### FIWK the NFL: Week 10

I obviously picked the Chargers this week. For everything else I'm going totally quantitative. For a while I had been trying to figure out how to relate points differential or yards/play differential into a way to attempt to pick against the spread. Bill Barnwell and Bill James provided the answer this week. Taking the Pythagorean winning % for the two opponents and Bill James' formula, you can determine how often the favorite should win. Then I used Spreadapedia to look at how often a favorite at each spread has won outright and added 3% for the home team. For example, based on the Chargers and Raiders points scored and points allowed, the Chargers should win 60.8% of the time. But 7 point favorites win 72.2% of the time. If we expect the Chargers to win 60% of the time, they should probably be favored by 3 or 3.5. So, strictly quantitatively, I would pick the Chargers to win outright and the Raiders ATS, but I never pick against the Chargers. I just can't.

Unfortunately, all of this makes for fairly boring analysis of each individual game.

SAN DIEGO (-7) over Oakland: Part of me is going to pretend this game hasn't happened yet. But the majority of me wanted to fire Norv Turner at half time.

New Orleans (0) over ATLANTA: The Falcons win 39.9% of the time. I still think the Saints are the second best team in the NFC, but the Giants and the 49ers are making me think long and hard about it.

CINCINNATI (+3) over Pittsburgh: Apparently I'm picking the Bengals straight up? When you look at team a vs. TEAM B with their stats, team a wins 38.9% of the time. There may be one huge flaw with my method and this game exposes it. Can you figure out what it is? *I'll reveal it at the end.

CLEVELAND (-2.5) over St. Louis: Cleveland should be favored by 7... you know, according to my arbitrarily chosen quantitative method.

Buffalo (+5.5) over DALLAS: No one is going to be surprised to pick against the schizophrenic Cowboys.

Jacksonville (-3) over INDIANAPOLIS: Wait, is Curtis Painter the quarterback or are the Colts starting someone worse?

Denver (+3) over KANSAS CITY: Not so much Tebow as the Chiefs and Matt Cassel being frauds. (Remember, I'm just making up reasons to go along with my arbitrarily chosen quantitative method.)

Washington (+3.5) over MIAMI: Um, taking the points between two sucky teams? But my method does indicate the Dolphins should win 53% of the time.

PHILADELPHIA (Off(0)) over Arizona: There probably is some bigger spread somewhere...

Houston (-3) over TAMPA BAY: Houston's defense is actually looking pretty good. Just by my numbers, the Texans should win 80% of the time. This would be an 11 or 12 point spread.

Tennessee (+3) over CAROLINA: I don't even need my quantitative method to pick a 4-4 team plus points over a 2-6 team.

Baltimore (-6.5) over SEATTLE: I don't need any method to pick in favor of Baltimore and against the Seahawks.

Detroit (+2.5) over CHICAGO: I think the Lions are the better team. Many, many stats agree with me.

SAN FRANCISCO (-3.5) over Nyg: SF probably gets a 2 seed with a win. The Giants are the leaders for the 2 seed with a win, but the NFC East is too competitive and the NFC South champ would also be pretty close. The 49ers are lucky this specific game is at home.

NYJ (-1.5) over New England: Pretty much every way you look at it, this is a 50/50, with the Jets being at home giving them the slight edge, maybe 52-53%? A team favored by 1.5 should win about 52% of the time, so this spread is right on. This is a coin flip guess and I see no opportunity. I will not be putting $ down on this game.

GREEN BAY (-13.5) over Minnesota: When you get this far out (on the scale of spreads) the sample sizes get really small and the minute differences are more like noise than statistically significant. The Packers should win this game 79% on a neutral field and with home field advantage, maybe bump it up to 82%. Every spread from 11 to 14 has the favorite winning 80-82% of the time. This doesn't feel like a coin flip... more like rolling two dice. Under 7, Vikings win; over 7, Packers win; exactly 7 roll again. You are still getting resolving a 50/50 decision, just doing it in a more complicated way.

*So, the big flaw in my system...did you figure it out? The Steelers have played and lost to the Ravens twice. The Bengals have played neither the Ravens nor the Steelers yet and these will probably be the four hardest games on their schedule. My strategy does not account for quality of opponents. Just like how the Chiefs made the Bills and Lions look amazing by losing by 34 and 41 in the first two weeks.

Labels:
NFL

Subscribe to:
Post Comments (Atom)

You may say the analysis is boring, but I actually this was super fascinating. I love that you didn't mope about your picks history, and just jumped into a new attempt at analysis.

ReplyDeleteThere are some really good games this weekend. Too bad I'll be driving around for about 3 hours in the middle of the day... hopefully I can catch the night games.

By the way, on your footnote - your strategy does account for quality of opponent to the extent that Pythagorean Wins are considered a good indicator of team quality (better, at least, than straight up wins-losses).

Man this is a good week. Okay, here's how I'm picking:

ReplyDeleteChargers -7: wait, what? awww crap... nevermind

Jacksonville -3, $100: This seems like free money. The Colts are clearly going tankariffic all season.

Tampa Bay +3, $50: I don't trust the Texans one bit, and I think their defense is worse now than it was earlier this year due to injuries. And the Bucs are looking feisty again (finally).

Picks part 2:

ReplyDeleteBaltimore -6.5, $100: What the hell is going on with this line. Did I go into a coma for four weeks and the Seahawks are now good? Or the Ravens started sucking? They should MURDER the Seahawks; this line should be -13.5 like the GB line. I am delighted.

NY Giants +3.5, $50: I am excited to get an extra 0.5 on this game which might end in a Niners field goal, might end in a Giants win, but will still win me the bet.

New England +1.5, $100: Unlike how Aaron is picking, there is no logic to why I'm doing this. I just feel like I trust the Patriots 200% more than I would ever trust the Jets, and also Tom Brady could throw for 500 yards any given week.

$20 on every game except the NYJ-Ne and GB-Min.

ReplyDeleteI, too, am excited to see you work through so many different theories & strategies week-after-week. This type of work is really what analysts do all day long to try and find that edge.

ReplyDeleteHere are my picks:

$20 on Chargers -7: Coming off the solid game against the Packers, and the Raiders shat against the Broncos, this line is easy money.

$50 on Ravens -6.5: Royce, I agree.

$20 on Niners -3.5: I dislike giving up the half-point, but I think the Niners D makes Eli look foolish tomorrow afternoon.

$20 on Steelers -3: This line confuses me.

So, the Chargers suck. Down to $1,060.

ReplyDeleteI went 2-3 but won two $100 bets and lost two $50 bets and a $100 bet meaning... I came up even.

ReplyDeleteI really wish I had wagered $1,000 on the Jaguars. Hell, $1,000,000. That was the biggest slam dunk of the season.

By the way, the Ravens game is the biggest shocker to me BY FAR. What the HELL is going on with that team... was that the classic, textbook definition of a trap game? Are they just inconsistent and only gameplan for the Steelers? I am so confused. Where is Ja!

ReplyDeleteWHERE'S JA?!?!

ReplyDeleteWon the Niners/Steelers bet (+$40) and lost the Ravens bet (-$50) to bring me down to earth (aka $1,050).

I went 6-7, down from $810 to $790.

ReplyDeleteRoyce - where are you in regards to overall $$?

ReplyDeleteDammit Scooter, I was afraid someone would ask me that... it forced me to go back (several weeks) and look up my history.

ReplyDeleteThe last time I picked was week 7. In week 7 I started with $900, lost $100 in week 7, so I am at $800 after going even this week.

I need to get back to my EPL picks, those were my moneymakers earlier this year.

So basically, MP is handily beating 2 of us and basically even with Scott, using his tried-and-tested "do nothing" strategy. Damn he's good.

ReplyDeleteYou mean, he's almost winning with the patented France method?

ReplyDeleteFrance is a country that loves a proverb, and I think the one that applies here is, "One who does nothing loses nothing" *

ReplyDelete*Note: not an actual French proverb.