WARNING: This page or the author does not condone gambling or betting on the NFL or other sports. This page is not affiliated with the NFL, Yahoo!, USA Today, Comcast Cable, any sports book, or corresponding affiliates or sponsors. All information on this page is for informational use only, and the author does not take responsibility for the accuracy or content of this page. Picks are based on simplistic data analysis and are OPINION. No guarantees or assertions are implied by the author's picks and are for discussion only.
This is not an "Insider" prediction site, and I am not a professional NFL handicapper. I am just looking at using seasonal statistical data to create a NFL "report card" ranking of how the teams are doing so far. The only input I am doing is through adjusting the power ranking categories allotment of points and home team advantage (HTA) points. If you want to gamble money (and I am not encouraging this), look at the handicapping suggestions below.
Click here for Spreadsheet examples.
USE THIS PAGE AT YOUR OWN RISK
How Statistical Data is used to compile the Reisterstown power rank.
Here is a general reference regarding statistical methods for computing sport team rankings, and a general bibliography.
The Reisterstown NFL Index predictor was started during the 2005 season. Changes and alterations to the index algorithm will occur for the first 3 or 4 years as I evaluate the divisors and determine which power point assignments are the most accurate/feasible. Please view the spreadsheets as a reference to what I am referring to on this page. The complexity of the spreadsheet itself, the exact numbers used in the power number assignments, and divisors used to determine how rankings and YPP influence the predicted scores are where the "meat" of this algorithm are. The reason I show the spreadsheets is that it is rare that ranking systems will show you what they are doing (probably because they don't want to share their secret.), and I have changed the points assignments, divisors and scaling system so many times that the earlier iterations of this spreadsheet are not reflective of what I am doing now. The spreadsheets (in image form) are a mere shell (or blueprint) of what I am doing now; however, if you have the time (and determination) to generate your own power ranking system using excel, you can use what I have done as a guide.
Please refer to 2005 NFL weeks 10-4 for more detail on how the excel charts are compiled. From NFL 2005 week 10, I will not post the excel images due to time and space (file size allocations) constraints. Refer to the existing excel sheet images to get a general "feel" for how the stats are used to generate a power ranking and score prediction.
However, If you decide to send me a donation, I might email you a copy of the most recent excel file for your scrutiny. (The spreadsheet it a bit of a mess, and quite complicated (NOT USER FRIENDLY)- I would only recommend experienced excel users ask for the file and you must make a formal request by emailing me before I will agree to send you the current spreadsheet file.)
The concept of generating a NFL team power scale - in the general scheme - is simple: Pick the factors that you feel determines a teams strength (points scored, points given, offensive rank, penalties, winning percentage, strength of schedule, key player rankings, Yards Per Point (YPP)......) and weight the factors to your liking . Look at each team and give each team a "scaled grade" for that individual factor based on your determined "weight." For example: let us say you determine the offensive rankings are scaled from 10 to 0 (A maximum weight of 10). Say the Colts are the #1 offense and the Jets are the #32 Offense, The Colts would get 10 points, the Jets would get 0 points for this factor. Add up the points given, for the various factors utilized, and see where the teams rank.
Basically, the results generated are from a compilation of the seasonal statistics. The team stats for offense and defense are obtained from a public data source (for example Yahoo! sports). The data is cut and pasted into a preformatted spreadsheet and sorted alphabetically. One this is done, all the power rankings are automatically generated. The ultimate goal is for the generated power rankings to be similar to a seasonal "Report Card" of how the teams are ranked.
Some factors (such as winning percentages, positional/injuries and strength of schedule) are calculated and adjusted (scaled and ranged). For example, Winning percentages are influenced by number of wins away (+), losses at home (-), and win streaks (+ or -). Strength of schedule is calculated from the opponents seasonal win-loss percentages and accumulated opponents scores (pf - pa). These opponent values are multiplied and scaled - (normalized) and ranged accordingly - then summed.
As of 12/7/05, I have removed determining HTA by a manual input. HTA is now set by calculating current win/loss percentage AT HOME, and subtracting that value from the overall w/l percentage. The values are then scaled and ranged from 2.0 to around 6.0. It may seem that teams are double dipped for wins away and severely punished for losses at home, and I feel this is justified. (perhaps these numbers will be scaled later to reduce the severity).
Key Positional Players (ie. top 15 QB, top 10 WR...) and injuries are linked together (for example, if a team has the #1 QB (32 pts given as an example) and #4 WR (14 pts given as an example), the team gets 46 pts. If a ranked player is injured and will not play that week, the points are not given to the team, and a 0 is given for that position. Injuries are determined by a graded scale (0 (none) - 10(all major players are hurt)) by looking at CBS NFL injury report - For example, If the scaled score for a team with multiple injuries to 5-6 key players is 7. The team gets - 28 (7*4) points for injuries. Thus the positional/injury number for the team is 18 pts (46-28 pts). The injury number is a qualitative number, and an obvious source of error. The divisors and multipliers will change for the various factors, but I hope you get the idea. So far, Injuries are the only manual input in this system that will affect the power ranking scores, and I have yet to find a way of determining a purely numerical method for determining a value for team injuries based on seasonal statistical data with no human "judgment." Also is seems that some teams will not reveal all injuries till late Saturday or over exaggerate their injury report, and this type of gamesmanship will also affect the power rankings through inaccuracy (error). It would be nice if injuries could just be ignored, but the reality is that injuries are a major determinant in determining game outcomes.
As the Season progresses, the power factor weights are adjusted in the power rankings (for example, at the beginning of the year W-L is obviously not important, and previous year's stats may be utilized for week 1-2), Typically by mid-season, stats such as points scored or given are weighed to 15-20% of the total value, W-L% are given 25-30% value and Yards Per Point (YPP) is given about 10-20% value with other factors (team rankings, strength of schedule, penalties, 3rd down conversion %, QB to RB rankings....) assigned a certain percentage to be combined in a summed total of power points. As noted above, the factor weights and scales are constantly changing with the hopes of obtaining optimal vales by reviewing the previous weeks score results to see how these weights and scales can be adjusted to produce more accurate predictions. Currently I have scaled the points scales of both YPP and stat power a scale of 0 to 70 +/- 10. this represents a large variation and I may tighten the scale even more.
Once the power ranking are determined, a spread is determined by subtracting the away power from the home power and adding the home team advantage. A spread from a simple statistical comparison (YPP (defense - offense) away - YPP (d-o) home + HTA) of the teams is then used to generate a spread. The 2 spreads are then averaged (typically with 70% weight to stats and 30% to YPP - this will change) to determine a final spread. Clearly you will have noted that YPP is a very important factor in this power and score predicting scheme. Click here to find out why. If these pages are down click here for a copy of the YPP pages.
Scores are then predicted from seasonal averages of points per game (PPG) scored and given, averaged according to the assigned opponents. For example the home team's PPG scored is averaged with the away teams PPG given to determine a central point for Home team points scored. This score is further adjusted according to YPP(d-o), and team defensive - opponents offensive rankings for pass and run. I have yet to determine how to scale the YPP and ranking scales to determine the most accurate means of determining an approximate home team score. As of now, I just tired to scale the YPP and rankings factors to influence the score +/- 7 points per game. Home team weather is then determined on Saturday, and if it is cold, snowing, heavy rain....the home team score is scaled down by 5 to 12 points.
I used the predicted home team score and simply added the predicted final spread to the away team to generate a final predicted score. If a negative number is generated for either (or both) teams, this is scaled accordingly to set the losing team to 0.0..
Posted scores and predictions are usually finalized on TUESDAY or Wednesday night (because I do not have time to make further corrections, but If I do have time I will alter the ranks for things like weather and injuries...). The posted predicted game spread outcomes may be changed by the ever changing spread lines.
Obvious errors and adjustments to be made
As noted above, the generated power rankings are thus a seasonal "Report Card" of how the teams are doing. The power rankings are influenced by seasonal stats, and early season biases may affect future power ranking generation. Some predictors throw out the first 1-3 week data, and reward more for current weekly results. While there is a early season bias, it is too difficult to weed this out, and hopefully it will be overshadowed by the sheer weight of the accumulated seasonal data. Some power systems focus on averaging the last 4 games played. While this sounds good, I don't have the time to keep my own stats, and I have to rely on seasonal compiled stats to cut into my spreadsheet and the time only takes about 1-2 hours to generate the power rankings.
Clearly the most important problem with this predictor is that there are far too many factors involved. Determining what the garbage stats versus the important stats is the first goal. After that, assigning meaningful power ranking percentages to these key stats are the main goals of this predictor. It is too early to determine what these percentages, number assignments and factor divisors are right now. Hopefully within 3-4 years, I will get a feel for how these numbers adjust during the season, and what divisors and percentages I should use in generating the power ranks and predicted scores. If not, then so be it.
As this power ranking is a summation of various factors, errors add up as well.....the standard deviation on the predictions are large. A factor that is not included in this ranking is motivation. Bubble teams that "need" to win a game to make the playoffs; teams that are riddled with chaos; and major rivals playing for sheer pride, are factors that statistics find difficult to emulate. Also, great teams inexplicably lose to pathetic rag dolls for no apparent reason each NFL week. Blame the spread, organized crime, a supreme will, entropy, or whatever you want, but there seems to some random factor involved in picking a perfect NFL week ATS (14-0) that thumbs its nose at the statistics. To try to insert such a random factor in this algorithm would make each game a coin flip, which is not the point.
Teams inexplicably have letdowns, and predicting blowouts or accurate spread results based on pure statistical data is almost impossible because factors that occur during the game will drastically affect how the scores are determined. Weather and team motivation are non statistical motivators that will influence the score. Injuries and game errors are also key factors that are not something stats can factor in. For example, an injury to a key player(s) or just a single stupid random error(s) by a key player(s) during the game will cause the score to flux dynamically in real time in relation to the given spread, and coaching philosophies during the game will alter spread predictions as well as over/under predictions. Also, how to factor "insider information" is impossible except for using a single HTA input. It is impossible listen to sportscenter and read every issue of The Sporting News, and filter out the sheer garbage from the important information. Whatever information, obtained in a brief scan of the papers, is adjusted for by Saturday night. or early Sunday morning.
All this being said, it is a wonder that any predictors are over the 50% mark in relation to the spread.
Once again, posted scores and predictions are finalized on TUESDAY night because I do not have time to make further corrections. YOU will have to make corrections (and factor in the nuances of each game) after the Tuesday night predictions are posted.
Current Adjustments in the works
I decided to add a scaled value for win and losses late in the season (week 14), and may not add it into the power ranking until week 16. I was hoping to use the Bradley-Terry model for paired comparisons to generate a scaled ranking based on games played. I decided try something else. Each week, a teams win or loss is scaled against the opponent faced. For example, lets say NE faces KC at KC. The score is NE 16 KC 26. NE will get a pf-pa of -10. KC has a w/l of 0.67, and this being a loss away use a scale of 0.5 ( 1 if win at home, 1.5 if lose at home or win away). The value of the loss is for NE is -1.68 (-10*.25*0.67). This is fine for wins, but a bad team is severely punished for losing big to a good team this way. I will probably just invert the negative (loss) values for this calculation and scale the overall weight from -10 (min) to +10 (max) for each game. I have changed this formula for w/l weighting a few times (using a delta of opponent w/l % - team w/l % as a multiplier, and utilizing the actual score - the predicted spread as a game index, and normalizing the result....but the idea is the same). Here is a sample line from the excel spreadsheet for ARZ from their game v. WAS week 14 2005.:
The win/loss weights, for each team, will be adjusted (scaled and ranged) and summed for the season. This win/loss weight for all the teams will be generated as I go along, and placed in the power ranking scale (giving it an estimated weight of 10-12% of the total overall power ranking) either by week 16 or perhaps by next year. I will have to see how the overall power numbers are affected by this w/l weighted rank and adjust the percentage of points allotted to the weighted rank of w/l as I go along......
General Handicapping suggestions (for amateur bettors)
1.
Use Power and stats to get a GENERAL sense of how (G)OOD or (B)AD a team is. Categorize the game as: G v. G, G
v. B, B v.B.
2.
EACH game must be analyzed independently
for winner,
points and spread.
Winners: Pre-determine who will likely win each game. Power and Overall team strengths are important here. Refer to multiple sources (see Todd Beck or David Wilson's sites) and use these systems to analyze/determine the winners for that week. Categorize game likely winners as Strong, Likely, and Questionable based on this research.
Points must be determined by comparing
previous weeks points made v. given and offensive v. defensive capabilities of
each team. Weather, injuries, Coaching philosophies are important points here.
Spread is a the most complex point since
bookmakers are only interested in splitting the betting field. Changes in the
spread are only reflections of how the field is betting, not reflections on who
will actually win the game. In determining if a team can cover the spread ALL
factors come into play, especially home team advantage, strength of schedule,
quality of previous wins or losses, injuries, weather, number of points
involved...... How a team covers may ultimately come to something a esoteric as
the team's defensive end having an argument with his wife before the game, a key
injury during the game, or a minor player making a single stupid vital random
error during the game, and cannot be reliably determined by statistics alone.
Ultimately, the coach may inexplicably sit on the ball with 10 minutes to go and
win by 4 with the spread at 4.5 points. (This is why the College BCS is loved by
gamblers, where the top teams are encouraged to pile up the points late in the
game. It is the sole reason why there is (and may never be) a college
playoff system) Only in circumstances where the spread is too high for the
favorite to cover is the spread more of a factor to be considered. Otherwise
look to take the favorite in a pickem, to take the underdog and the points most
of the time (as analysis dictates), and flip a coin for the others.
Ask
yourself.
1.
Is the spread less than 9 points?
2.
Is the favorite team's offense capable of generating a lead quickly and the
defense capable of stopping the underdog's offense cold?
3.
Is the favorite playing against a major rival at home? (motivation)
4.
Is the underdog having to deal with debilitating injuries?
1.
Is the favorite playing an underdog coming off a bye week? (rest and
preparation)
2.
Is the favorite, (playing away) having to deal with debilitating injuries or
travel to a hostile or bad weather opponent?
3.
Does the underdog have an offense or defense ranked in the top half of the
league?
4. Is the Underdog playing at home, playing a major conference rival, and coming off a strong showing against a top 10 team last week? (motivation)
5. Is the favorite undefeated with over 8 wins and coming off a tough game last week and against a supposedly weak opponent? (due for letdown)
If
any 3 of the last 5 answers are yes....then the underdog and the points become
more tasty, and further analysis of the underdogs chances to win + points may be
merited.
Otherwise,
what the heck, take the big favorite minus big points and pray they can cover.
3.
All the Stats and assumptions about the game can become GARBAGE with a single factor overshadowing all other factors.
(For example - Vick out means ATL's offense is shattered by 30%, and the
previous weeks offensive stats and rankings are generally useless) Weather,
Altitude, Travel, Determination, Team disarray or internal problems, Revenge,
Motivation...are examples of factors that must be used in your independent game
analysis. Try to avoid using trends (they are waiting to be broken).
5.
Use your bankroll wisely.
From here on in I must file this disclaimer: I am NOT a professional handicapper. All the material below is sheer personal opinion, and based on my personal experience and discussions with others and research on the internet. I have had numerous losses and wins, and do not pretend to have any authority on this subject. I just enjoy gambling in a recreational manner on the NFL. I try to avoid college because of the complexities involved (but I do love the BCS as it is because it promotes good teams to blow out their competition in the middle of the season). How I bet (and the reasons for betting in the way I do) are continually evolving; however my NUMBER ONE RULE of sports-betting is NEVER GO BEYOND YOUR INITIAL BANKROLL.
I like to bet about $35-45/week, and at 16 weeks+ playoffs that is about $740 (more on the $40 later) for the season. I am wary of the 1st - 3 weeks (teams unsure of who to play and sheer rust) and the last 2 weeks of the season (teams are looking to the playoffs and sitting key players). Never go beyond the initial bankroll (this is a recipe for disaster and financial ruin) , and remember this is just for fun, not something to do for a living ( for 99.9 % of the population). You must ask yourself, is this $700 disposable? - If not, then you should find another hobby and use the money for what you REALLY need, or wait till the 8th to 10th week to start and go from there with what you have.
I like to make parlay/teasers bets. Parlays are actually considered a bad bet by most because you can go 4-1, 3-1 and lose. For the professional gambler, it is best to bet each game individually, and (having a good system) bet appropriately. However, for the recreational gambler, that would mean staking over $70-140/week (at $5-$10/bet) and at 16 weeks that is $1,120-$2,240. My bank roll is about a half to a third of that amount, and I want to spread this out over an entire season. Thus, I decided to go with a riskier betting format, and just have fun (remember $700 disposable is MY MAX and I desperately try to stick to that....) Parlays are a way of spreading out my bankroll, have a reason to keep working at my power index, and still have a potential (be it a small one) at a profit. Also hitting a parlay/teaser is pure fun. Ideally, I hope to get a minimum of 4-7 correct (out of 14 games ATS per week = 50%) right. Thus it is never really acceptable to take a parlay greater than 7 teams. Taking a parlay beyond 7 teams is a true longshot, and not worth the investment (unless you have even more disposable income). Nevertheless, I do a 8 or 10 team parlay just for fun because the potential profits are huge. (The sports-books love parlays because it is rare for anyone to hit a 9 team parlay and above, and anything the sports-books love must be bad.) Here is an article regarding parlays.
Teasers are a more touchy subject. The profits are less, and, since you have to bet more to win more, it may be better to just bet games straight up. I guess I am a sucker here because I love getting max points towards a teaser bet, but once again, if you don't know what you are doing, a 3-1 loss is a total loss, and a unexplained blowout will screw your supposedly rock solid $20 13 point 4-team sweetheart tease. I always try to tease a good team back towards even or to a close win. It seems to be more likely that a bad team will lose by 25 points or more to an exceptional team, and it is more likely a good team will beat a bad team by more than 2 points (unless injuries or weather become a factor). Thus, a good team at -10 sweetheart teased +13 points seems to be a better deal than teasing the bad team to +23. There are exception, of course, and those are trickier to spot. I could be wrong on this, and this is why I just try to avoid taking too many sweetheart teasers after the 10th week.
Anyway,
I like to make 5-6 parlay/teaser bets. 1 - moneyline/sweetheart teasers ($15-20
for 3-4 teams), 2 -teaser/parlay ($5 for 3-4 teams), 3 - parlay ($ 5 for 3-4
teams), 4 - longshot parlay for fun ($5 for 6-8 teams) . The 5th bet is to hedge
the MNF game or just tease ($5 for 4 teams). If any of the 4-5 previous
bets are successful to the MNF game, to recover the money wagered at the
beginning, a 6th bet may be used as a sidelight hedge, but this is only for the
8:30pm game and the MNF game.
Remember, you are NOT a professional sports gambler, and this is just for fun. Never, never, never, never go beyond your set bankroll. If you do, then you have failed.
** It is important to note that each bet should include a MNF game bet (all the same) and/or a 8:30 bet for hedging purposes. Bets should avoid having a single team that is common to all bets to avoid a total collapse. Spread it out! I cannot tell you how many time a pure rock solid NFL lock team flushed hundreds of profit dollars down the toilet when they inexplicably collapsed to a team, that the week before could not beat your local JV practice squad.
6. Be prepared to lose. Bad weeks happen for no apparent reason, and bad weeks can turn into bad months and a bad season. Accept them and understand you have a bankroll to play with. Do not try to recover your losses by betting big next week or just keep depositing money into your account. Just do the analysis, make your bets and try to recover by taking more teams if the circumstances look good. Remember - this is for fun (remember $700 disposable...), and not something to base who you are or how successful you are. If so, then seek medical attention.
7. Where should I bet (and is it legal?)
I would say, going to Las Vegas and betting at the legal sports books there are the safest and legal way for an American citizen to wager on sports events in the United States. However, for most of us that is just not practical. We have to rely on internet sports-books and there are a number of pure scams out there. If you really want to place relatively obscene (> $10,000) amounts of money on the internet - You should not be reading this page, and you should be seeking help. Otherwise (you filthy rich bastard) do it in a legal and safe manner and go (fly/send your associate/move) to Las Vegas.
If the United States Government would get its head out of its rear end, it would legalize, regulate, and tax internet sports gambling. Apparently Great Britain (along with most of the Caribbean Islands) is smart enough to do so, and they have stolen (potentially) billions of taxable American dollars from this misconceived decision of the American Government. The hypocrisy of American Government to allow Horse Racing gambling, State lotteries, off track betting, and the proliferation of Indian and riverboat gambling casino's while proclaiming the evils of internet gambling is a reflection of the backward and non-progressive nature of the system - and an example of how technologically ignorant (and artificially moral) the people Americans elect to public office really are - and I do love the USA. My personal opinion is that if you are an addictive or underage gambler, you will find ways to gamble (bookies, high stakes card games, office pools, trips to casinos....) and no amount of finger pointing will stop that. The only way for the US Government to stop internet gambling it to 1. Shut down the internet or 2. Have the international community ban internet gambling outright, and either notion is sheer fantasy. By legalizing, regulating and monitoring betting on the internet, there is a possibility of identifying pathological gamblers (be it slight), and the dollars gained by taxing internet gambling could be used to educate and fund recovery clinics. More importantly, government regulation of internet gambling will provide hundreds of millions (to potentially Billions) of dollars to local, state and federal organizations. (Ok, enough ranting for now.)
I cannot give you my personal book choice since I am not sponsored, but there is a internet site reference below on how to choose a good sports book. I do utilize one of the top rated sites there, and (so far) I have had no problems with deposits or payouts (fingers crossed). Once again, I cannot vouch for the legal validity of all the sites listed there, and I take no responsibility for the information provided by that site, and YOU take a personal risk regarding utilization of these internet sports-books, and this page, or the author, does not condone gambling or betting on the NFL or other sports. But if you chose to do so, Sportsbook Review (independent watchdog) is a good start: http://www.sportsbookreview.com/default.aspx
I would recommend looking at the top rated books and making a comparison before making a choice. If you are currently using a book, refer to the blacklisted sites to see if you are in a good spot. If not, get the heck out NOW, and pray you can recoup some of your hard earned money! Be careful about bonus money though, because there is often a catch involved, and no internet book can be 100% reliable (site outages, possible bankruptcy....). I also recommend using more than one book because at times you will get better odds that way.
SBR WARNING 12/4/2005:
also, Read: How to spot a scam Book from SBR (pdf).
My ultimate recommendation is to test the book you want to use (with a small deposit, less than $150) and see how reliable the book is regarding deposits and payouts. See how good the book is regarding your questions and problems. If the book proves dependable, then up your deposits. However, I would not recommend putting outrageous amounts of money in offshore betting agencies (technically/legally this could be construed as money laundering), and I would limit the amounts to less than $1000 (disposable). With a big win, I usually withdraw the large amount won or funds over $1000 (or amounts greater than 30% of your starting bankroll = profit). If there are ANY problems with payouts, STOP all betting activities at this site IMMEDIATELY until you get the payoff you expect. After getting the payout, determine the viability of the book you use and decide to pull out or not. Unfortunately, if the book decides to take your money, there is little you can do about it except ask assistance from SBR or file a review with SBR and hopefully other players will just avoid that site (But please do so in a responsible manner, because I see a lot of frivolous and obvious scam reviews in there.). This is another reason why I wish the US Government would legalize American companies to provide a regulated internet sports book, security.
Also, there are ongoing misconceived efforts by the American Senate and Congress to limit the ways to deposit and withdraw funds from internet sports books, and you must stay aware those possibilities. Here is are some references:
http://www.ipmall.info/hosted_resources/crs/RS21487_041129.pdf
http://sportsgambling.about.com/od/legalfacts/a/Betting_Laws.htm
discussion: http://www.cato.org/dailys/06-13-02.html
Ok the final bottom line question the nitpickers out there want answered: Is it legal to place sports bets on the internet?
As of 2005, it is technically illegal to place bets using an off shore sports book for Americans doing so on American soil under the Federal Interstate Wire Act. (18 USC 1084)
Specifics: http://www.ipmall.info/hosted_resources/crs/RS21984_041129.pdf , http://gsulaw.gsu.edu/lawand/papers/su01/feldman/#federal_
On the other hand, there is a debate on the legalities: see: http://sportsillustrated.cnn.com/2003/writers/mike_fish/12/26/internet.gambling/,
and as of this date I have not heard of any private individual (who is of legal age, and doing so in a manner that does not violate the rights of other individuals) that places sports bets on the internet who has been arrested for this activity. Remember, It is technically illegal for you to operate your in-home low limit Friday night poker games; however, I don't see the FBI busting the millions of these low-limit home based US poker games since the FBI should have their hands full tracking down terrorists, drug pushers, murderers, and kidnappers. (unless some of these agents are in your poker game and losing big time - see http://news.yahoo.com/s/ap/20051121/ap_on_re_us/poker_clubs_losing_hand.)
Enough of the lecture, Good Luck!
f you are requesting a copy of the most recent excel sheet please email me a formal request so we can negotiate a reasonable price / donation (only experienced excel users please). Thank you.
Following week 10 of the 2005 season, I will not post the raw spreadsheet images due to time constraints. Use these references to get a feel of how the power rankings and predicted spreads and scores were generated. The current spreadsheets for this season have been adjusted and changed with increased game and seasonal experience, and are not representative of the sample spreadsheets presented here. Goto Index and Handicapping philosophies for more details.
Abbreviated raw excel spreadsheets for week 10 of the 2005 season click here
Raw excel spreadsheets for week 9 of the 2005 season click here
Raw excel spreadsheets for week 8 of the 2005 season click here
Raw excel spreadsheets for week 6 of the 2005 season click here
Raw excel spreadsheets for week 4 of the 2005 season click here