Weekend Football Review (9/13/11)
Tuesday, September 13 at 2:00 AM ET
I will always recap the weekend in football each week (usually on Mondays). It is very important to us to be transparent and honest about our picks. At this point, through Monday, the Paul's Picks, which include the top three ATS plays Saturday and Sunday as well as the top weekday college against-the-spread (ATS) plays and the Monday Night play (in this case, the top MNF ATS play since there are two, though one is unplayable and would not count in our record anyway), are 10-1 ATS (91%) to start the season (plus the Week 1 Michigan "Lock of the Week" that did not count in Vegas because it did not reach 55 minutes), the ATS Top Plays of the Day, the strongest opinion ATS each day overall in football 6-2 ATS (75%) (two weather-shortened games are removed from that record) and the "Locks of the Week" are 2-0-Weather. Including last season, this brings our all-time record during football to 108-63 ATS (63%) for Paul's Picks, 79-38 ATS (68%) in ATS Top Plays of the Day and 30-12 ATS (71%) for Locks of the week. We'll dive deeper into those records, performance with our other picks thus far, expectations, money management in this week's blog (we're trying to make this specific blog a more universal entry we can point people to for common questions, so I'll table game-by-game pick analysis and MLB discussions until next week). I've also included the Play Analyzer, Updating Plays and Previous Blogs blurbs from last week's blog because they are still at least as relevant and I know the full football season does not really kick off for many until the start of the NFL season, so I want be sure everyone catches these recommendations.
As a reminder, soon after the last game of each day, we make all of our subscriber content available for registered users. We're never going to hide anything. So even though we have to swap out articles in the archive to focus on new ones, articles never go away just make sure to use the correct week and date in the URL - or ask us for the link)...
We have received many questions regarding money management as it relates to our plays. Naturally, it is expected, and is especially notable due to the performance of our top plays in relation to the rest of our picks. It is a reality of this business that, unless we go undefeated or don't win a single play on a specific day or football week, some subscribers will likely win with our information and some of our subscribers will lose. While the losses may not really be anyone's fault more than the other, and we have been very fortunate to produce far more winning/happy/profitable weeks for our football subscribers, it is important to note some of the best ways to utilize the information.
Paul's Picks are 9-1 ATS (plus Michigan Week 1 - I will now try to remove this caveat from future discussion; you get the point) so far in the football season. However, one of those original picks from Sunday, Tennessee +3, closed yesterday at Tennessee -1.5. A Jacksonville win over Tennessee by two points, 16-14, gives the original pick a win despite a cover for Jacksonville on the closing line. This is a great example about the benefits of our technology and how best to use our resources to manage money.
The Paul's Picks have several advantages in that they highlight the top plays throughout the week in college and the NFL that have the highest chance of covering, adding extra analysis to clarify why the numbers appear as they do and what to look for in these teams (especially if there is any uncertainties with injuries). As we'll discuss next, three plays a day or so seems to be around the appropriate amount of games to play on a Saturday or Sunday based on our information and many subscribers appreciate and take comfort in the insight into the numbers for such significant plays - beyond just seeing the numbers. The Paul's Picks, along with the Locks of the Week and ATS Top Plays of the day are what we stake our football reputation on by promoting them and our performance with those picks over time. Fortunately, as noted above, those picks have performed very well in roughly 25 weeks of publishing football picks information on this site.
However, that does not necessarily mean that the side (and/or totals) from the Paul's Picks should be blindly played by everyone. Going forward, NFL and college football picks will be published at 8:00 pm ET on Wednesday nights. The lines that are used in those picks will be accurate at that time. Once we post the Paul's picks, we can and will not change them because maintaining the integrity of our record keeping is most important. (Note: While we do not change picks, in cases like we saw with Tennessee +3 where a major roster move occurred - the Jaguars cutting quarterback David Garrard - we do add a message to the Paul's Pick write-up acknowledging the change and linking to the updated Game Picks, Customizable Predictalator and Play Analyzer.) As soon as we post the picks, lines change and can already by different in specific books (we use real-time consensus lines from over 30 books). It would be impossible to track records for the real-time top three ATS plays at all times for every book after the picks are published until the games start. We have to stake our record on the original published picks. Every time the line moves though, even if it is just the juice, it affects the pick. Tennessee at +3 in Jacksonville covered 61.7% of the time. Tennessee at -1.5 covers 52% of the time. The play went from essentially a "2X" play at ~62% to a "no pick" on Sunday morning. At Tennessee -1.5, the Titans should not have been played.
We stick our flag in the ground on the Paul's Picks on Wednesdays at 8:00 pm ET. However, that doesn't mean that if someone cannot access those picks or those lines at that exact time that he/she is in the dark. The Customizable Predictalator and Play Analyzer were built to evaluate all picks at all lines. Specifically, with the Play Analyzer (which now includes the ability to print the table - as does every blog... see the bottom of this page), plays can be evaluated against the real-time, current consensus lines (and then tweaked to fit specific books). In some cases, like in the Baylor-TCU example (below in the more thorough explanation of the Play Analyzer) or with our Week 1 NFL Lock of the Week, Philadelphia -5 at St. Louis, which moved to -3.5 at one point Sunday morning (we'll try to Tweet when we see these type of games), line changes may turn a "normal" play as it was published into a "2X" or even "3X" play. It is very important to put our play information in the right context when it comes to lines. Even if there are only a couple plays that are being considered by a wagerer, if the line is different, they should be evaluated in the Customizable Predictalator or Play Analyzer.
Once the plays are loaded and re-run at new/consensus lines, determining a total amount to play becomes the next challenge. While there can be many motivating factors that lead people to take a position on games, the smartest way to do this from a wagering standpoint is to play the top plays at the current lines. How many? Unfortunately, there is not a perfect answer that covers every situation, but our rule of thumb based on what we have seen is to play somewhere between two and six individual or parlay plays in a day USING THE PLAY VALUE RECOMMENDATIONS. Technically, this depends on many things like the game times, when the bankroll is open and up-to-date from previous plays, our confidence in the group of plays in general and how that confidence varies amongst the top plays. Remember that every time a weaker play is added, it is not only less likely to win, but less should be wagered on it AND less should be wagered on all other plays at the same time because adding plays takes up more bankroll. Also remember that our confidence is exactly what it says. If five games are all exactly 60% to cover, our expectation is that we will go 3-2 in those plays.
To-date, 60%+ has proven to be a strong threshold that provides 1-6 plays a day, while also being as accurate as intended. It's not a steadfast rule that all or only 60%+ plays should be wagered on, but that range has seemed to perform very well. Unless it has occurred this year in the Play Analyzer with book-specific lines, I do not believe we have ever had more than three "green" (2X or 3X) plays on any given day (or at least start time), so playing those should be a no-brainer when they are available. It should go without saying, but "green" plays should be wagered on more than non-green plays. Furthermore, 63.1% "green" plays should be wagered on more than 62.0% plays. For optimal bankroll management, please follow play value recommendations. In a hypothetical situation where there are ten 70%+ confidence plays at one time, I am not telling you to choose seven and only play them. In that case, you should play significant amounts on all ten, expect to hit seven of them and clean-up. I'm just saying that, given a normal week, which includes 0-3 "green" picks per sport and usually around six 60%+ picks on Saturdays and Sundays, playing 1-6 plays a day based on real-time Play Analyzer and Play Value Calculator data is appropriate (the Play Value Calculator only goes up to 16 anyway. If we see a day when more than 16 plays should be wagered on at one time, we'll adjust and let everyone know.)
Resist the urge to play 50 games a week, especially without accurately applying the play value calculator. Why do we post every pick then? The most important reason for this is to be sure that we catch all line changes that may work in our favor and become strong plays (or vice versa - strong plays that become unplayable). Beyond that though, we understand that there are other reasons to wager on (or even just to know our data for) other games. If we are 53% confident that your alma mater is going to cover and you want to be in on the action (or if you want to do the opposite as a personal "hedge"), that's okay. If you want to roll 10-16 plays up into one parlay (also known as the "lottery ticket" to some) with big payout odds, that's cool too. Chilling out on a Monday night and want to liven up the Monday Night NFL game? We have the best answers we can come up with for every game. These may not always be "optimal" or very likely to succeed and it would not be considered our recommended best practices, but we get it. Our goal is to always put our subscribers in the best position to succeed and we work very hard at doing that. We only ask that expectations are managed relative to our confidence.
First and foremost on this topic, our confidence is our confidence and value is value. All comments regarding expectations throughout the season are relative to both our confidence and our accuracy. At no point do we ever expect that our performance will exceed (or fall short of) our projected confidence in each pick. That being said, having simulated football games professionally for seven previous seasons, I have seen trends relative to our projections. The most notable performance trend that we have is that, the more the public has an opinion, the better our projections (and higher our confidence) tend to be. We've seen this as obvious in the NFL playoffs and the first weekend in the NCAA Tournament, but it also plays a role throughout the football season. It makes total sense in that, the more public/steam we can exploit the better. The sports market is more exploitable because of the emotion and ignorance with which people base their wagers. With less intriguing matchups, the public stays away and we end up competing, to a good degree, against people with access to similar information. I feel as though we'll always have a technical advantage, but it's tougher to exploit the sharps (if you will - though that Philadelphia @ St. Louis lock felt good).
In college football, we do better (and our confidence is typically higher), especially ATS, earlier in the year than later. The public is so eager to wager on football that they jump in strong in college for a couple weeks until they get in trouble and focus elsewhere. In the NFL, whether it's Week 1, Week 9 or the Super Bowl, the public is all over every game. In that case, as one may expect, we consistently do well against the public, yet do improve as we learn more information on players and teams (and the public usually does not keep up as well) and interest gains towards the playoffs. So, if one were to graph our performance/confidence in college and the NFL, the lines would likely cross after the first few weeks of the season. College football picks typically start strong, then plateau (until the bowls, which I have generally had a lot of success with, yet we had a rougher than expected stretch last year after the "green" plays), whereas the NFL starts in decent shape and steadily climbs throughout the season. It is what is. That may not be exactly how things work this year (we'd love for the public to be strongly on every game so we can take advantage), but, following the blog information below and our/my pick history would suggest that this should be our expectation for the season.
That we have performed well ATS with our better plays and not as well with our O/U plays should be considered as much of a fluke as anything else. Save for some of the Play Analyzer examples noted below, our published O/U opinions have not typically been as strong as our ATS plays. And, for the O/U plays that have been high, yet not succeeded, we have run into a few "OVER" plays where the game has reached within a score of covering with more the eight minutes left before the teams stopped trying to score (Tulsa @ OU, FAU @ Michigan State, etc. - clearly we understood the pace and success that the team/s would have, but didn't quite predict whether or not any of the coaches would keep trying to score). Our O/U plays should be trusted as much as our ATS plays (relative to our confidence). We promote our ATS plays a little more for the simple fact that there is significantly more interest in sides than totals - which correlates to the degree to which they are played relative to totals.
The Football Numbers:
While I genuinely enjoy and find catharsis in the College and Pro best and worst of the weekend, that pick-by-pick analysis will take a back seat for this thorough and already long enough blog entry. To highlight the Paul's picks, we succeeded with our Locks of the Week, Alabama (-10 opening, -9 at kickoff), which won 27-11, and Philadelphia (-5 opening, moved to -3.5 early Sunday morning), for which the line moved to -3.5 on the consensus at one point Sunday morning and covered when the Eagles won 31-13. With the other Paul's Picks, we won with Oklahoma State (-14) over Arizona, 37-13, Washington State (-14) over UNLV, 59-7, in college and Tennessee (+3) at Jacksonville, 16-14, and Detroit (+1.5) over Tampa Bay, 27-20. The losing Paul's Pick this weekend occurred when San Diego State (-9.5) won over Army by just a score of 23-20. Aside from the huge swing in our favor with the NFL Lock of the Week, the other biggest Play Analyzer move ATS for the week occurred in college when Kentucky -12.5, which was a "weak" pick, to -10.5 over Central Michigan, which became one of the strongest sides of the college week as a normal+ pick, covering greater than 57% of the time. The Wildcats struggled out of the gate (for the second week in a row), but still won at home by a final score of 27-13 to cover the plays regardless of those lines.
Here are our win/loss stats from Week 2 in College Football and NFL Week 1 using playable picks (53%+ to cover) from our published articles (Play Analyzer/Customizable Predictalator plays appear to have provided even better results and stronger plays). We will update these at least until we have the TrendFinder, at which time the TrendFinder will likely supplement the blog posts:
- All Playable Games ATS: 53% (20-18)
- All Playable Games O/U: 49% (17-18)
- Paul's Picks ATS: 88% (7-1)
- Lock of the Week: 100% (2-0)
- ATS Top Plays of the Day: 80% (4-1)
- Picks that Cover 60%+: 70% (7-3)
- Normal+ Picks (57%+): 48% (14-15)
And here are the combined numbers for the season thus far:
- All Playable Games ATS: 55% (35-29)
- All Playable Games O/U: 54% (36-31)
- Paul's Picks ATS: 91% (10-1)
- Lock of the Week: 100% (2-0 and Michigan in Week 1)
- ATS Top Plays of the Day: 75% (6-2 and Michigan and West Virginia in Week 1... last time, I swear)
- Picks that Cover 60%+: 67% (8-4)
- Normal+ Picks (57%+): 52% (27-24)
These following sections were published in last week's blog, yet warrant posting in this "one-stop shop" blog for the season... The Play Analyzer is a new application that was on full display throughout the first weekend of the college football season. The tool allows you to import "consensus lines" - the most common side and total line for each game from over 30 books - and replay all games (you can look at a smaller, more specific subset if you prefer) in real-time. The plays are reordered and color-coded by confidence level and the play value recommendation is noted next to each pick (based on your normal play).
One great example of this came on the first Friday night of the college football season when the Baylor-TCU total line inexplicably dropped to 52.5 (52 in some places). The OVER (52.5), which had been originally published as a 59.4% play at a line of 56, became the top overall play of the weekend (and remained such based on our records for consensus line movements), nearing 67% confidence, which would be a "3X" play. The OVER covered with five minutes left in the second quarter.
They are not all going to work that way, but, win or lose is not necessarily the point here. Finding value is. And this example is great evidence of the power of the Play Analyzer. Without having to change any lines or replay any individual games, loading the data at the consensus lines quickly and easily illustrated the strength of that play as it rose to the top of the week's picks. It is highly recommended that subscribers check in periodically to quickly run the Play Analyzer to look for advantageous line movement - while following play value recommendations when moves are made. No matter when or how often plays can be made, this tool ensures the best information possible.
Of course, there can be lines that move for specific reasons. In the event of drastic line movements, a quick look into recent news for those games is advised. When important players get injured, cut, suspended or demoted (more on that later) or if the weather changes drastically from our expectations, we will update the picks as soon as we can. In these situations, books typically take plays off the board, re-evaluate and then put up new lines. We hope to update each team's information before the new lines are up, but that is not always possible. Take note of the "last updated" note at the top of the Play Analyzer and trust that we are making necessary changes if the data has not been updated since the breaking news. We will also send out Pick Availability Alert Updates when major changes are made. If you are not sure what is causing the line movement and/or if we are going to update the play, feel free to ask us and we'll let you know as soon as we can. You should feel comfortable trusting the Play Analyzer in helping you find the best value (like it can right now since the NFL Week 1 "Lock of the Week" has already moved a half a point in our favor, making it a "green" play at almost 63%).
To learn more about the Play Analyzer, check out our Football Video Tutorials.
Many of you have already seen that we updated the Tennessee @ Jacksonville play due to the Jaguars' release of quarterback David Garrard. When doing so, we updated the Predictalator Picks, Boxscore and Fantasy Projections. We also sent out a Pick Availability Alert email to note the updated play. While all player personnel moves can result in slight differences in our projections, we will update the picks, data, boxscores and fantasy information and send out emails when important information is learned that is different than our original expectation and has a direct impact on the value of the play. Typically, quarterback changes will evoke this update, as will mass suspensions, injuries to elite players and dramatic weather changes. We believe that those who subscribe to our information want the most accurate pick information possible and that those who subscribe to our emails regarding the availability of such picks would also like to know when they are updated, so it is important to provide these services.
However, since the performance of our picks is vital to our reputation and to our integrity, there are some things that we will not do. We will not ever change Paul's Picks or the Lock of the Week after publishing. Much of the foundation of our business is staked on these plays and many subscribers (rightfully) play these picks as soon as they are posted. We cannot change them after the fact. Fortunately, Locks and Paul's Picks rarely include games with uncertain situations at key positions. Also, as with our daily sports, we will not make any changes to the Predictalator Picks tables within two hours of the first game of a day (changes will never be without notice and only occur in necessary situations). Data for the Play Analyzer and Customizable Predictalator will be maintained at all times, updating as needed (and noted).
Previous blogs can be a great resource for tips on bankroll management and expectations. While I do not want to rehash everything we have ever discussed, to save you the time of hunting through my ramblings to find interesting information, below are some of the more valuable entries. The most important theme to the blogs and for getting the most out of this site: Value is value - value should be reflected in our confidence for each pick and enacted upon appropriately. Here are some useful blogs:
- Bankroll Management (9/7/10)
- Football Expectations and More Management (9/27/10)
- Parlays and Hedging (10/4/10)
- Setting Expectations (in Rough Weeks) (11/29/10)
- Finding the Holy Grail (and What is Simulation?) (1/17/11)
- By Sport, In-season Expectations (see Q&A) (1/24/11)
- Setting Expectations (in Great Weeks) (2/8/11)
- New Football Products (7/27/11)
- Video Tutorials (8/23/11)
As usual, if you have any of your own comments about this article or suggestions about how to improve the site, please do not hesitate to contact us at any time. We respond to every support contact as quickly as we can (usually within a few hours) and are very amenable to suggestions. I firmly believe that open communication with our customers and user feedback is the best way for us to grow and provide the types of products that will maximize the experience for all. Thank you in advance for your suggestions, comments and questions.