Ian Allan dissects each category for defenses and special teams, and when he totals it up at the end, Seattle projects to have the best defense in all scoring formats he can think of.

I will dip into the three remaining positions -- RB, WR, TE -- on Thursday.'

Here are the defenses ...

DEFENSES:

Sacks:

For sacks, New England is definitely last. The Patriots actually led these teams in sacks in the regular season, with 48, but Peyton Manning is really good at getting the ball out of his hands. He was sacked only 20 times in the regular season – half as often as any of these other teams. The Patriots got 2 sacks in the earlier game against the Broncos, but that’s probably their ceiling. We’re putting them down for 1.8 sacks, and that might be overly generous.

The other three groups all look pretty similar. The Broncos had 41 sacks in the regular season; the Patriots allowed 40. The Seahawks had 44 sacks, the 49ers allowed 39. And San Francisco had 38 sacks, while Seattle gave up 44. So the combined numbers for all three of those matchups are 81-83, or about 2.6 per game.

Denver doesn’t have Von Miller; he’s their best sack guy. But they managed to sack Philip Rivers 4 times on Sunday, and that’s impressive – Rivers was sacked only 30 times in the regular season. Tom Brady traditionally is tough to sack, but that really hasn’t been the case this year – he’s been sacked 40 times. The Patriots are playing a little bit better now than earlier in the year, but that sack number has held pretty steady – 22 in his last 10 games. We’re putting Denver down for 2.4 sacks.

With Russell Wilson, he’s really elusive, but defenses have been able to get to him. He was sacked 44 times in the regular season – 4 more than any of these other teams. Seattle was missing its offensive tackles for a big chunk of the season, but there’s more to it than that. He’s been sacked 4, 4, 4 and 3 times in his last four games. He was running for his life the last time San Francisco played here, taking 4 sacks (he took 2 in the rematch down at Candlestick). I’m going with 2.7 sacks.

The Seahawks, I think, belong in that same neighbor. Of these four teams, they ranked 2nd in sacks in the regular season. Kaepernick took 39, including 5 in the two games against Seattle. We’ll go with 2.6 for them.

Interceptions:
These offenses are all good at avoiding interceptions – 8-11 picks for each of them. While these defenses have all shown some ability to take it away. They all averaged over an interception per game. Only 10 of the league’s other 28 times finished with more than 16 interceptions.

Seattle leads the way. With 28 interceptions in the regular season, it had 10 more than any of these other clubs. Plus it’s playing at home, where it will be helped along by crowd noise. Kaepernick threw 3 interceptions in his last game up here, and he had 1 interception in each of his other two starts against this defense. If any quarterback is going to throw more than 1 interception, he’s probably the guy. That said, it’s far more likely that he’ll throw 1 rather than 2. He’s thrown only 5 interceptions in his 16 games against other teams this year.

Russell Wilson is very careful with the ball. He’s thrown only 9 picks all year – fewer than Manning or Brady. It’s as if Alex Smith is whispering in his ear, reminding him that it’s better to take a sack or throw the ball away, rather than turning it over. But San Francisco seems to match up well against him; they’ve intercepted him once in each of his four games against them.

We’re putting New England a little ahead of Denver in this category. That’s out of respect for the potential that Belichick might bait him into some kind of mistake with a Jedi defense. Manning doesn’t throw interceptions, but he’s been picked off 6 times in his last four by the Patriots. Andrew Luck doesn’t throw interceptions either, but he served up 4 to New England last week.

Denver comes in last, but no big difference between it and the 49ers and Patriots. The Broncos intercepted 17 passes in the regular season, while Brady threw 11. So that combined number (28) is actually a little higher than either of those other two defenses.

Fumbles:
I don’t worry too much about fumbles. I think it’s mainly a luck-driven category. In the regular season, half of the teams in the league recovered either 9 or 10 fumbles. The Carolina Panthers, who were awesome defensively, finished with only 6. Anyway, when you get up to 11 fumble recoveries, you’re starting to beat the odds and be something special. So on my board, I tend to rank defenses with .60 fumbles in a game. If they’re facing an error-prone quarterback like Geno Smith, we can think about maybe moving them up as high as .70. But the vast majority of the time, the rankings should be .50-.70. Not a lot of difference between these units.

It may surprise you, but Denver lost 16 fumbles in the regular season – 6 more than any of the remaining teams. That’s the one area of his game where Peyton Manning has declined. He had 10 fumbles (6 lost) in the regular season. In none of his 10 seasons did he have more than 6 fumbles or 2 lost fumbles. So while he doesn’t get hit too often, if you get to him, it seems that ball can slip out a little more than it used to.

San Francisco was the best team at not fumbling in the regular season (just 20). But I think they’re right there with the other teams (who all finished between 24-27 regular-season fumbles). The Seahawks, 49ers and Patriots all lost 9-10 fumbles in the regular season.

You can talk all you want about this guy being a hard hitter or that guy being a playmaker, but ultimately it doesn’t really play out much in the results. Denver recovered 9 fumbles in the regular season, while the other three teams were all at 11-12. I’m just putting all four teams down with a 60 percent probability of recovering a fumble.

Touchdowns (defensive):
It doesn’t look particularly likely that any of these teams will return an interception or fumble for a touchdown. The Seahawks, Patriots and Broncos all had only 2 turnovers returned for touchdowns against them, while the 49ers didn’t allow any.

The Seahawks managed to score 4 of those touchdowns (an average of one every four games), but they’re going up against the team that didn’t allow any.

The Patriots and 49ers each scored only 2 TDs that way, and the Broncos scored only one. Combine all that garbage up, and you’re looking at only about a 12 percent probability for each of these defenses to score. So there’s about a 50 percent chance that one (and only one) of these defenses will get a touchdown.

Seattle has the opportunistic secondary and is playing at home. We’ll rank them a little higher than the others. Denver is down a pass rusher and just lost a starting corner, so we’ll stick them in the 4 spot.

Out of respect for New England’s coverages and history against Peyton Manning (plus his increased fumbles this year), I’m putting the Patriots 2nd and the 49ers third.

Touchdowns (special teams):
I think the order needs to be Broncos-Seahawks-Patriots-49ers. The Broncos have Trindon Holliday, and he’s capable of the big play. He scored 2 TDs on returns against Baltimore in the playoffs last year, and he had one called back on Sunday against San Diego. He scored 2 TDs on kick returns in the regular season. Holliday has been a little shaky with fielding them, so I don’t think they’ll use him on punts. Eric Decker doesn’t have the same return skills, but he’s safer with the hands. The Broncos lost the earlier New England game, recall, on a bungled punt in overtime (Wes Welker didn’t fumble it, but he allowed it to land and bounce off one of their blockers). I don’t think of Decker as being much of a return guy, but he had one over 40 yards against the Chargers.

Seattle is close behind Denver, I think. Golden Tate has been a threat on punt returns all year. He hasn’t taken one back for a touchdown, but he’s been close a few times. I believe I heard that he’s had six or eight returns over 20 yards. I know he had one 71-yarder. To me, he’s definitely the best punt returner playing this weekend, and there are about twice as many touchdowns on punt returns as kickoff returns. Depending on how he grades out physically (coming off a concussion), Percy Harvin might be used on kickoffs, and if so, he’s almost as good as Holliday.

New England comes in 3rd on my board. Julian Edelman is a skilled, veteran punt return. He averages 12.3 yards on punt returns for his career, which might be best in the league for guys with at least 50 returns. He scored 1 TD on a punt return ine ach of three years leading into 2013 – 3 TDs in 37 games in those years, which works out to about an 8 percent probability. But Edelman in all of those years was a reserve receiver. Now he’s their starter, so he’s not going out there fresh. I don’t think he’s going to return one for a touchdown. With kickoffs, their using LeGarrette Blount. He’s kind of the kickoff version of Eric Decker. I don’t think too much of Blount as a returner – too slow – but must concede that he had an 83-yarder for a near-score in Week 17. With weather being good on Sunday, I doubt he’ll even return one in this game. Defensively, the Broncos are the only remaining team that allowed a kick return touchdown in the regular season.

I’m sticking San Francisco last. Not hapless, but definitely below-average by NFL standards. In the first game of the Jim Harbaugh era (at Candlestick against Seattle) Ted Ginn returned a kickoff and a punt for a touchdown. But those are the only touchdowns they’ve scored – they’re on a 54-game dry spell. James had a 62-yard kickoff return that was pivotal in a win at New England last year, and he had a 40-yard punt return this year, but he’s been pretty ordinary. Seattle in the regular season almost set a record for fewest punt return yards allowed (they tend to put punts high, so few are even returned).

When an average NFL team is playing, we give it a 6 percent probability of scoring on a kick return – 4 percent for punts, and 2 percent for kickoffs. On this Sunday, I will go with: Denver (8 percent), Seattle (7 percent), New England (5 percent), San Francisco (4 percent).

Points allowed:
The gambling lines suggest Denver will win about 30-25 and Seattle will win about 21-18. No need to stray too far away from that. I’ll just drop those numbers into our database.

Recommendations:

Putting all those projections in a spreadsheet, we get the following.

For a standard league (6 points for TDs, 2 for takeaways and 1 for each sack):

For the same kind of league but also giving bonuses for fewest points allowed (specifically, a point for every 2 points allowed under 24 – shutout = 12 fantasy points, 10 points = 6 fantasy points, etc.)

And for a TD-only league …