 rating - beat Oak 16-13*

2001 - Super Bowl - 16-27 (59.3%), 145 yds, 1 td, 0 int, 86.2 rating - beat StL 20-17

2002 - Week 8 - 15-29 (51.7%), 130 yds, 1 td, 0 int, 75.4 rating - lost to Den 24-16*

2003 - Week 10 - 15-34 (44.1%), 212 yds, 0 td, 0 int, 64.8 rating - beat Dal 12-0

2003 - Divisional Round - 21-41 (51.2%), 201 yds, 1 td, 0 int, 73.3 rating - beat Ten 17-14

2003 - Super Bowl - 32-48 (66.7%), 354 yds, 3 td, 1 int, 100.5 rating - beat Car 32-29*

2004 - Week 3 - 17-30 (56.7%), 298 yds, 2 td, 0 int, 112.9 rating - beat Buf 31-17*

2004 - Super Bowl - 23-33 (69.7%), 236 yds, 2 td, 0 int, 110.2 rating - beat Phi 24-21

2005 - Week 7 - 14-21 (66.7%), 199 yds, 1 td, 0 int, 113.0 rating - beat Buf 21-16*

2006 - Week 6 - 18-27 (66.7%), 195 yds, 2 td, 0 int, 112.4 rating - beat Buf 28-6

2007 - Week 10 - 31-39 (79.5%), 373 yds, 5 td, 0 int, 146.1 rating - beat Buf 56-10

2007 - Divisional Round - 26-28 (92.9%), 262 yds, 3 td, 0 int, 141.4 rating - beat Jax 31-20

2007 - Super Bowl - 29-48 (60.4%), 266 yds, 1 td, 0 int, 82.5 rating - lost to NYG 17-14*

2008 - N/A

2009 - Week 8 - 25-37 (67.6%), 332 yds, 1 td, 1 int, 93.5 rating - beat Mia 27-17

2010 - Week 5 - 27-44 (61.4%), 292 yds, 1 td, 2 int, 69.5 rating - beat Bal 23-20

2010 - Divisional Round - 29-45 (64.4%), 299 yds, 2 td, 1 int, 89.0 rating - lost to NYJ 28-21

2011 - Week 7 - 24-35 (68.6%), 198 yds, 2 td, 0 int, 101.8 rating - lost to Pit 25-17

2011 - Divisional Round - 26-34 (76.5%), 363 yds, 6 td, 1 int, 137.6 rating - beat Den 45-10

That's 18 games following a bye. Of these 18, here's where they would fit in my rubric:

excellent: 3 (16.7%)
very good: 6 (33.3%)
good: 3 (16.7%)
average: 2 (11.1%)
poor: 3 (16.7%)
terrible: 1 (5.6%)

So 12 of the 18 (66.7%) were "good" or better.

Of these 18, just 6 of them followed a performance that was "average" or worse. Here were the QB ratings in those 6 instances: 70.4, 75.4, 100.5, 112.9, 113.0, and 82.5.

So again, a bit of a mixed bag. He's never been hideous in that situation, but he's never been insanely brilliant either. He's either been below average or very good. 50-50 proposition, really.

Wish I had better news to share!

Nice work, but there are still a few too many variables...Can you also please account for: home v. away v. neutral site; dome v. outside; for outside games, weather conditions according to the following categories: temperature, rain, snow; game time; number of Pro Bowl players on each opposing teams DL; composition of Pats OL for each game; composition of Pats receiving corps for each game; identity of Pats Offensive coordinator; Bridgette v. Giselle v. just some groupie; and breakfast cereal consumed on game day.

I will check back in about 20 mins.

 

[Seacoast Fan]

No doubt. And a QB can play really well but still finish with just a so-so QB rating. For example, let's say Brady hits 90% of his completions and gets the team down to the 1 every time, but they run the ball in from there. He won't have the TD on his stat line which really would have boosted his QB rating.

So to you, and to Andy, I agree, it's not a crystal clear picture. But it really is about all I have available, unfortunately. Baseball stat geeks have done a ridiculously great job of breaking down everything that happens on a field and the stats are incredible. There is something called "linear weights" that quantifies everything. I wish there was something like that for football but, to my knowledge, there is not. So this is all I have to go with. Sorry!

Stop apologizing and start working...you've been given your assignment, now MOVE!

Just kidding, of course...great thread, a ton of work and time went into it. I say this alot here, but its threads like this that keep me coming to this board. I dont always agree with the posts, and sometimes the posts show, like yours did, that things are not alway clear cut, but its a great read anyway! thanks.

 

[Seacoast Fan]

Disagree... The Ravens are just as good as the Giants in my opinion.

Agreed. While the Giants O may be better, you can make a strong case their D is not as good as the Ravens.

 

[ivanvamp]

Stop apologizing and start working...you've been given your assignment, now MOVE!

Just kidding, of course...great thread, a ton of work and time went into it. I say this alot here, but its threads like this that keep me coming to this board. I dont always agree with the posts, and sometimes the posts show, like yours did, that things are not alway clear cut, but its a great read anyway! thanks.

Thanks for the kind words. This stuff is very fun for me. If you can't tell.

 

[PatriotSeven]

Those are passer ratings, not quarterback ratings. Brady had a bad passer rating. He played great as a quarterback against the Ravens.

He tossed a meaningless interception which hurt his passer rating, but was the right risk to take, and he still scored the winning touchdown himself. The other bounced off of his receivers hands. That wasn't even on him.

This is BB and Brady playing with the minds of the Giants.

 

[Palm Beach Pats Fan]

He tossed a meaningless interception which hurt his passer rating, but was the right risk to take

Say WHAT?

The bomb to Slater was a terrible risk to take. You have the ball at mid-field, mid 4th Q, the other team has no quick strike offense, so you methodically take the ball down the field instead of madly bombing into double coverage at a guy with one catch on the year.

Horrible risk to take, one of the worst play calls ever, and you are the first person I know of that things it was the right play call.