I didn’t see any of the games this week past Thursday night (ok, I saw half of the Rams-Jaguars game, but that’s nothing to brag about), so I don’t have too much to say. Instead we’ll just go through how the models did and update the rankings/total win predictions.
I’ll start with the Hilton SuperContest picks. This past week Luigi had the Jaguars, Jets, Texans, Eagles, and Cowboys, for a 3-2 record. I wasn’t really sure the Rams could win by double digits, but they did it. And the Texans are imploding. But the Jets and Eagles won outright as underdogs, and the Cowboys nearly won a game that we’ll conveniently forget about if the Broncos win a few more and start getting real ‘undefeated season’ talk. Since I usually follow Luigi, that would put my SuperContest ‘entry’ at 14-10-1 for 14.5 points. That’s currently out of the top 30 pay spots, but in striking distance and well above average (which I have at 11 points). The average this week was actually under 2.5 correct, so 3-2 is solid. Yoshi 1 also went 3-2 for a total of 12-12-1, and Yoshi 2 went 3-2 for a total of 13-11-1. So as far as it goes, all three models are doing better than the average SuperContest entry, but none have worked their way into the money yet.
How about the predictions as a whole? In all games, Luigi was 6-8, but would have passed on two games and gone 6-6 on the rest. That gives Luigi a two week total of 33-27-1/30-22-1. Yoshi 1 was 7-7/6-5 for a total of 31-29-1/26-22-1, and Yoshi 2 was 6-8/6-6 for a total of 30-30-1/26-23-1. The consensus picks, where all three models agree, are 20-14-1 through three weeks of picks.
Using his lines for the past weekend, Bill went 9-5. Luigi and Yoshi 2 went 6-8 and Yoshi 1 went 8-6. That makes the two week counts Bill 30-27-4, Luigi 32-25-4, Yoshi 1 33-24-4, and Yoshi 2 30-27-4. Bill made up good ground, but I still have a couple games on him.
Luigi was 5-8 on moneyline bets for a positive week, thanks mostly to the Jets upset on Monday night (the Jets were +365). Oakland at +180 was also helpful. Yoshi 1 had a pretty good week to just barely move into the black, and Yoshi 2 didn’t have quite as good a turnout this week but is a bit more into the black. Luigi is well ahead of both Yoshis though.
How about some rankings?
Unsurprisingly, the Broncos are still number 1. Their defense is not so hot, but their offense is shockingly higher than any other team’s. To draw a bit of a comparison, Luigi has Denver and Detroit’s defense as being about the same, but the Broncos’ offense is better than the Lions’ by the same extent as Chicago is better than Jacksonville. That puts Denver in first while Detroit is in the middle of the pack. At the sad end of the list, the Texans are making an appearance. It isn’t so much the pick-sixes, although they don’t help, but generally poor play and a 3-point output is going to catch up with you. And even with the big win over Jacksonville, the Rams are still 5th-worst.
We have five teams aimed solidly at 10 wins: Denver, Seattle, New Orleans, Kansas City, and Indy. But I wouldn’t be surprised if a couple ended up at 12 or 13 wins. This seems like a good place to discuss luck. When I look at the difference between actual wins and expected wins, I usually call it ‘luck’. That doesn’t mean luck in the usual sense; the Broncos have certainly played well enough to win 5 games. What it means is that with their stats and the stats that their opponents have put up, a team would be expected to win more like 3 or 4 games. The luck comes from how and when those stats were accumulated: fumbles and interceptions weren’t at critical points in the game or happened away from scoring position (for either team), good offense (or bad defense) came in circumstances where it was needed (or not hurtful), and so on. Similarly, the Broncos will probably be favored in almost every game they play the rest of the season. But that doesn’t mean you really peg them to go 16-0; you have to assume that something will break the wrong way at some point. After all, if you have a 80% chance of winning (which would be pretty high for the NFL), you have a 20% chance of losing. So when I say a team was lucky or unlucky, that means that they’ve had more of those 80%’s become wins and fewer of the 20%’s become losses.
It’s still early in the season, but we can start to look at some important games. The Chiefs are playing the Raiders, which shouldn’t necessarily be a close game but the Chiefs will need division wins to keep up with the Broncos in case a tie-breaker comes into play. The Saints-Patriots game is important for New England, who hasn’t quite looked as good as its 4-1 record suggests. A win or at least solid showing against the Saints will tell us something about how good the Pats really are. The 49ers-Cardinals game looks like it should be a pushover for San Fran, but both teams are 3-2. Whoever wins will have picked up a valuable division win in the chase against Seattle. Every gamed played by a NFC East team is important because they’re so close, so games against each other are especially important, making the Sunday night Washington-Dallas game a good one. And I think everyone will be interested in Denver-Jacksonville just to see how much the Broncos win by and what kind of stats Peyton Manning will put up before he gets pulled.