Thursday, November 29, 2007

Week 13 - Capper and Box lovefest - Preliminary picks

Well well, Excelius, we have another week where we actually agree! Last time it worked well for me, and you got hosed. Let's hope for more of the same. I mean different. Possibly same...

Because my actual job took up too much time this week, I wasn't able to give you the Vegas vs Capper table. However, these are the games that I love this week, and unbelievably the box agrees, mostly. Must be my birthday.

1. CLE +1 v. ARI - One of atleast 3 games where Vegas favored the opposite team I did? Can this be true? and it's the box's #1 pick! This definitely must be my birthday. I said to bet the house if CLE is favored by less than 3 points... so i suppose it's good I'm a renter. This would be my double-down pick, if i was so inclined, though.

2. NYG (-1.5) at CHI - Love seeing Chicago get credit for their special teams' insanity last week. No way they get 3 touchdowns from special teams each week. This game might just be the case study for my Wave Theory. I had capped this game at NYG (-6), so I feel like I'm playing with 4.5 free points.

3. DET +3 v. MIN - I hate betting against the Minnesota Spreadbusters, but this is almost too good to be true. MIN gets a huge upset against the G-Men, but Manning gift-wrapped 21 points for them. MIN still has no offense, an ineffective QB and no Peterson. My own laws state not to bet on DET (multiple personality reasons) and against MIN (spreadbusters)... do I have the discipline? I don't think i'll know til Saturday.

4. JAC +7 v. IND - Indy is still banged up, and JAX is rolling. This should be a close game, so the 7 points seems excessive. The box likes it at #5, so that's even more encouragement

5. TB +3 v. NO Away - The box shocked me with this one. The box loooves NO. Every week I can't believe you bet NO. Over and over again. Saints Saints Saints.... Well, welcome to the dark side, my friend. I actually have TB favored in this game (-2), so this is another potential moneyline pick. I love erasing the juice.

6. HOU +3.5 v. TEN - As Excelius has pointed out, TN has been in a slide and it may be that it should have been expected. HOU is still a scary team to me with AJ and MS, so i liked this to be a game they could pull out. The spread might be too low for me to take a shot, though.

7. OAK +3 v. DEN - This is one we disagree on. OAK seems ranked suspiciously high in your financials (we need a better name for this list), but maybe they are better than common knowledge suggests? Or maybe Den is an wave theory pick... They are good enough to beat PIT , KC and TN. But recently they've barely lost big to DET, then barely to GB (overtime), and barely to CHI (despite giving up 3 tds on special teams). I dunno, i'll probably stay away from this one, since we agree on so many others.

8. NYJ +1 v. MIA - I just figured NYJ are a little better, and pretty weird that they're the 'dog. As you can see my conviction basically matches the box (#16), so we'll see if i need an extra pick to make my parlays work.

9. BUF +5.5 v. WAS - You have to figure WAS is distracted this week, plus they now have to start their backup safety. Unfortunately I heard they benched JP again, and haven't heard if Lynch is healed. You know the more I write this, the more I think I'm staying away from this game. Too many variables.

Week 13 - Box's (Semi)Final Bets

The (semi)final bets are in. I haven't decided if I'm adding in a few moneyline bets this week, so I'll wait until Friday or Saturday to make those bets. Right now, I decided to stay conservative and bet my top 5 spread picks only. I will NOT bet a parlay or teaser this week... well, maybe a teaser...

Here are the bets so far:

Bets for Week 13
Spread: GB +7 (-110) v. DAL - $8.99 to win $8.17
Spread: JAC +6.5 (-110) v. IND - $11.83 to win $10.75
Spread: CLE +1 (-115) v. ARI - $10.97 to win $9.54
Spread: NYG -1.5 (-115) v. CHI - $8.71 to win $7.57
Spread: BAL +20.5 (-110) v. NE - $9.50 to win $8.64

Results for Week 13
Bet Totals: $50.00
Potential Winnings: $44.67

Week 13 - Best Bets and Your Mom

Your Mom called - she said to throw your money away today. You'll see why with this week's top 5:

The Rankings - Spread

* considers lay in pick order Under
1. JAC +6.5 v. IND Away 44.0
2. CLE +1 v. ARI Away 42.0
3. BAL +20.5 v. NE @Home 44.0
4. GB +7 v. DAL Away - Bet Soon 46.5
5. NYG -1.5 v. CHI Away 42.0

This is only the second week of the year that the Box has decided winning is a terrible thing. Baltimore?? NOOOOOOO!!!

After a close game last week, Tom Brady will be out for blood. And I heard he hates creepy black birds, too. So when Baltimore comes to town next Monday with a shaky Kyle Boller, no offensive firepower, and a mediocre defense; could we be back to 30 point drubbings? Well, I'll find out firsthand when I lose money on the game... sigh...

I like the Jacksonville pick, but I liked it better early this week when it was +7 and pushable. I like the GB pick, it's pushable and GB is on a roll. But my favorite pick is the NYG pick. The G-Men will be itching to take out that last loss on someone. And I have a feeling Tom Coughlin will shove Brandon Jacobs right up the Bears a-hole, keeping in a galaxy far far away from "The Other Manning". It must suck to be the red headed stepchild of the Manning family.

With the Wednesday moneylines out, here are the Box's pick ranks:

The Rankings - Moneyline

* considers lay in pick order
1. SD -260 v. KC @Home x2!
2. NO -180 v. TB Away
3. PHI -190 v. SEA Away
4. CAR -150 v. SF Away
5. TEN -210 v. HOU Away
6. WAS -250 v. BUF Away
7. DEN -200 v. OAK @Home
8. MIN -190 v. DET Away
9. IND -300 v. JAC Away
10. DAL -280 v. GB Away
11. PIT -300 v. CIN Away

Not a lot of ridiculous lays this week, which I like. 5 off games makes the picking easier, since they've already narrowed the focus. I feel like the Jets/Dolphins game being OFF and Miami being -1 indicates a trap game. The same is true of the Cleveland/Arizona game. That's a gut feeling, but if the spread is basically a pick 'em, why make the moneyline off? Of course, the CLE game is one of the top 5 this week, so I'm prepared to throw money away there, too.

I love the SD pick, though. Herm Edwards is imploding, and the team is playing accordingly. In this year's rematch of the Super Terrible Coaching Bowl, SD will be out for blood I think. The lay is low, the odds are good, and the box says to double it.

Last tidbit for the afternoon, a look at my first draft of a teaser module. I will not be betting this, but any feedback on methodology could be useful here. Note that "Two Score" means a 6 point teaser makes it at least a two score game, and "Pushable" means the spread becomes pushable. This may be a bad teaser week, but it highlights a preliminary top group as picked by the Box:

The Teaser - Playlist
Games: 4
Pick Spread Status Rank
GB 13.0 Two Score 5
SEA 9.0

BUF 11.5 Two Score 3
DET 9.5

SF 9.0

ATL 9.5

JAC 12.5 Two Score 7
NYJ 7.0 Pushable 4
HOU 9.5

KC 11.5 Two Score 8
OAK 9.5

CLE 7.0 Pushable 6
TB 9.0

NYG 4.5

CIN 13.0 Two Score 1
BAL 26.5 Two Score 2

I went with a 4 game teaser for example purposes, but pick ranks are on the right. I hate this module right now, ideas are welcome. Time to place my bets!

Wednesday, November 28, 2007

Week 13 - Wall Street Ranks

So the power ranks are in, and I added a new feature - the overall rank. It's a combination of Sharpe and Alpha total convictions (Beta is a component of Alpha, and Sharpe has its own risk adjustment, so adding Beta into the mix would be redundant). I have also further refined the Sharpe rank to account only for points scored ABOVE the spread. By doing this, I'm actually ranking a team's consistent ability to bust a spread, and hopefully making it far more useful for parlays in the future. The alpha and beta are very much the same, but I've slightly tweaked how they are scaled for ranking purposes, so there may be a minor rank change from last week's post. This has had little effect on the rankings, but has improved the double bet performance.

Here are the financials - if you want to see any specific game breakdowns this week, let me know, and I can post it:

Sharpe Alpha Beta
Team Rank Rank Rank
ARI 15 13 18
ATL 28 32 31
BAL 30 19 9
BUF 23 30 19
CAR 24 12 5
CHI 19 28 32
CIN 18 5 7
CLE 12 7 26
DAL 3 1 1
DEN 25 25 16
DET 14 6 4
GB 4 11 21
HOU 17 10 15
IND 2 3 3
JAC 6 27 27
KC 26 31 30
MIA 32 17 24
MIN 11 16 10
NE 1 2 12
NO 16 24 28
NYG 13 4 2
NYJ 31 14 20
OAK 21 9 11
PHI 10 22 25
PIT 5 8 6
SD 9 18 29
SEA 7 15 13
SF 29 23 8
STL 27 29 14
TB 8 26 23
TEN 22 21 22
WAS 20 20 17

A lot of the same from last week, but I did note a trend that most of the press has given a lot of attention. Tennessee's fall from grace has been nothing if not meteoric, and it has largely been attributed to the absence of Albert Haynesworth. But is it really just one man that makes the difference?

If I look back in the rankings, we see the following:

Sharpe Alpha Beta
Week 8 TEN 4 20 22
Week 9 TEN 5 20 21
Week 10 TEN 5 20 22
Week 11 TEN 10 20 22
Week 12 TEN 16 21 22
Week 13 TEN 22 21 22

The most obvious drop is in the Sharpe, and boy does it drop. Nearly 20 spots in 3 weeks. But the alpha and beta exactly the same since week 8. So what does that tell us?

Well, since Sharpe measures point output over the spread adjusted for risk, it's clear they can't score, not that they can't play defense. The loss of Haynesworth has been a problem, but what it's highlighted is the poor play of Vince Young and the Titan offense more than anything else. Teams are stacking 8 in the box and making Young beat them with his arm, something he has yet to prove he can really do, especially with this young receiver core. The real Titans show up in the alpha and beta ranks, which take into account statistics that correct for league averages as well as point output. They have consistently ranked in the bottom half of the league. So, will the real Titans please stand up? The Box will be hard pressed to take them as favorites for the rest of the year, with or without Haynesworth.

Here are the overall financial rankings, the new segment. This should closely compare with most of the power rankings out there, with a few exceptions.

Overall Rank
1 NE
5 GB ▲1
6 NYG ▼1
10 SEA ▲1
11 OAK ▲1
12 HOU ▼2
13 SD ▲2
14 MIN ▲3
15 ARI ▼2
16 PHI
17 CAR ▼3
18 TB
19 JAC ▲2
20 NO ▲2
21 WAS ▼1
22 TEN ▼3
23 CHI ▲3
24 DEN ▲1
25 NYJ ▼2
26 BAL ▼2
27 SF ▲3
28 BUF ▼1
29 MIA ▼1
30 STL ▼1
31 KC
32 ATL

The arrows indicate a change from last week's overall rank. Big movers are Tennessee, SF, Chicago, Carolina, and Minnesota. Upward motion from Chicago, SF, and Minnesota reflect the great play and wins in Week 12, while the drop in Tennessee (see above) and Carolina reflect losses. The triumphant return of Steve Smith couldn't even outshadow the QB troubles for the Panthers.

Biggest surprise: Jacksonville at 19 despite it's record. Jacksonville has a long history of playing up or down to competition. Their consistency, especially against their statistical expectations, is terrible. It may be brought down mostly by the residual effects of the 41-24 drubbing New Orleans gave them a few weeks ago, but they currently sit at number 19. As I get deeper into the stats, maybe I'll get a specific explanation for that posted here this week.

Tuesday, November 27, 2007

Week 13 - Madder Capping

Gotta get even this week. So here's how I see the games going down:

GB at DAL - The NFC's version of Colts/Pats. Very even matched, but I'll give Dallas the edge in their house. GB -4.

BUF at WAS- Terrible news about Sean Taylor. How will the team respond? BUF had been playing good ball until they ran into the Pats and Jaguars, now they've benched Losman again and Marshawn has been hurt for 3 weeks now. I'll take an emotional WAS-2.

DET at MIN- DET's playing for their playoff lives, and gets a break in their brutal 2nd half schedule in a trip to Minnesota. Hey! Minnesota has astro-turf too! Peterson is still out, and no way MIN duplicates it's lucky defensive touchdowns. Let me say one last thing... I said it last week, I'll saying again. I hate betting against MIN, they are spread busters. DET -4.

HOU at TEN- Man oh man I want to favor HOU. TEN seems to be in a free-fall, but this is still supposed to be that overachieving team that toughed their way to the playoffs.... TEN is at home settles it. TN -1.

JAX at IND- JAX seems to be rolling, and they know how to play the colts. Rumor has it Joseph Addai could be back, though. Plus this is still Indy's division until someone takes it from them. IND -4.

NYJ at MIA- Two lowly AFC East times bump into each other this week. MIA is down to it's 3rd string QB, 3rd string RB and traded its #1 WR. There is no reason to pick them. Ever. NYJ -2.

SD at KC- KC was my madcapping darling for a few weeks there... until they kicked me in the balls 3 weeks in a row. Now they are my enemy. I hate cheering for LaDanian Whiningson but, what can you do? SD-6.

SEA at PHI- Philly showed some serious moxy against NE last week. SEA is definitely a Multiple Personality team. I'll give it to the home team again. PHI -3.

SF at CAR- Stinker of the week! I have no reason to pick either team. None. So let's go with SF on a 2-win streak, shall we? SF-1.

CLE at ARI- I really thought the wave theory would have caught up with the brownies by now, but they continue to be under-capped. If they are favored by less than 3 here, bet the house on it. CLE -4.

DEN at OAK- DEN would be riding a wave right now if Shannahan hadn't stubbornly kicked to Frickin' Hester last week. Fortunately OAK has nobody of that ilk on its team, so I like DEN in a landslide. DEN-5.

TB at NO- TB is the 3rd best team in the NFC right now. How 'bout that? NO is a spoiler team that will beat spreads when you least expect it. Not this time, though. TB-2.

NYG at CHI- Last week was a mulligan. For both teams. No way CHI gets the free points it got last week, and no way NYG gives away the free points again. NYG -6.

CIN at PIT- You know... PIT is starting to make me nervous. They lose to the Jets, and squeak past MIA. And bullcrap to you who says the field was messy. Isn't that supposed to be an advantage to the running-game steelers? In any case, I can't in good conscience bet the Bengals at 3-river. PIT-6.

NE at BAL- So will the spreads calm down a little bit this week? They probably should, but I am thinking this game might be a shutout 19-0. NE-19.

Week 13 - Wall Street Says Bet the Browns

So the early lines are out (the STL line is currently OFF almost everywhere, so I'll have to update this again on Thursday). This week is a little startling, though. Not because the Box suggests all but one dog - that's pretty normal. More because of the teams that ARE dogs this week. MIA favored by 1? Are the Jets that bad? DET a 3 point dog to MIN? After a 4 interception week, much of which can be attributed to luck and poor play by Manning, MIN warrants 3 over DET?

Here are this week's bets ranked:

The Rankings - Spread

* considers lay in pick order Under
1. CLE +1 v. ARI Away 42.0
2. SEA +3 v. PHI Away - Bet Soon 42.5
3. NYG -1.5 v. CHI Away 42.0
4. ATL +4.5 v. STL Away 36.5
5. JAC +7 v. IND Away - Bet Soon 44.0
6. GB +6.5 v. DAL Away 46.5
7. SF +3 v. CAR Away - Bet Soon 34.0
8. CIN +7 v. PIT Away - Bet Soon 42.0
9. BUF +5.5 v. WAS Away 42.0
10. TB +3 v. NO Away - Bet Soon 43.5
11. DET +3 v. MIN Away - Bet Soon 40.5
12. OAK +3 v. DEN @Home - Bet Soon 38.5
13. HOU +3.5 v. TEN Away 38.0
14. KC +4.5 v. SD @Home 42.5
15. BAL +20 v. NE @Home 44.0
16. NYJ +1 v. MIA Away 39.0

I will get to commentary in a later post, but those are the prelims.

Monday, November 26, 2007

Week 12 - Box's Results and the Balls Therein

Ok. I'm tired of not winning. It's not so much that I'm losing. I'm just not winning.

I discovered a few blips in the system over the weekend when I was doing a small audit of the math - the Sharpe was using an improper array, and the Moneyline was reading from the wrong module. Neither of these things proved earth shattering for the week or the historical results, it was a matter of a few percentage points here or there.

What I have discovered, though, is that parlays are terrible. Parlays are like leverage in a hedge fund - they have the ability to enhance the returns by putting very little down, but the "very little down" can add up if the returns aren't there. The payouts are large for a reason: they are near impossible to predict. I've tried several permutations, and had little to no success on a consistent basis. And yet, when I do the historical review, had I chosen my top 5 spread picks ONLY, I'd be up huge this year.

Lesson learned?

Probably not.

Despite my discipline in terms of analyzing data, I can't seem to separate it into placing the bets. The parlays are fun when they hit. I'm going to spend some serious time this week exploring the parlay potential of each module and the combinations. Maybe if I combine Sharpe, stats, and rank by beta it'll work? Maybe the alpha and Sharpe is all I need? Also, I need to develop a teaser module - basically, I need to mirror the entire system and add the teaser points to each team to see what the best bets are. I don't know if they're any different, but we'll see.

Since I have no money on the game tonight, here are this week's results:

Results for Week 12
Moneyline: SEA (-160) v. STL - $19.05 wins $30.96
Moneyline: CLE (-170) v. HOU - $5.00 wins $7.94
Spread: SF +10 (-110) v. ARI - $11.14 wins $21.27
Spread: MIN +7.5 (-120) v. NYG - $7.07 wins $12.96
Spread: STL +3 (-115) v. SEA - $11.48 wins $0.00
Spread: BUF +7.5 (-110) v. JAC - $6.25 wins $0.00
Spread: DET +3 (110) v. GB - $5.37 wins $0.00
Parlay: STL +3 (-115) v. SEA, CAR +3 (-120) v. NO, DET +3 (110) v. GB - $5.00 wins $0.00
Parlay: STL +9 (-110) v. SEA, WAS +9 (105) v. TB, BAL +15 (-110) v. SD, DET +9 (105) v. GB, PHI +28.5 (-110) v. NE, ATL +18.5 (-110) v. IND - $5.00 wins $0.00

Results for Week 12
Bet Totals: $75.36
Potential Winnings: $113.43
Payouts: $73.13
Actual Profit: -$2.23
%s: Week 12 -2.96%

Box Results - through Week 12
Initial $s: $139.16
Bet Totals: $513.91
Potential Winnings: $942.99
Payouts: $518.46
Actual Profit: $4.55
Gain/Loss 3.27%

Notice the teaser at the bottom of the bets that wasn't there in earlier posts. Yeah, that's called "5 minutes to game time, let's get crazy!" betting. I developed a slapdash teaser methodology that I didn't backtest that I decided to just go ahead an bet. Without that bet (or the parlay), I would have had a nicer week. Unfortunately, I am a moron.

Week 12 - Sunday Blues - Weekly Roundup

Well it started off so well with Thursday football. My last second changes to include indy and GB into the mix hit on Thankgiving, giving me a huge leg up on Sunday. Then it all went to crap.

First, the good:
IND vs ATL - I correctly added indy into the mix wednesday night, figuring they were a deep enough team to beat up lowly ATL. They are, and they did. I look forward to betting with IND and against ATL the rest of the way.

GB vs DET- It all came true.... DET was overrated and not given enough points, and GB went in there and spanked 'em. I added GB to my parlay and was looking forward to cashing in on Sunday.... Look for GB to keep winning, and keep an eye on any game they are not laying more than a field goal.

JAX vs BUF- I got this one right, too. JAX is the real deal, built for cold-weather power football. BUF is scrappy, but not ready to handle a team like this. BUF is still a good pick against the middle teams, though.

The Bad:
TN vs CIN- wow. 35-6. it wasn't even a game. My comment last week was that TN should just run over the Cincy defense. They didn't. The TN defense either misses Albert Haynesworth that much, or CIN just remembered how to play. Either way this pick blew up my parlay and half of my single-game picks.

KC vs OAK- I wisely moved them out of my parlay, but then stuck with them in two teasers, figuring that with a +0.5 at home against 2-win Oakland, it was a lock. It wasn't. I'm not sure what happened in this game, but it coast me two teasers.

ARI vs SF- The cards got beat by SF. Pathetic. Pathetic. Pathetic.

DEN vs CHI- Devin frickin Hester. Good for my fantasy team, bad for my single-game bet.


net losses = $25.00

Thursday, November 22, 2007

Week 12 - Matty is Mad, Capper is Cool - Final Picks

Matty Matty Matty.... What the hell are you doing?! (yes, i use multiple punctuation marks).

I completely agree that you are nuts for hedging the STL vs DET game. How can you be banking on a 1 or 2 point victory... let alone that you are relying on inconsistent SEA. You should hedge your algorithm by calculating the likelihood of a 1-2 point victory. Historically I bet it's less than 10%. Nuts I tell you. NUTS!

Another criticism... you're gonzo with your CLE vs HOU multi-bet. You pick the o/u on the game... fine. But picking CLE to cover -3.5 and then picking them to win the moneyline in the same parlay just doesn't make sense. If they win the game by less than 4 points, you will win your moneyline, but lose your spread pick anyway - destroying the parlay. now this would be fine if the moneyline was a higher risk-higher reward bet, but it's not! CLE -3.5 will pay 1:1 (less the juice), while the premium on the moneyline pick (since CLE is favored ) is ~70%! Why not just stick with the spread pick? If CLE wins but doesn't cover, you're parlay loses.

This parlay could hold some water if you were betting on an underdog. Say you thought HOU would win... Then you bet HOU +3.5 and pay the standard 15% juice, and you parlay this with a HOU +150 moneyline... Now you pay the same juice, but your moneyline pick ups your payout. This week the ideal game to play this game with would be DEN +2. We both think they beat CHI, and you'd be willing to bet them with the 2 points, and get the bonus payout for moneylining the 'dog.

As for the rest of your bets... this is more of the typical week where we completely disagree. I won't touch most of the games you bet, and the few games we both bet on, I am on the other side. This season you have kicked my ass in these situations, but I'm sticking to my guns.

I've made a last minute (albeit 5 glasses of wine) switch. I heard today that Priest Holmes retired. This means that KC is starting inexperienced Brady Croyle and now is down to their 3rd string back. While I still love KC at home vs OAK, I am no longer confident in their ability to cover a -5 spread. Therefore... I am adding a 2-team teaser to take advantage of an expected KC victory (KC +1) and adding in IND to cover a -6 spread vs. ATL... ATL has no quarterback, and no defense. They are f**ked. Will Indy win? YES! Will they win by a touchdown... I think so.

The second half of this new outlook is that KC needs to be dropped from my 3-team parlay. I'm sticking to my initial opinion that DET is overrated and that GB destroys them. Insert GB-3! Lastly, since the spread is only 2 in the DEN vs CHI game, and I like the 'dog (DEN), I'm switching from a spread pick to a moneyline bet to take advantage of the bonus payout, and avoid paying juice. If DEN loses by 1 point, I guess I deserve it. But again, what are the odds of that?

6-point Teaser:
[IND-6 at ATL -- KC +0.5 vs OAK]
[GB+3 at DET -- JAX -1.5 vs BUF -- ARI -4 vs SF -- KC +0.5 vs OAK]

3-team Parlay: [DEN +2 vs CHI -- GB -3 vs DET -- TN -1 vs CIN]

Straight up:
[DEN +110]
[TEN -1]

Friends shouldn't let friends bet drunk, but I'm here by myself with a bottle of Spanish Tempranillo Red Wine.


Week 12 - Box's Final Bets

Let me just say that I don't like my bets this week. In fact, I may prefer to have my face rub up against a cheese grater than make these bets. Too bad the Box says go. I am just a drone...

Bets for Week 12
Moneyline: SEA (-160) v. STL - $19.05 to win $11.91
Moneyline: CLE (-170) v. HOU - $5.00 to win $2.94
Spread: SF +10 (-110) v. ARI - $11.14 to win $10.13
Spread: MIN +7.5 (-120) v. NYG - $7.07 to win $5.89
Spread: STL +3 (-115) v. SEA - $11.48 to win $9.98
Spread: BUF +7.5 (-110) v. JAC - $6.25 to win $5.68
Spread: DET +3 (110) v. GB - $5.37 to win $5.91
Parlay: STL +3 (-115) v. SEA, CAR +3 (-120) v. NO, DET +3 (110) v. GB - $5.00 to win $30.99

Bet Totals: $70.36
Potential Winnings: $83.43

The stupidest part is the $20 SEA to win bet, coupled with 2 STL +3 bets. I'm throwing money away somewhere, especially when the Box is so convicted both ways. The Box is like a lawyer, arguing every side of everything.

Let's try to figure this out, shall we?

STL is the 3rd ranked spread pick, and is a statistical pick that wasn't overridden by they new Less Than Three module due to the 3 point spread. It's also a double down bet as the overall lowest game beta this week. So far, I like those odds. Seattle is overrated, and St. Louis is getting healthier. Not to mention they're playing on the turf. All good arguments for a St. Louis pick.

But then we move on to the moneyline. Moneyline is a combination of statistical conviction, alpha, Sharpe, and beta all rolled into one. The Box picks STL because statistically it's pegged the game at a 2 point Seattle win, not 3 as the spread suggests. So the moneyline statistically favors Seattle as well. But here's where it gets crazy. Seattle's Sharpe and alpha are OFF THE CHARTS compared to STL (see below for chart).

Pick: STL Opp: SEA
Pick: 16.75 Diff: -7.00
Opp: 9.75 Adv: SEA

STL has a minor beta advantage, but both of their betas are so high, it shouldn't matter - they should play as consistently as they have all season hypothetically. So not only am I caught betting on Seattle to win with a big bet, but the Box throws STL against the spread in its parlay suggestions and straight wagers. I now require the perfect storm: a 1 or 2 point Seattle victory. Small margin for error I would say.

I think I need to figure a way to exclude any moneyline bet if I've already got a spread bet going and the spread is 3 or less. This way, I won't be praying Seattle gets that safety to win the game, and I can rest easy thinking it'll just be a field goal. Sigh. Another losing week on the horizon? Am I outsmarting myself? Well, I'm a genius, so if anyone could outsmart me, it would be ME.

Cross your fingers, I need a big week. Lots of money down, let's get there.

And I haven't back tested this at all, but worth a giggle - the Box now suggests a single game parlay bet, and here it is:

O / U
I haven't backtested this module, but it was fun to make. I like the idea of having $5 or $10 on a single game with 3 factors going at once. The O/U represents what the over/under should be. So if Vegas has it at 45, you bet the under. If it's 40, you bet the over. Since I haven't tracked o/u historically, getting and back data will be very hard if not impossible, but at least it's entertaining.

Wednesday, November 21, 2007

Week 12 - Final Pick Ranks

Despite travel fatigue, I am true to the cause - here are the final pre-bet rankings for both spread and moneylines this week:

The Rankings - Spread

* considers lay in pick order Under
1. MIN +7.5 v. NYG Away 42.5
2. BUF +7.5 v. JAC Away 42.5
3. STL +3 v. SEA @Home x2! - Bet Soon 38.5
4. SF +10 v. ARI Away x2! - Bet Soon 35.5
5. DET +3 v. GB @Home - Bet Soon 42.5
6. WAS +3 v. TB Away - Bet Soon 43.0
7. CIN +1 v. TEN @Home 40.0
8. CAR +3 v. NO @Home - Bet Soon 38.5
9. DEN +2 v. CHI Away 40.5
10. BAL +9 v. SD Away - Bet Soon 40.5
11. NYJ +14 v. DAL Away - Bet Soon 44.0
12. MIA +16 v. PIT Away - Bet Soon 40.5
13. OAK +5 v. KC Away 38.0
14. CLE -3.5 v. HOU @Home 40.5
15. ATL +12 v. IND @Home - Bet Soon 42.0
16. PHI +22.5 v. NE Away 47.5

No changes in the spread, I don't think, with the exception of NE officially being 22.5. I read the line opened at 17 and within minutes it was 22 as money poured in. When they play the Jets later this season, I expect nothing less than a 55 point spread. Here are the moneyline ranks:

The Rankings - Moneyline

* considers lay in pick order
1. SEA -160 v. STL @Home x2!
2. CLE -170 v. HOU Away x2!
3. GB -190 v. DET @Home
4. TB -175 v. WAS Away
5. DEN +130 v. CHI Away
6. NO -140 v. CAR @Home
7. ARI -550 v. SF Away
8. NYG -320 v. MIN Away
9. JAC -350 v. BUF Away
10. SD -440 v. BAL Away
11. KC -240 v. OAK Away
12. IND -700 v. ATL @Home

I've altered the moneyline system slightly for this week, this is a trial run. I don't have much in the way of historical lays on moneylines, if anyone knows where I can get this info, I would love to have it. Right now the moneyline is chosen by the best Alpha, Sharpe, and Win Percentage. The new caveat is the rank order, which is now calculated using cost (the lay) per unit of risk (beta). This will move high cost games to the bottom if they are risky. For the small historical sample, it seems to work well at weeding out expensive and risky games. No telling yet how well the top 3 perform yet with any solid historical perspective, so we'll revisit this again later when I have better data.

I will be posting my final bets shortly once I decide on the best combination of risk. Lots of double bets this week in my top ranks, and also some nice hedges (STL +3 x2, SEA -160 x2) which will mitigate some of the risk while giving up some returns - worthwhile considering last week's debacle. What the Box says, Matt does.

Week 12 - Get out the Red Pen! Vegas vs the Mad Capper

Red Ink Everywhere! This means I'm smarter than the gambling world, or completely dilusional. I do find it interesting that Vegas upped the NE spread like I suggested, I won't be betting on a 22.5 spread, i wonder what the gambling world will do. Here's the breakdown:

VEGAS ................THE MAD CAPPER............... DIFF.
GB (-3) at DET....... SD (-11)..................... 8.0
NYJ (+14) at DAL..... DAL (-11).................... 3.0
IND (-12) at ATL..... IND (-8)..................... 4.0
TEN (-1) at CIN ..... TEN (-5)..................... 4.0
BUF (+7.5) at JAX.... JAX (-4)..................... 3.5
OAK (+5.5)at KC ..... KC (-3)...................... 2.5
HOU (+3.5)at CLE .... CLE (-2)..................... 1.5
SEA (-3) at STL...... SEA (-5)..................... 2.0
MIN (+7.5) at NYG.... NYG (-8)..................... 0.5
NO (-3) at CAR....... NO (-7)...................... 4.0
WAS (+3) at TB....... TB (-1) ..................... 2.0
SF (+10) at ARI...... ARI (-9)..................... 1.0
DEN (+2)at CHI....... DEN (-4)..................... 6.0
BAL (+9) at SD....... SD (-3)...................... 6.0
PHI (+22.5) at NE.... NE (-39)..................... 16.5
MIA (+16) at PIT..... PIT (-10).................... 6

The Thanksgiving Holidays create some urgency this week. My #1 prelim pick was GB over DET, figuring DET would be given too much credit, and I was dead on. It's GB-3 and I need to get my picks involving the Packers in today. So here's how my week is currently looking:

6-point Teaser: [GB+3 -- JAX -1.5 -- ARI -4 -- KC +0.5]

3-team parlay: [DEN +2 -- KC +5.5 -- TEN -1]

Straight Up:
TEN -1
DEN +2

Strangely enough, the more I looked at the games, the more nervous I became about the GB-DET game. DET plays much stronger at home, and is capable of a big game. It's also a holiday weekend, the home crowd will be fired up, and seriously... doesn't GB have to slip up once? If they are going to slip, this is a candidate. Plus, the box disagrees with me, so I just used them in the Teaser.

Happy Capsgiving!

Tuesday, November 20, 2007

Week 12 - Early Picks and Wall Street Ranks

The early picks are definitely preliminary, as Vegas can't seem to decide how to handicap the NE game. Again. In the rankings, I've highlighted to DET/GB game, where GB is currently favored by 3 to give an example of how the financial rankings can make the difference in a pick. The Sharpe and Alpha rankings are exactly flip flopped, and these teams execute very well, but the Beta indicates that GB is far more likely to do it again. Of course, because it's a 3 point game, the Box picks DET (the game is outside the <3 point margin for financial override), but at least it illustrates some things to consider. Here are the rankings and picks:

Sharpe Alpha Beta
Team Rank Rank Rank
ARI 9 13 6
ATL 31 32 30
BAL 27 19 11
BUF 28 30 9
CAR 26 11 19
CHI 21 28 32
CIN 7 5 15
CLE 3 7 25
DAL 2 1 31
DEN 22 25 2
DET 12 6 22
GB 6 12 12
HOU 11 10 1
IND 5 3 26
JAC 16 27 23
KC 29 31 29
MIA 23 15 21
MIN 20 17 7
NE 1 2 8
NO 14 24 24
NYG 8 4 28
NYJ 25 14 13
OAK 24 9 3
PHI 15 22 20
PIT 4 8 17
SD 10 18 27
SEA 13 16 5
SF 32 23 16
STL 30 29 4
TB 19 26 18
TEN 18 21 14
WAS 17 20 10

* considers lay in pick order Under
1. MIN +7.5 v. NYG Away 42.5
2. BUF +7.5 v. JAC Away x2! 42.5
3. STL +3 v. SEA @Home x2! - Bet Soon 38.5
4. SF +10 v. ARI Away - Bet Soon 35.5
5. DET +3 v. GB @Home - Bet Soon 42.5
6. WAS +3 v. TB Away - Bet Soon 43.0
7. CIN +1 v. TEN @Home 40.0
8. CAR +3 v. NO @Home - Bet Soon 38.5
9. DEN +2 v. CHI Away 40.5
10. BAL +9 v. SD Away - Bet Soon 40.5
11. NYJ +14 v. DAL Away - Bet Soon 44.0
12. MIA +16 v. PIT Away - Bet Soon 40.5
13. OAK +5.5 v. KC Away 38.0
14. CLE -3.5 v. HOU @Home 40.5
15. ATL +12 v. IND @Home - Bet Soon 42.0
16. PHI +22.5 v. NE Away 47.5

I expect the picks to change when the NE spread is finalized and by tomorrow evening when the spreads move a bit further off their opening lines, so stay tuned for a repeat pick post tomorrow or early on Turkey Day. The only favorite on the roster this week is CLE, but I'm glad it's the Box's 14th ranked pick, since HOU's return of Johnson and Schaub makes me nervous. The combo has the ability to be a Box buster, even against a scrappy Derek Andersen.

Notice in the WS Ranks how NE's beta actually DROPPED from last week because they outscored themselves, hence adding a layer of volatility. I'm not sure how DAL is even in the same league as NE in terms of output, but I think the Alpha rankings are the only rankings in the country that have DAL number 1 and NE number 2. That should at least make Dallas fans happy, until of course Dallas plays the Pats again for the ring and loses by 20 or 30 or 40 points. It's got to make you feel warm and fuzzy to see San Francisco and Miami ranked above anyone in SOMETHING, even if it is for consistent poor play.