Last week's post: viewtopic.php?t=505532
If you've been following the column, Yahoo and this analyzer have been pretty close in their predictive ability and the prior week, Yahoo had pretty soundly beaten the analyzer's predictions so I'll keep the comparisons up until there's a solid winner or just for information.
Week 8: Average score 7.8, Median 6. 11 scores of 10 or higher
Yahoo's predictions had 9/10 above average inscore and had 7/11 scores that were 10 or higher.
Mine had 7/10 above average scores and had 5/11 scores that were 10 or higher. Yahoo had listed NO who scored 12, SD who scored 11 and TEN who scored 16 on theirs which I hadn't on my list. I had DAL that scored 8, BUF that scored 23 and MIN that scored 7 that were on mine but not theirs. I had said that I hadn't liked the DAL call or MIN call as I felt that Vick and Newton are holding onto the ball better. Would have liked to bring NO into my top 10.
I think I'd tip towards using Yahoo on a more regular basis. They've done well for weeks now and although the analyzer has come close to their level of predictive ability and sometimes picks a break out game that they would miss, on the whole, it does seem to tip in Yahoo's favor.
However, I think that combining information from both may be your best strategy looking at those that were called on both lists.
Turnovers/Sacks are the most reliable predictor that I can find for above average and breakout games. Other additions into the analysis only served to reduce reliability. The following lists are the rank, the defending team, their matchup for the week, the average sacks/turnovers obtained by the defending team, the same given up by their matchup and then the overall score. These are averages which consider the number of games played as some teams have now gone through bye weeks.
I've previously said that NYJ, BAL and GB are teams that should probably be started every week regardless of matchup. I've also been impressed with NYG, DET, SF and CHI and think they may be good starts regardless of matchup as well. The BYE teams are still listed. The bye 'score' btw, is just the alphabetically best fit for the computer so please ignore it.
1) SF vs. WAS (One of the best defenses with a juicy matchup against a faltering skins)
2) BAL v. PIT (Very different from week 1 but BAL is too studly to sit)
3) NYJ v. BUF (Tougher matchup against a strong BUF team but NYJ too studly to sit)
4) GB v. SD (SD is a mess right now)
5) KC v. MIA (Sneaky pick here but I think an appropriate one)
6) CHI v. PHI (CHI is a strong defense but I lowered the ranking because of recent performance by Vick and company)
7) BUF v. NYJ (Buffalo has had some elite weeks making them a higher rank)
8) HOU v. CLE (Houston a bit of a surprise to me, not sure what to make of it)
9) DAL v. SEA (DAL has to be hungry for a win and SEA is the main course)
10) CIN v. TEN (CIN is turning more and more into an elite or need good reason to sit defense)
Considers: OAK v. DEN - Playing in Oakland, denver with a lot of change and I think Oakland will have fewer turnovers than previously.
PIT v. BAL - PIT is a quality defense, I think they've played much better in the past several weeks. NYG is in my 'would sit for good reason' category and I worry about them vs. NE this weekend. We'll see what Yahoo says but I think this may be a more objective way of doing it than my seeing theirs and commenting. I'd be interested to see if they don't have CHI on their top 10.
PASSING AND RUSHING DEFENSE:
The NFL ranks passing or rushing defense based on yards per game. However, from a fantasy perspective, TDs and turnovers are weighted much more heavily than yards in most fantasy leagues. Thus, the NFL's ranking doesn't always make sense. The following fantasy passing ranking is based on 1 point per 25 yards, 6 per TD and -2 per turnover. It does NOT take into account the matchup for the week.
Each line shows my rank, the team, the NFL's rank, the average score they are yielding opposing passers, and the toughness of their schedule so far this year. For example, the NYJ are my number 1 passing defense, they are the NFL's number 7 ranked defense. They are yielding opponents only 8.2 points per game on average and have faced the 26th toughest schedule so far.
You should also note that this is passing only and if your qb has good legs, look also at the rushing rankings.
You should also note that this doesn't take into account their opponent that week.
At this point, the top and bottom 10 passing and rushing defenses tends to stabilize quite a bit for the rest of the season with perhaps one team moving in and out of the top or bottom. Too much data now forces the trends a bit.
The jets have allowed only 4 passing TDs in 7 games and have had a solid 11 picks. The average for the league is 10 with the Colts giving up 15 in 8 games folllowed by the Rams, Broncos, Dolphins and Minnesota each giving up 14 TDs. Despite how good Cincy looks against the pass, they have only 3 interceptions (league averages 7 or one per game) with buffalo having the most at 14 picks. A surprise to me is that KC has the 2nd highest number of picks at 13 but one has to remember 6 given up by OAK in one game. Still, they've been good at keeping the ball away. And the primary reason that PIT isn't higher on my ranking scale is that they've only had 2 picks in 8 games which is tied for worst in the league with Miami.
Based on 1 per 10 yards, 6 per TD, -2 per fumble.
For example SF is my number 1 rushing defense, ranked 11th by NFL, giving up an average of only 5.3 ppg to opposing rushers but with the easiest schedule this year.
SF's streak continues. No rushing TDs allowed so far this year. The Ravens have similarly allowed only 2 rushing TD all year. Rough day for Mendenhall coming up. CLE, DET, GB and MIA have also only allowed 3 rushing TDs in their games this year with CLE being the surprise there. Normally one would think of them as a very weak defense. MIA is also a surprise in there given how soft they are in general.
CAR and ARI have yielded the most rushing TDs with 11 apiece and Indy has allowed 10. The jets and giants are solid passing defenses but they have each yielded 9 rushing TDs which averages at more than one per week. As for Fumble Recoveries, again the Steelers would be ranked in my top 10 but for the fact that they've only had 1 FR in their 8 games of play so far.Combined with their low picks, they end up very low from a turnover standpoint.
The sack leaders are NYG at 26 and BAL at 25 but a surprise right behind is MIN with 24. DET has the same mark. And WAS has 23 although on the decline and PHI with 22. I would have thought DAL to be in the top 5 but they and SF have 21. On the low end of that spectrum is KC with only 9 sacks on the year. TB, IND have 12 and the Bills, for as good as they are on defense, only have 13.
I've always said that Defensive and Return TDs are difficult to predict other than pressure such as sacks and turnovers.
However, just for completeness, the Ravens and Lions have had 4 defensive TDs. Buffalo, Cincy and the jets have each had 3 defensive TDs and CHI, KC, NO and TEN have had 2. But as I've always said, these are very hard to predict from week to week other than looking at sacks and turnovers.
Special Teams returns are even harder to predict. CHI, SF and ARI each have 2 returns for TD. A smattering of others have 1.
Points allowed are typically a poor predictor of overall score but for the record, the Ravens and 49ers have only allowed 104 and 105 points. Close behind are the Bengals at 117. The Jets are fairly close to the pack of 128 along with PIT at 133. But looking at the other end of the spectrum, Indy has allowed 234 points (gah!) and Carolina and Minnesota have allowed 201 and 197 points. At the extremes, I do think that points against will matter more.
So that's the tour of the statistical look at defenses for the week.
This information is for entertainment purposes only. Please do not replace common sense and good judgment with a bunch of statistics.