- Joined
- Mar 19, 2006
- Messages
- 33,964
- Reaction score
- 14,418
A lot's been said about the strength of the Pats' victims er opponents this year. First, since they played the Pats, each of them must have already lost at least 1 game (out of either 4 or 5.)
But what if each team that played the Pats is almost destined to lose after playing the Pats?
Going back to Week 4 of 2006 (the Cincinatti game,) the following stats pertain:
- excluding the last game of the year (it's hardly relevant what happened in the previous season's last game,) and excluding last week's game (they have not played again yet) 13 team have lost to the Pats, and then played a game within 1 or 2 weeks afterward.
- Of those 13 games, Pats' opponents have a record of 4-9 in their post-Pats game.
- Even when "strength of schedule" is weak at the start of the season, opponents rarely add up to under 45% wins.
- After losing to the Pats, teams achieve a 30.8% winning percentage. (4/13)
- The Pats' strength of sched this year going in was about 55%, rather than 45%.
-----------------------------------------------
In other words, a team is much more likely to lose again, after losing to New England.
So, in retrospect, each team had a 100% chance of losing to NE, and we could expect about a 69% chance (better than 2 out of 3 chance) that they would lose the following game, assuming that the "beatdown factor" actually exists. If it does exist, one could easily explain it many ways: it could be the result of losing at all, or it could be the result of BB exposing a team's weaknesses, for the next team to exploit, or it could simply be luck, that a team always seems to be going through a rough patch when it plays NE because God loves us.
I favor the "BB exposes them for the next team's benefit" theory.
There are 2 flaws here, of course:
1) small sample size, which is unavoidable, although it could be addressed if someone wants to check out similar data dating back to BB's first year, and
2) the question of what usually happens after a team beats you. It would seem that "bouncebacks" are at least as likely as "losing streaks," particularly since over time the teams involved would average somewhere around .500 records. I simply don't have these data, but I don't think that a team that loses is more or less likely to lose again; thinking about the Pats -- admittedly a good team -- they had a streak of something like 6 years before back-to-back losses actually happened to them in 2006; it was certainly not a near-unavoidable outcome.
Preliminarily, it seems likely that our five unfortunate adversaries in the 2007 season have to be judged not only with a bit of understanding in the face of a loss to the Pats, but also with a nod toward the notion that they are statistically unlikely to win the week following a Pats beatdown.
So first let's look at records absent the loss to the Pats. Had they not played us, their records would be:
NYJ 1-3
SD 2-2
BUF 1-3
CIN 1-2
CLE 2-2
----------
7-12
Certainly not awe-inspiring either. But let's consider their chances in their next game. In a normal game, their chances would average out to 50%. Going in, you would think their chances higher, since you are looking at 3 teams with playoff potential at the start of the season. But we'll say 50%.
After playing the Pats, we're saying that their 1 in 2 chance becomes a 1 in 3 chance. since Cleveland has not played an opponent after us yet, we have 4 games this season, of which we could expect them to win not 2, but 1.33 games total (after the beat-down.) And lo and behold, that's exactly what happened (well, almost exactly, you can't win 1/3 of a game): of our 4 opponents who have played after playing us, they have won a combined total of 1 game.
Had they not been "beaten down", we could expect that 7-12 mark above to be an 8-11 mark.
Now don't get me wrong, that's still bad. But it's more like average bad, far from the horrendous "unadjusted" record of 7-17 those teams have compiled, when you include their thus-far inevitable losses to the NE Patriots juggernaut.
Now, I freely admit this is total homer logic. But I like it. I like it a lot.
PFnV
But what if each team that played the Pats is almost destined to lose after playing the Pats?
Going back to Week 4 of 2006 (the Cincinatti game,) the following stats pertain:
- excluding the last game of the year (it's hardly relevant what happened in the previous season's last game,) and excluding last week's game (they have not played again yet) 13 team have lost to the Pats, and then played a game within 1 or 2 weeks afterward.
- Of those 13 games, Pats' opponents have a record of 4-9 in their post-Pats game.
- Even when "strength of schedule" is weak at the start of the season, opponents rarely add up to under 45% wins.
- After losing to the Pats, teams achieve a 30.8% winning percentage. (4/13)
- The Pats' strength of sched this year going in was about 55%, rather than 45%.
-----------------------------------------------
In other words, a team is much more likely to lose again, after losing to New England.
So, in retrospect, each team had a 100% chance of losing to NE, and we could expect about a 69% chance (better than 2 out of 3 chance) that they would lose the following game, assuming that the "beatdown factor" actually exists. If it does exist, one could easily explain it many ways: it could be the result of losing at all, or it could be the result of BB exposing a team's weaknesses, for the next team to exploit, or it could simply be luck, that a team always seems to be going through a rough patch when it plays NE because God loves us.
I favor the "BB exposes them for the next team's benefit" theory.
There are 2 flaws here, of course:
1) small sample size, which is unavoidable, although it could be addressed if someone wants to check out similar data dating back to BB's first year, and
2) the question of what usually happens after a team beats you. It would seem that "bouncebacks" are at least as likely as "losing streaks," particularly since over time the teams involved would average somewhere around .500 records. I simply don't have these data, but I don't think that a team that loses is more or less likely to lose again; thinking about the Pats -- admittedly a good team -- they had a streak of something like 6 years before back-to-back losses actually happened to them in 2006; it was certainly not a near-unavoidable outcome.
Preliminarily, it seems likely that our five unfortunate adversaries in the 2007 season have to be judged not only with a bit of understanding in the face of a loss to the Pats, but also with a nod toward the notion that they are statistically unlikely to win the week following a Pats beatdown.
So first let's look at records absent the loss to the Pats. Had they not played us, their records would be:
NYJ 1-3
SD 2-2
BUF 1-3
CIN 1-2
CLE 2-2
----------
7-12
Certainly not awe-inspiring either. But let's consider their chances in their next game. In a normal game, their chances would average out to 50%. Going in, you would think their chances higher, since you are looking at 3 teams with playoff potential at the start of the season. But we'll say 50%.
After playing the Pats, we're saying that their 1 in 2 chance becomes a 1 in 3 chance. since Cleveland has not played an opponent after us yet, we have 4 games this season, of which we could expect them to win not 2, but 1.33 games total (after the beat-down.) And lo and behold, that's exactly what happened (well, almost exactly, you can't win 1/3 of a game): of our 4 opponents who have played after playing us, they have won a combined total of 1 game.
Had they not been "beaten down", we could expect that 7-12 mark above to be an 8-11 mark.
Now don't get me wrong, that's still bad. But it's more like average bad, far from the horrendous "unadjusted" record of 7-17 those teams have compiled, when you include their thus-far inevitable losses to the NE Patriots juggernaut.
Now, I freely admit this is total homer logic. But I like it. I like it a lot.
PFnV