College Football Week 10 Rankings

It must have been a pretty unpredictable week in college football as my model only got 52% of the games correct. Prior to this, my worst week for the season prediction wise was week 7 at 64%. All other weeks I predicted over 70% of games correctly. The overall rate for the season still sits at 73%.

1. Alabama
2. Clemson
3. Oklahoma
4. Ohio State
5. Georgia
6. Notre Dame
7. Louisiana State
8. Central Florida
9. Michigan
10. Washington State
11. Penn State
12. Wisconsin
13. Utah
14. Mississippi State
15. Boise State
16. Iowa State
17. Northwestern
18. Kentucky
19. Auburn
20. Michigan State
21. Iowa
22. Florida
23. Stanford
24. Washington
25. North Carolina State

CFL Week 20 Picks

A bit late with my picks this week. This looks especially dubious as I have Winnipeg beat Calgary, but their home advantage made them a 51% favourite. Hamilton at 54% over Ottawa, Saskatchewan are 67% versus the BC Lions and Toronto are 55% to beat Montreal.

College Football Week 9 Picks

Picks for all of this week's (week 9) 56 College Football FBS games, as predicted by my Elo model. My overall rate of prediction for the season is 75%.

[Troy] at South Alabama
Ball State at [Ohio]
Baylor at [West Virginia]
Toledo at [Western Michigan]
[Appalachian State] at Georgia Southern
Georgia Tech at [Virginia Tech]
Louisiana Tech at [Florida Atlantic]
[Miami (FL)] at Boston College
Indiana at [Minnesota]
Wyoming at [Colorado State]
[Utah] at UCLA
[Army] at Eastern Michigan
Bethune-Cookman at [Nebraska]
Central Michigan at [Akron]
[Clemson] at Florida State
Purdue at [Michigan State]
Texas Tech at [Iowa State]
Massachusetts at [Connecticut]
Vanderbilt at [Arkansas]
[Wake Forest] at Louisville
[Wisconsin] at Northwestern
North Carolina at [Virginia]
Coastal Carolina at [Georgia State]
[Southern Mississippi] at Charlotte
Oregon State at [Colorado]
[Texas Christian] at Kansas
Arizona State at [Southern California]
[Duke] at Pittsburgh
[Georgia] vs Florida
Illinois at [Maryland]
Iowa at [Penn State]
Kansas State at [Oklahoma]
[Middle Tennessee State] at Old Dominion
Northern Illinois at [Brigham Young]
Cincinnati at [Southern Methodist]
[South Florida] at Houston
[Kentucky] at Missouri
New Mexico at [Utah State]
Rice at [North Texas]
[Nevada-Las Vegas] at San Jose State
[Washington] at California
Arkansas State at [Louisiana-Lafayette]
[Boise State] at Air Force
[North Carolina State] at Syracuse
[New Mexico State] at Texas State
Texas A&M at [Mississippi State]
[Tulane] at Tulsa
Washington State at [Stanford]
[Florida International] at Western Kentucky
Tennessee at [South Carolina]
[Alabama-Birmingham] at Texas-El Paso
Navy vs [Notre Dame]
[Texas] at Oklahoma State
Hawaii at [Fresno State]
[Oregon] at Arizona
[San Diego State] at Nevada

NFL Week 8 Picks

Week 8 NFL picks from my Elo model are detailed below. Winners are [these ones], as ever. I do not think any of these are particularly egregious. The only ones that stick out to me this week are the Cardinals  and Raiders winning. Both are still probably rather overrated in my model. San Francisco are not that good and Arizona are at home, hence they come out ahead as a 60% favourite. It is a similar story with Oakland and Indianapolis.

Pittsburgh are my biggest favourites this week, at close to 90%. Let us hope that they can do more than just tie with the Browns when they host them. The Chiefs, Rams and Patriots are also big favourites, no surprises there.

Miami Dolphins at [Houston Texans]
[Philadelphia Eagles] vs Jacksonville Jaguars
New York Jets at [Chicago Bears]
Tampa Bay Buccaneers at [Cincinnati Bengals]
Seattle Seahawks at [Detroit Lions]
Denver Broncos at [Kansas City Chiefs]
[Washington Redskins] at New York Giants
Cleveland Browns at [Pittsburgh Steelers]
Baltimore Ravens at [Carolina Panthers]
Indianapolis Colts at [Oakland Raiders]
Green Bay Packers at [Los Angeles Rams]
San Francisco 49ers at [Arizona Cardinals]
New Orleans Saints at [Minnesota Vikings]
[New England Patriots] at Buffalo Bills

College Football Rankings Week 9

The top 50 college football FBS team from my Elo model going into week 9 of the season are listed below. I have extracted the top 50 from my model this week as I saw some grumblings online about Appalachian State being included in the AP top 25 so I wanted to see where I had them rated compared to the other teams. The answer was 46th:

1. Alabama
2. Clemson
3. Oklahoma
4. Ohio State
5. Georgia
6. Notre Dame
7. Louisiana State
8. Central Florida
9. Michigan
10. Wisconsin
11. Penn State
12. Washington
13. Stanford
14. North Carolina State
15. Southern California
16. Washington State
17. Iowa
18. Utah
19. Florida
20. Texas A&M
21. Miami (FL)
22. Texas
23. Auburn
24. Virginia Tech
25. Mississippi State
26. Boise State
27. Iowa State
28. South Florida
29. Michigan State
30. Kentucky
31. Fresno State
32. West Virginia
33. Florida State
34. Northwestern
35. Oregon
36. Texas Christian
37. Virginia
38. Texas Tech
39. South Carolina
40. North Dakota State
41. Army
42. Purdue
43. Duke
44. Oklahoma State
45. San Diego State
46. Appalachian State
47. Boston College
48. Mississippi
49. Houston
50. Kansas State

CFL Week 19 Picks

Picks for this week's CFL games below. I realise I did not get this out in time for the first two picks so you will just have to trust me on those ones. I had Ottawa as a 55% favourite and BC Lions as a 59% one.

Toronto are a 65% favourite over Montreal and Calgary 'only' a 67% favourite over Saskatchewan.

NFL Week 7 Picks

Picks for the week 7 NFL games listed below. Predicted winners are [these ones].

I do not think (m)any of them are particularly contentious. The two more irrational ones that jump out to me are the Cardinals winning the Thursday Night Football game against the Broncos, although this is only due to home advantage, and predicting the Bills to beat the Colts on the road with a backup quarterback (although their first choice is only a rookie). The latter seems rather speculative and the Colts are a big favourite with the oddsmakers.

The closest game in my model is the Saints at the Ravens. Both teams are fairly highly rated by my model (four wins each this season) but the Saints just edge this, even with the Ravens' home advantage.

There are no huge favourites this week. No team has a win probability of over 78%. But the moderately big favourites, with percentage win chance in the seventies, in descending order are: the Falcons, the Buccaneers (somewhat surprisingly but the Browns are really lowly rated still) the Rams, the Patriots, the Chiefs and the Vikings.

Denver Broncos at [Arizona Cardinals]
Tennessee Titans vs [Los Angeles Chargers]
[New England Patriots] at Chicago Bears
[Buffalo Bills] at Indianapolis Colts
Detroit Lions at [Miami Dolphins]
[Minnesota Vikings] at New York Jets
Carolina Panthers at [Philadelphia Eagles]
Cleveland Browns at [Tampa Bay Buccaneers]
Houston Texans at [Jacksonville Jaguars]
[New Orleans Saints] at Baltimore Ravens
[Los Angeles Rams] at San Francisco 49ers
Dallas Cowboys at [Washington Redskins]
Cincinnati Bengals at [Kansas City Chiefs]
New York Giants at [Atlanta Falcons]

Happy punting!

College Football Week 8 Picks

I noticed a few errors in my college football FBS model so had to re-run the whole thing, back from the 2000 season. This resulted in my current top 25 college teams, here. 

My overall prediction rate for the season through the first seven weeks (plus week 0) is 74%. Predictions for all 55 of the week 8 FBS games below. Winners be [these ones]. Highlight of this week is the Auburn Strawberry Blondes versus the Mississippi Cream Pies.

Georgia State at [Arkansas State]
[Stanford] at Arizona State
Colorado State at [Boise State]
[Air Force] at Nevada-Las Vegas
Auburn at [Mississippi]
Buffalo at [Toledo]
Cincinnati at [Temple]
Illinois at [Wisconsin]
Maryland at [Iowa]
Miami (OH) at [Army]
Michigan at [Michigan State]
[Northwestern] at Rutgers
[Oklahoma] at Texas Christian
Tulsa at [Arkansas]
North Carolina at [Syracuse]
Virginia at [Duke]
Bowling Green State at [Ohio]
Idaho State at [Liberty]
[Florida Atlantic] at Marshall
[Utah State] at Wyoming
Charlotte at [Middle Tennessee State]
[Eastern Michigan] at Ball State
[Western Michigan] at Central Michigan
[Akron] at Kent State
[Alabama] at Tennessee
[Coastal Carolina] at Massachusetts
Colorado at [Washington]
[Houston] at Navy
North Carolina State at [Clemson]
[Penn State] at Indiana
Kansas at [Texas Tech]
Louisiana-Lafayette at [Appalachian State]
[Minnesota] at Nebraska
Southern Methodist at [Tulane]
Texas-El Paso at [Louisiana Tech]
Wake Forest at [Florida State]
[California] at Oregon State
Memphis at [Missouri]
[Georgia Southern] at New Mexico State
Mississippi State at [Louisiana State]
Texas State at [Louisiana-Monroe]
[Central Florida] at East Carolina
Connecticut at [South Florida]
Texas-San Antonio at [Southern Mississippi]
[Fresno State] at New Mexico
North Texas at [Alabama-Birmingham]
[Ohio State] at Purdue
Old Dominion at [Western Kentucky]
Oregon at [Washington State]
Rice at [Florida International]
Vanderbilt at [Kentucky]
[Southern California] at Utah
Arizona at [UCLA]
San Jose State at [San Diego State]
Nevada at [Hawaii]

NFL Weeks 5 and 6 Re-Cap

I skipped on my NFL betting update last week for a few reasons. Firstly, I noticed errors in both my NFL and college football models so was spending time correcting those. Secondly, I was annoyed at my bad results in week 5 so did not feel like posting, the details are all below though. However, as Week 6 was my biggest winning week to date, I am happy about my model again now.

Outright predictions wise, I got 60% (9/15) right in week 5 and 73% (11/15) correct in week 6. This has pulled my overall prediction rate up to 58%. This is starting to get respectable but my target is around mid-60 percent correct.

I had four bets in week 5 and went 2-2. This was an annoying week for me as I very rarely have losing weeks and this was my second in three. A lot of my bets tend to be on favourites (they are favourites for a reason and generally win). This means that when they do lose you tend to lose quite big(ly) and winning the other bets does not tend to make up for the shortfall.

The four bets were:

Carolina Panthers (hosting the New York Giants) at -303
Baltimore Ravens (at the Cleveland Browns) at -150
New England Patriots (hosting the Indianapolis Colts) at -520
Dallas Cowboys (at the Houston Texans) at +150

The two winning bets were the Panthers and Patriots, the two losing bets were the Ravens and Cowboys. This gave me an ROI for the week of -32% and resulted in an overall loss-making record for the season to date (the first time this year). Fortunately, this was recovered in week 6, with my biggest winning week of the season, but more on that later.

I was very relieved that the Panthers won with their last-minute field goal, that was very close. I had resigned myself to losing and the Panthers probably deserved to. 

Baltimore losing to the Browns was a big upset, in my opinion. I felt that getting the Ravens at -150 against the Browns was a steal. I had just been bragging about betting against the Browns and how great it was when this result went and happened. This took me into a slight losing record (money wise) against the Browns, which annoyed me. This was corrected in week 6 though, with another bet against the Browns and it looks like I may bet against them again in week 7.

The Dallas bet was much more speculative and I was not particularly happy about placing it. They do seem like a pretty bad team yet they are 3-2 for the season and just put 40 points on Jacksonville. Yet I still have the feeling my model over rates them (and under rates the Texans a bit). This bet was right on the cusp of the allowable limit (set by my model) of being placed. But as it was still saying it was a good bet at closing odds time, I placed it. I always place a bet if my model tells me to, even if it feels bad to me. In the end they only lost by 3 points, so it does not look too terrible in hindsight, given they were +150.

The Patriots won at home to the Colts – no surprise there.

Did I mention week 6 was my biggest winning week thus far this season? The four bets placed were;

The Philadelphia Eagles (at the New York Giants) at -125
The Buffalo Bills (at the Houston Texans) at +360
The LA Chargers (at the Cleveland Browns) at -120
The Miami Dolphins (hosting the Chicago Bears) at +225

The Eagles were only -125 at the Giants, how could one not bet on that? They opened at -160 and then the money went in on the Giants to reduce the odds on the Eagles. Have people seen the Giants play? The Eagles might not have looked great this season but neither have the Giants. A pretty easy win in the end.

The only losing bet, on the Bills, was a very speculative bet and only came about because of their long odds. The Bills probably are over-rated in my model, and the Texans relatively under-rated, but the bet threshold was +300, so it was not suggesting betting unless the odds were particularly wide. The Bills only lost because of a pick six from their backup quarterback in the last 90 seconds of the game, so I do not feel too bad about this one. The longer the odds, the smaller my bets tend to be, so it was not a large amount of money to lose.

For the second week in a row, I felt like the Browns were massively over-rated in the odds. I bet on the Chargers at -120 because I thought the line was about to drift back, but they closed at about -105 after going as wide as +105 so the Browns started as -116 favourites (the Browns!). The Chargers are not that bad a team. The Browns are not terrible (not AS terrible as they were anyway) but I do not see how they are a favourite against this year’s Chargers. In the end, the Chargers won comfortably so my take was justified. This meant I was profitable against the Browns again, saved from embarrassment. Current BOI is +13%.

I was quite the braggart on twitter (follow me @FzBtz) after the Dolphins win because these odds seemed particularly good for them. They have had a good start to the season and it is not as if the Chicago Bears are some amazingly awesome team, just pretty solid. What I did not realise was that it was the Dolphin's quarterback playing. 

My model does not take injuries and trades into account explicitly. However, to a certain extent this is rolled up into the thresholds for when a bet should be placed or not. What one will often find is that a team becomes a suggested bet in the model if a star player is injured. I would say this is due to over-compensation of bookmakers and bettors to such occurrences and still provides some value situations. I won a big bet on the Packers last year against the Bears (again, ha!) after Rodgers was injured. Without that bet, the Packers would have been a slight losing team for me in 2017.

None of that matters now as the Dolphins won. However, the odds did not provide quite the value I initially thought they had. I do not play close attention to injury reports and other news going into the games. I do not need to as I just look at the games schedule, the odds and the results. These are the only inputs to my model and I barely even watch the games. I especially do not watch the games I have bet on as I find it too intense to watch. I live in the UK so the first round of games begin at 6 pm, then 9 pm, then 1 am. I have to get up early on a Monday so I am usually in bed just after the second round of games start. This does not give me much opportunity to really watch the games. Frankly, I find American Football pretty boring to watch in real time. The condensed games, with all the action in 40 minutes, are fine though.

All that blathering is irrelevant now though as I have an ROI of +10% for the year. This is a respectable rate and I would be happy to finish the season with an overall rate around this level.

College Football Rankings Week 8

I found a bit of a glitch in the matrix of my college football FBS model this week so I had to re-run the whole thing from scratch, going all the way back to 2000. There were a few games from my source material that have the home and away teams the wrong way around and I did not realise that 'Louisiana' and 'Lafayette' were the same team in the source games I had (even though they had a different name for different years) so had to consolidate all of that.

This means that some of my predictions in this season to date and the individual team ratings may have been a touch off, but I do not think it made too much difference to the top of the rankings. At some point I will post all of the predictions that the model came out with for the previous weeks - that will be one long post!

I believe week 7 was pretty random in the FBS. I cannot really comment as all I do is enter the games, then enter the results and my model does the rest. There are far to many games and teams for me to really keep track of what is going on - I struggle with just the NFL.

1. Alabama
2. Ohio State
3. Clemson
4. Oklahoma
5. Georgia
6. Notre Dame
7. Louisiana State
8. Central Florida
9. Wisconsin
10. Southern California
11. Michigan
12. North Carolina State
13. Penn State
14. Washington
15. Michigan State
16. Iowa
17. Florida
18. Texas A&M
19. Stanford
20. Washington State
21. Miami (FL)
22. Texas
23. Mississippi State
24. Virginia Tech
25. Iowa State

CFL Week 18 Picks

I have re-taken the lead from my CFL.ca game picking arch-nemeses.

I am bucking the trend a bit this week by picking Saskatchewan against Winnipeg. It is close though, I have it as virtually a coin-flip. Hamitlon and Edmonston are not big favourites either as both are less than 60% to win. 

I do not know why anyone would pick against Calgary at home. I have them at over 80% to win. The only reason could be as a hail Mary to try and regain ground. Here's looking at you Jamie Nye, the non-science guy, and Chris Delirious O'Leary.

They say that the ability to draw a perfect circle is the sign of a psychopath. Everyone in my life should be worried when they see my Edmonton Eskimos logo below.

College Football Week 7 Picks

I am in the process of re-running my historical College Football from scratch but in the meantime, here is what it is predicting for this week's college football FBS games. Winners are [these ones].

[Appalachian State] at Arkansas State
[Georgia Southern] at Texas State
Texas Tech at [Texas Christian]
[South Florida] at Tulsa
Air Force at [San Diego State]
Arizona at [Utah]
Akron at [Buffalo]
[Florida] at Vanderbilt
[Iowa] at Indiana
Minnesota at [Ohio State]
Nebraska at [Northwestern]
[Oklahoma State] at Kansas State
Rutgers at [Maryland]
Tennessee at [Auburn]
[Toledo] at Eastern Michigan
Duke at [Georgia Tech]
Louisville at [Boston College]
[Alabama-Birmingham] at Rice
Southern Mississippi at [North Texas]
[Troy] at Liberty
Kent State at [Miami (OH)]
Pittsburgh at [Notre Dame]
Ball State at [Central Michigan]
[Western Michigan] at Bowling Green State
[Army] at San Jose State
Baylor at [Texas]
[Georgia] at Louisiana State
Michigan State at [Penn State]
[Purdue] at Illinois
[Marshall] at Old Dominion
Ohio at [Northern Illinois]
Temple at [Navy]
Texas A&M at [South Carolina]
[Central Florida] at Memphis
[Washington] at Oregon
[Western Kentucky] at Charlotte
New Mexico at [Colorado State]
Nevada-Las Vegas at [Utah State]
Alabama State at [South Alabama]
New Mexico State at [Louisiana-Lafayette]
Louisiana-Monroe at [Coastal Carolina]
[Houston] at East Carolina
[Louisiana Tech] at Texas-San Antonio
[Miami (FL)] at Virginia
Missouri at [Alabama]
UCLA at [California]
[Virginia Tech] at North Carolina
West Virginia at [Iowa State]
[Middle Tennessee State] at Florida International
[Mississippi] at Arkansas
[Wisconsin] at Michigan
Hawaii at [Brigham Young]
[Boise State] at Nevada
Colorado at [Southern California]
Wyoming at [Fresno State]

NFL Week 6 Predictions

I have had a busy week with other things so not many posts this week. I also noticed a glitch in the matrix in my NFL model in the 2017 season which has caused a rejig of some of the team's ratings. Essentially this means the Kansas City Chiefs are a bit higher rated than they previously were (now top of my rankings).

Predictions for this week (I know I missed the Eagles game, whom I had a bet on, woohoo!). Winners identified like [this].

[Philadelphia Eagles] at New York Giants
Tampa Bay Buccaneers at [Atlanta Falcons]
Pittsburgh Steelers at [Cincinnati Bengals]
[Los Angeles Chargers] at Cleveland Browns
[Seattle Seahawks] at Oakland Raiders
Chicago Bears at [Miami Dolphins]
Arizona Cardinals at [Minnesota Vikings]
Indianapolis Colts at [New York Jets]
[Carolina Panthers] at Washington Redskins
[Buffalo Bills] at Houston Texans
[Los Angeles Rams] at Denver Broncos
Jacksonville Jaguars at [Dallas Cowboys]
Baltimore Ravens at [Tennessee Titans]
Kansas City Chiefs at [New England Patriots]
San Francisco 49ers at [Green Bay Packers]

CFL Week 17 Picks

Picks for this week's CFL games artistically displayed below.

Calgary are the biggest favourites again this week, what a surprise. The other three picks (BC Lions, Ottawa Redblacks and Saskatchewan Roughriders) are all fairly equally, decent sized favourites.

College Football Week 6 Picks

Predicted winners for all of this week's (week 6) college football FBS games, as determined by my Elo model. Picked winners are identified [thusly].

I do not know enough about college football and its teams to really comment on any of these match-ups (I hear Alabama are good, right?) I just paste it the output from my model. The game that stands out to me this week is the Liberty Bell-Ends against the New Mexico State Border Patrol Dogs.

Georgia State at [Troy]
Tulsa at [Houston]
Georgia Tech at [Louisville]
Middle Tennessee State at [Marshall]
Utah State at [Brigham Young]
[Alabama] at Arkansas
[Buffalo] at Central Michigan
East Carolina at [Temple]
Eastern Michigan at [Western Michigan]
Illinois at [Rutgers]
Kansas at [West Virginia]
Maryland at [Michigan]
Missouri at [South Carolina]
Northwestern at [Michigan State]
Syracuse at [Pittsburgh]
[Oklahoma] vs Texas
Tulane at [Cincinnati]
Boston College at [North Carolina State]
[Northern Illinois] at Ball State
[Iowa] at Minnesota
Bowling Green State at [Toledo]
[Clemson] at Wake Forest
Florida State at [Miami (FL)]
Iowa State at [Oklahoma State]
[Kansas State] at Baylor
[Louisiana State] at Florida
Miami (OH) at [Akron]
Navy at [Air Force]
[Notre Dame] at Virginia Tech
[Ohio] at Kent State
San Diego State at [Boise State]
South Alabama at [Georgia Southern]
[South Florida] at Massachusetts
Arizona State at [Colorado]
Indiana at [Ohio State]
New Mexico at [Nevada-Las Vegas]
Louisiana-Monroe at [Mississippi]
Old Dominion at [Florida Atlantic]
[Auburn] at Mississippi State
[Kentucky] at Texas A&M
Vanderbilt at [Georgia]
Alabama-Birmingham at [Louisiana Tech]
Connecticut at [Memphis]
[Louisiana] at Texas State
Southern Methodist at [Central Florida]
[Texas-San Antonio] at Rice
Nebraska at [Wisconsin]
[North Texas] at Texas-El Paso
Utah at [Stanford]
[Washington] at UCLA
[Liberty] at New Mexico State
[Washington State] at Oregon State
California at [Arizona]
[Colorado State] at San Jose State
[Fresno State] at Nevada
Wyoming at [Hawaii]

CFL Power Rankings Week 17

The rankings of the CFL teams, according to my Elo model, following week 16 are listed below.

The Stampeders are way out in front, unsurprisingly. Edmonton have dropped from 3rd to 5th after their heavy defeat to Winnipeg. Saskatchewan have a reasonable gap to Ottawa (in terms of Elo rating) and then Ottawa have their own clean air to the tight clump of teams from 4th to 7th. 

Montreal are all out by themselves too, right at the bottom. Toronto are trending in their direction though as the Aloutettes have won more recently.

I would do a snazzy chart of the trends, like I did for the NFL this week, but I cannot be bothered formatting it right now...

1. Calgary Stampeders
2. Saskatchewan Roughriders
3. Ottawa Redblacks
4. Winnipeg Blue Bombers
5. Edmonton Eskimos
6. Hamilton Tiger-Cats
7. BC Lions
8. Toronto Argonauts
9. Montreal Alouettes

NFL Week 5 Picks

Game winner picks for each of the week 5 NFL games are listed below. 

The Patriots are the biggest favourite of the week, I have the Colts on a really low rating. The Panthers and Ravens are also pretty big favourites as the Giants and Browns are also two low rated teams. The closest game is the Broncos at the Jets with home advantage pushing up the Jets chances somewhat.

I do not think any of the picks are too egregious. The Lions beating the Packers might be a touch optimistic but neither side is particularly amazing this year and the Lions have home advantage. 

Indianapolis Colts at [New England Patriots]
[Tennessee Titans] at Buffalo Bills
Miami Dolphins at [Cincinnati Bengals]
[Baltimore Ravens] at Cleveland Browns
Green Bay Packers at [Detroit Lions]
Jacksonville Jaguars at [Kansas City Chiefs]
[Denver Broncos] at New York Jets
Atlanta Falcons at [Pittsburgh Steelers]
New York Giants at [Carolina Panthers]
Oakland Raiders at [Los Angeles Chargers]
Minnesota Vikings at [Philadelphia Eagles]
Arizona Cardinals at [San Francisco 49ers]
[Los Angeles Rams] at Seattle Seahawks
[Dallas Cowboys] at Houston Texans
Washington Redskins at [New Orleans Saints]

NFL Week 5 Power Rankings

The Patriots were back in business in week 4 of the NFL and rise back up to the top of my Elo model's relative team rankings, full list below. This was also helped but the Eagles losing to the moderately rated Titans which also assisted the Chiefs in their rise to now second in the list. The Rams are now also hot on the heels of the Eagles as their impressive start to the season continues.

The relative gaps and transitions can be seen in the graph at the bottom of the post. Some of it can be a bit hard to make out, due to the big clump of teams in the middle. Also, there are only a finite amount of different shades of blue and red the human eye can really, categorically discern (no matter what Dulux try and tell you) when placed next to each other.

The teams have generally closed up, especially if you ignore the Cleveland Browns (look at that gap to them!). This is mainly due to the Patriots and Eagles losing relative ground to the rest of the teams.

The top 6 teams are fairly well separated but then it is very close between the Vikings in 7th down to the Bengals in 14th.

At the bottom of the rankings is the lonely Browns and then the Colts, Texans, Jets and Giants in a group of 4 with the Cardinals desperate to join them. The Raiders downward trend has head a bit of a stay, only because they got to play the Browns at home, but otherwise i am sure they will be down there soon too.

1. New England Patriots
2. Kansas City Chiefs
3. Philadelphia Eagles
4. Los Angeles Rams
5. New Orleans Saints
6. Carolina Panthers
7. Minnesota Vikings
8. Tennessee Titans
9. Pittsburgh Steelers
10. Baltimore Ravens
11. Jacksonville Jaguars
12. Dallas Cowboys
13. Atlanta Falcons
14. Cincinnati Bengals
15. Seattle Seahawks
16. Los Angeles Chargers
17. Miami Dolphins
18. Green Bay Packers
19. Buffalo Bills
20. Washington Redskins
21. Detroit Lions
22. Tampa Bay Buccaneers
23. Chicago Bears
24. Denver Broncos
25. San Francisco 49ers
26. Oakland Raiders
27. Arizona Cardinals
28. New York Giants
29. New York Jets
30. Houston Texans
31. Indianapolis Colts
32. Cleveland Browns

NFL Week 4 Re-Cap

My model predicted 10 of 15 games right in week 4 of the NFL. Finally, some success. I am only 52% correct for the season having got 9, 8 and 6 games correct in each of the first three weeks of the season thus far (respectively?).

In general, this week seemed a lot more straightforward (predictable) compared to the relative carnage of the first three weeks of the season. Although Philadelphia losing to Tennessee was a bit of a surprise. I only had the Eagles as a 58% favourite so it was not the biggest upset by any means. Baltimore beating the Steelers in Pittsburgh and Cincinnati beating the Falcons in Atlanta were considerably bigger upsets in my model as those two away teams only had a 33% chance of winning.

It was also a pretty straightforward week at the bookies as my model did not pick up many spots for betting i.e. the lines were set at about the right levels without many edges. Buffalo were an early pick but their odds soon tightened which took them off the table, fortunately. They were only a potential bet due to their relatively long opening odds of +330, my model was never picking them to actually win the game.

Indianapolis were also identified as a bet at the start of the week but as punter confidence in them shrunk, causing their odds to lengthen, this bet also evaporated. Another bullet dodged. This is the opposite case of the context. In the first example the odds are shortening, in the second lengthening yet they both became nonviable bets. My model looks at both the difference between the game prediction and the odds but the way bet/no bet decision also incorporates market confidence.

The only other bet it threw up, the only one I did end up placing, was on the Oakland Raiders at -140 against the Cleveland Browns. At least the Raiders were at home, which is a considerable advantage in the NFL. Oakland's pitch has a few tonnes of sand in the middle of it which is not taken into account in my model, Nate Silver would probably build it in though.

I would say the Raiders are probably over-rated by my model but they are still the 6th lowest rated team. The Browns, in the other hand, are the perennial bottom feeder, both in the league and in my model. But they do probably have a more pessimistic rating than they deserve. This means they often come up as a team to bet against. I have bet against them in weeks 1 and 2 this year and it was a very close call against the Jets in week 3 (another bullet ultimately dodged by my model). I will probably bet against them in week 5 too.

What a nail-biter of a game it was, with the Raiders taking it to over-time with an 8 point drive in the final minute of regular time and then winning on their second possession of over-time after missing a field goal (from the aforementioned sand) earlier in the period. This win gave me an ROI of 71% for the week (duh, the single bet was at -140), 62% against the Browns (BOI - Browturn On Investment) and lifting my overall ROI to 6% after the previous week's slump. All of this is captured in the snazztastic graph below.

College Rankings After Week 5

I did not really pay much attention to the college football this weekend. I had other things on, such as styling my toe hair.

I got 44 of my 58 picks correct, giving me an overall rate of 76% for the season.

The latest rankings from the model are below. A few switches here and then and more shifts towards the bottom end of the rankings. The harder the opposition you play, the more points you gain, so even a winning week means you can drop if another team plays a relatively stronger team.

1. Alabama
2. Ohio State
3. Clemson
4. Georgia
5. Oklahoma
6. Louisiana State
7. Wisconsin
8. Notre Dame
9. Penn State
10. Washington
11. Central Florida
12. Stanford
13. Southern California
14. Auburn
15. Miami (FL)
16. North Carolina State
17. Oklahoma State
18. Michigan State
19. Texas Christian
20. Boise State
21. Virginia Tech
22. Texas
23. Washington State
24. Kentucky
25. West Virginia

CFL Week 16 Re-Cap

I went 3 for 4 of my CFL picks this past weekend. Edmonton decided they did not want to turn up so got destroyed by Winnipeg and broke my winning streak.

I actually had a bet on the CFL this week, for the first time in a month. Nothing too exciting though, the Stampeders as a big favourite (-800). This is actually the first time I have bet on the Stampeders this year, surprisingly. They have mostly been just too big a favourite for my model to pick them.

Saskatchewan were bouncing around the required odds (by my model) for a bet but ended up falling the wrong side of it. Part way through the game I was glad of this but in the end they managed to close it out. Johnny Football managed to through his first CFL touchdown (two even!) and led the Alouettes in rushing too, what a guy. One day he might actually win a game, like if the other team gets held up in a snowstorm and they are forced to play their under-18 squad instead or if the manager distracts the referees and he hits the other team’s quarterback with a steel chair. But one of these days Montreal will win another game and I will probably be on the losing end of it.

Latest snazzy graph of winnings below. Still just riding on the back of three games were Hamilton were over-rated earlier in the season.