When To Buy The Half Point * Revisited

We put together an analysis of when it makes sense to buy the half point in football.

The sample data set consists of the last five years of history for College Football and the NFL.

We selected all games where the Closing Line fell on a “football number”, then recalculated the results based on if we would have bought the Line down when playing a Favorite, or bought the Line up when playing an Underdog.

The data is presented in both percentages and in net dollar figures based on simulated $100 wagers.

We used the assumption of 110 to win 100 for the actual data, and 120 to win 100 for the buy down/up data. If you wish to use different odds for the buy down/up data you can simply download the spreadsheet and apply whatever variables you see fit.

The “Net Advantage” column in the spreadsheet shows the plus or minus net that would have occurred from buying the Line up or down at each data point.

Some of the findings are obvious, and some of them not so obvious.

Getting The Most Out Of Our Website – How To Use The “Select/Sort and Print Games” Function

We had always found it frustrating, that while there were countless sites that provided handicapping information, there was never a simple way to print a roster of the current days games from them.

You could try to copy and paste from somewhere, but that always screwed up the formats, and for the most part, unless you possessed a lot of Excel expertise, you were stuck with whatever Cntl-V left you with.

No matter what your handicapping regiment is, pretty much everyone wants or needs a list of the games and we have helped to make that a lot easier to get.

We created functionality that gives our users the ability to Select, Sort, and Print games, that we believe is as unique and universally useful a handicapping tool as you can find anywhere on the web.

Below you will find an easy to understand primer on how to use this section of our site.

First of all, to launch the “Select/Sort/Print Tool”

1. Go to any of the League Drop Downs on our site
2. Mouse Over “Current Games”
3. Mouse Over and choose “Select and Print Games”

The initial screen offers a top -line view of the complete current card of games.

Some of the data elements include:

• Date and Time
• Rotation Number
• Opening and Current ML/Line/Total Information
• Our Projected Scores and Projected ML/ATS/Total Results
• Notes Section updated with comments on any sides/totals that we may like

To Filter the display:

1. Select any combination of the following variables-

• Current Line – Select a specific Current Line Range
• Current Total – Select a specific Current Total Range
• Date – Select a specific date or date range
• Time – Use the buttons bar to select a specific range of game start times

2. Then click the “Filter Schedule” button and your results will be displayed.
3. Clicking the “Clear Filters” button will reload all games for display.

At any time you can sort your displayed results by clicking any of the report headings including:

• Rotation Number
• Date
• Time
• Current Line
• Current Total
• Projected Outcome
• Team Name

To print your current selection:

NFL – An Historic Analysis Of Q3 ML/ATS/Total Performance By Team

We posted data through Q2 about a month ago, and thought it was worth continuing into Q3.

You can find the original post here.

To recap, we have tracked  how teams have done comparatively across different time frames with in a season.

The time-frames are defined as follows for NFL:

• Weeks 01-04 are Q1
• Weeks 05-08 are Q2
• Weeks 09-12 are Q3
• Weeks 13-17 are Q4

The analysis of the NFL Q3 history indicates there are several teams whose ML/ATS/Total performances have been measurably different as compared to other time frames.

For example, the Saints have historically been a very strong Q3 team ATS covering 64.7%, 17.3% over their historic Q2 performance and 12.3% over their historic baseline performance.

We say this all the time,  but we want to be absolutely clear that any of these trends in a vacuum are no reason in and of itself to play a side or total.  This data should be used as just one more piece of information in your handicapping.

Some of this we said in our original post but it bears repeating:

• There has been a noteworthy consistency to this data for several teams, and we believe it is worth your while to familiarize yourself with the information.
• There are a lot of arguments that can be made as to why these trends occur for some teams from quarter to quarter over time.  Coaching philosophies, scheduling, QB’s that start slow, and odds-makers’ adjustments all play apart in the fluctuations we see between time-frames for some of the teams.
• When using this reference it is important that you keep in mind situations such as coaching or impact roster changes that would render historic data much less useful.  Obviously for teams that have enjoyed coaching and player continuity, the data should tend to hold more true.
•  The baseline represents all results for a given team across all time-frames.

In the embedded Excel workbook below, you will find the following worksheets for both the NFL and NCAAF:

• ML_HISTORY_Q1-Q4 – ML WIN % BY TEAM
• ATS_HISTORY_Q1-Q4 – ATS WIN % BY TEAM
• TOTAL_HISTORY_Q1-Q4 – TOTAL OVER WIN % BY TEAM
• SUMMARY_HISTORY_Q1-Q3 – ML/ATS/TOTAL COUNTS BY FAV/OVER FOR Q1, Q2, Q3, AND A BASELINE PERIODS

Please note the following regarding the data:

• THE DATA REPRESENTS A 5 YEAR HISTORY AND IS GROUPED BY QUARTER
• ML AND ATS PERCENTS ARE IN TERMS OR WIN %
• TOTAL PERCENTS ARE IN TERMS OF PERCENT OF GAMES THAT WENT OVER
• THE BASELINE PERIOD REPRESENTS ALL DATA FOR ALL TIME-FRAMES BY TEAM
• For Q3 any Historic percentages >= 60% are highlighted in green
• For Q3, any Historic – Baseline differences of >= 10% are highlighted in green and any <=-10% in red
• For Q3 as compared to Q2, any differences of >= 20% are highlighted in green and <=20% in red
• Under the column “Best Quarter” is Q3 it is highlighted in green, and if “Worst Quarter” highlighted in red
• ALL PUSHES HAVE BEEN OMITTED

 

NFL DATA

*

We hope you find the information an asset to your handicapping and please visit our free site, GreyMatterStats, where we put this information at your fingertips.

If you find something of interest that you would like to share, please feel free to leave a comment or Tweet us  @GreyMatterStats

College Football – An Historic Analysis Of Q3 ML/ATS/Total Performance By Team

We posted data through Q2 about a month ago, and thought it was worth continuing into Q3.

You can find the original post here.

To recap, we have tracked  how teams have done comparatively across different time frames with in a season.

The time-frames are defined as follows for NCAAF:

• Weeks 01-04 are Q1
• Weeks 05-08 are Q2
• Weeks 09-12 are Q3
• Weeks 13-Through the Bowl Games are Q4

The analysis of the Q3 history in College Football indicates there are several teams whose ML/ATS/Total performances have been measurably different as compared to other time frames.

For example, Ole Miss has historically been a very strong Q2 team ATS covering 66.7%, but for Q3 their ATS cover rate drops dramatically to just 37.5%.  The current Rebel team has held true to this trend and are undefeated Q2 season-to-date Q2 ATS. We will be using this information to look to make a case to fade Ole Miss ATS as part of our overall handicapping process.

We say this all the time,  but we want to be absolutely clear that any of these trends in a vacuum are no reason in and of itself to play a side or total.  This data should be used as just one more piece of information in your handicapping.

Some of this we said in our original post but it bears repeating:

• There has been a noteworthy consistency to this data for several teams, and we believe it is worth your while to familiarize yourself with the information.
• There are a lot of arguments that can be made as to why these trends occur for some teams from quarter to quarter over time.  Coaching philosophies, scheduling, QB’s that start slow, and odds-makers’ adjustments all play apart in the fluctuations we see between time-frames for some of the teams.
• When using this reference it is important that you keep in mind situations such as coaching or impact roster changes that would render historic data much less useful.  Obviously for teams that have enjoyed coaching and player continuity, the data should tend to hold more true.
•  The baseline represents all results for a given team across all time-frames.

In the embedded Excel workbook below, you will find the following worksheets for both the NFL and NCAAF:

• ML_HISTORY_Q1-Q4 – ML WIN % BY TEAM
• ATS_HISTORY_Q1-Q4 – ATS WIN % BY TEAM
• TOTAL_HISTORY_Q1-Q4 – TOTAL OVER WIN % BY TEAM
• SUMMARY_HISTORY_Q1-Q3 – ML/ATS/TOTAL COUNTS BY FAV/OVER FOR Q1, Q2, Q3, AND A BASELINE PERIODS

Please note the following regarding the data:

• THE DATA REPRESENTS A 5 YEAR HISTORY AND IS GROUPED BY QUARTER
• ML AND ATS PERCENTS ARE IN TERMS OR WIN %
• TOTAL PERCENTS ARE IN TERMS OF PERCENT OF GAMES THAT WENT OVER
• THE BASELINE PERIOD REPRESENTS ALL DATA FOR ALL TIME-FRAMES BY TEAM
• For Q3 any Historic percentages >= 60% are highlighted in green
• For Q3, any Historic – Baseline differences of >= 10% are highlighted in green and any <=-10% in red
• For Q3 as compared to Q2, any differences of >= 20% are highlighted in green and <=20% in red
• Under the column “Best Quarter” is Q3 it is highlighted in green, and if “Worst Quarter” highlighted in red
• ALL PUSHES HAVE BEEN OMITTED

 

COLLEGE FOOTBALL DATA

*

We hope you find the information an asset to your handicapping and please visit our free site, GreyMatterStats, where we put this information at your fingertips.

If you find something of interest that you would like to share, please feel free to leave a comment or Tweet us  @GreyMatterStats

Football Is At The Quarter Mile Post And History Suggests That Some Teams Perform Measurably Different In The Next Leg Of Their Seasons

If you’re like us, whenever football is not being played you can’t wait for it to start, then once it does the weeks just fly by.

It’s hard to believe that we are already one quarter of the way into the NFL season, and one quarter plus into the College Football season.

One of the things we have tracked for some time but never reported on, is an analysis of how teams have done comparatively across different time frames with in a season.

For football, we have found that by taking the regular season and grouping it into quarters we have gotten that best results overtime when it comes to this type of comparative analysis.

The time-frames are defined as follows:

• For both the NFL and NCAAF – Weeks 01-04 are Q1
• For both the NFL and NCAAF – Weeks 05-08 are Q2
• For both the NFL and NCAAF – Weeks 09-12 are Q3
• Q4 for the is NFL Weeks 13-17, and for NCAAF Weeks 13-Through the Bowl Games

Everyone reading this is probably well aware that at the end of any football season, there is a fairly even distribution of Favorites and Dogs that cover ATS.  But you will never see an even distribution of Favorites and Dogs from week to week.  A few weeks might be top heavy with Favorites followed by a run of Dogs, that at the end of the season results in relative ATS parity.

The same basic philosophy can be applied when analyzing result data at the individual team level across various time-frames.

For example, the Florida Gators historically have been overall a very strong 60.6% ATS in the past 5 years.  But when we review how the Gators have done across our time-frames we see an interesting and consistent trend.  For Q1 the Gators have been an off the charts 76.2% ATS, but for Q2 their ATS win rate drops to only 33.3%.  Currently the Gators are 3-1, or 75.0%, ATS and host LSU as a 2.5 point Dog in this Q2 matchup.

In another example, the Minnesota Vikings have gone Over the Total 52.4% of the time for the past five years of history.  But for Q1 only across that five year history the Vikings Total has gone Over only 39.1% of the time, while for Q2 they have been Over 72.2% of the time.  This week Minny, who is 1 Over/3 Under on the year, hosts the Titans with the Total currently at 44.

We want to be absolutely clear, that any of these trends in a vacuum are no reason in and of itself to play a side or total.  This data should be used as just one more piece of information in your handicapping.

That said, there has been a noteworthy consistency to this data for several teams, and we believe it is very much worth your while to familiarize yourself with the information.

There are a lot of arguments that can be made as to why these trends occur for some teams from quarter to quarter over time.  Coaching philosophies, scheduling, QB’s that start slow, and odds-makers’ adjustments all play apart in the fluctuations we see between time-frames for some of the teams.

When using this reference it is important that you keep in mind situations such as coaching or impact roster changes that would render historic data much less useful.  Obviously for teams that have enjoyed coaching and player continuity, the data should tend to hold more true.

The screenshots and spreadsheet show historic Q1/Q2 and baseline data for the NFL and College Football.  The baseline represents all results for a given team across all time-frames. We will revisit this post at the end of Q2 and see how the data fared versus the actual results, as well as add data for Q3.

In the embedded Excel workbook below, you will find the following worksheets for both the NFL and NCAAF:

• ML_HISTORY_Q1-Q2 – ML WIN % BY TEAM FOR BASELINE/Q2/Q1 – SORTED BY ALPHA AND BY Q2 WIN % AND BY Q2 DIFFERENCE FROM BASELINE AND BY Q2 DIFFERENCE FROM Q1
• ATS_HISTORY_Q1-Q2 – ATS WIN % BY TEAM FOR BASELINE/Q2/Q1 – SORTED BY ALPHA AND BY Q2 WIN % AND BY Q2 DIFFERENCE FROM BASELINE AND BY Q2 DIFFERENCE FROM Q1
• TOTAL_HISTORY_Q1-Q2 – TOTAL OVER WIN % BY TEAM FOR BASELINE/Q2/Q1 – SORTED BY ALPHA AND BY Q2 WIN % AND BY Q2 DIFFERENCE FROM BASELINE AND BY Q2 DIFFERENCE FROM Q1
• SUMMARY_HISTORY_Q1-Q2 – ML/ATS/TOTAL COUNTS BY FAV/OVER FOR Q1, Q2, AND A BASELINE PERIODA

Please note the following regarding the data:

• THE DATA REPRESENTS A 5 YEAR HISTORY AND IS GROUPED BY QUARTER
• ML AND ATS PERCENTS ARE IN TERMS OR WIN %
• TOTAL PERCENTS ARE IN TERMS OF PERCENT OF GAMES THAT WENT OVER
• THE BASELINE PERIOD REPRESENTS ALL DATA FOR ALL TIME-FRAMES BY TEAM
• ALL PUSHES HAVE BEEN OMITTED

NFL DATA

COLLEGE FOOTBALL DATA

We hope you find the information an asset to your handicapping and please visit our free site, GreyMatterStats, where we put this information at your fingertips.

If you find something of interest that you would like to share, please feel free to leave a comment or Tweet us  @GreyMatterStats

A Public Service Announcement On The Evils Of Betting Money Line Favorites

The following is a Public Service Announcement for those that are unaware, and a reminder for those that are aware, of the Evils Of Betting Money Line Favorites.

What we write about below is most applicable to College Basketball and Football, but the lessons should be applied to all your sports handicapping.

*

Those of you familiar with our site know that we project the outcome of every game played and track those projected results as simulated $100 wagers against the Money Line, ATS, and for the Total.

We do this for many reason, but most importantly to provide an easy to understand barometer as to how well our projection methodology is working overall and how it would have performed in any of the “what-if” queries that can be submitted through our site.

So last night we were poking around looking for some trends for the current season of College Football., and we happened to notice that the dollar-ized output for our season to date Projected Money Line results was –$16,265.

And we were like, “Ummmmm, do we have a bug somewhere…it’s only Week 05 that can’t be right, can it?”

So we started to double check everything and uncovered what you will always uncover when you bet Money Line Favorites; Sooner or later you are going to get eaten alive by a huge Live Dog.

Please note that all of the following are based on straight up Money Line wins for the 2012/2013 College Football Season.

Season to date there are been 232 games that we have forecast.

Of the 232 games, we had projected the Favorite to win 195 times and were correct 151 times, or 77.4%, a respectable figure considering that we include every single game played.

But a funny thing happens on the way to the bank with this simulation, and that is we’d have nothing to deposit since we would be $15,323 in the hole, even after winning 77.4%.

Three of the Favorites we had selected straight up were Houston –9000 over Texas State, Arkansas –8000 over UL-Monroe, and Georgia Tech –3000 over Middle Tennessee State.

All three lost outright, and those three games alone contributed $20,000 is losses in our simulation.

Our 151 wins would have gained $15,100 and our 44 losses would have resulted in a staggering loss of $30,423.

Three out of 195, or 1.5% of the games, accounted for 65.7% of our simulated monetized losses.

To further put this in perspective, just to get even from a deficit of -$15,323, betting $X to win $100 you would need to win 154 games in a row.

Once in awhile we come across a thread in a forum or a Twitter question where someone asks, “Hey wouldn’t it make sense to take the Money Line with these big Favorites and not worry about them covering a huge number ATS?”

The answer is, NO NO NO NO NO.

If you don’t want to lay the points in these types of games at something much closer to $110 to win $100, THEN PICK ANOTHER GAME, OR MAKE NO PLAYS AT ALL OR BOX A TRIFECTA AT SOME TRACK.

In our opinion for football/basketball in both college and pro, you should never be playing Money Line Favorites, and never ever ever ever playing huge Money Line Favorites.

Now if you are taking Money Line Dogs, we are all for that and if you have a good please pass it along.

This Public Service Announcement has been brought to you by GreyMatterStats, and we have approved this message.

We now return you to your degenerate gambling habits.

@GreyMatterStats

Twitter Question – Using Opening Lines/Totals Instead Of Closing Lines/Totals, How Much Variance Would We See In The Results For NCAAF and The NFL?

We got a question via Twitter from @SkatingTripods, a friend of our site from @BTBFourm, who asked how much variance we would see for football ATS results if the outcomes were calculated using Opening Lines instead of the Closing Lines.

Just a quick FYI, @SkatingTripods and @lindetrain, AKA Adam Burke and Matt Lindeman, host a weekly podcast focusing on sports handicapping.   The discussion is never about 100 Star Locks of the Week and instead offers intelligent debate and useful information.  They have two shows under their belt as of this writing and I think they have done great job, so if you get a chance, please check them out.

Some miscellaneous notes and observations:

• The data is presented for the Previous 2011/2012 Season as well as  All History, which represents all data going back to the 2007/2008 season
• Analysis is presented for NCAAF and the NFL by ATS and Total
• For both ATS and Totals we have calculated what would have been the outcomes using the Opening Line/Totals and compared that against the actual data
• Specific to ATS, there is a somewhat confusing situation that occurs when the Favorite Flipped from the Opening Line to the Closing Line.  In this scenario for example, what was an Opening Home Favorite would be a Closing Home Dog.  So for some of the line items you will see Closing Result Favorite/Open Result Dog – NO CHANGE
• In my initial exchange with @SkatingTripodsI told him my off the cuff guess is that there would be minor variance in NCAAF and the NFL a bit more, but still not too significant.  The finding were interesting to me in that more games were affected than I would have anticipated.
  • For NCAAF ATS All History showed that 4.5%, and for the Previous Season 5.2%, of the games would have changed
  • For NCAAF Totals All History showed that 4.4%, and for the Previous Season 2.8%, of the games would have changed
  • For NFL ATS All History showed that 4.4%, and for the Previous Season 5.2%, of the games would have changed
  • For NFL Totals All History showed that 5.2%, and for the Previous Season 5.6%, of the games would have changed

• It is important to note that for the percentages differences listed above, it does not necessarily mean that you would go have gone from a loser to a winner or a winner to a loser.  In many cases a Push comes into play where you either had a Push based on the Closing number and an outcome based on the Opening number, or an outcome based on the Closing number and a Push based on the Opening number.  The percentage differences simply indicate that there was some sort of change between the two calculations.
• The right most column in the spreadsheets indicates the percentage difference from the Previous Season as compared to All History for each line item.

NCAAF ATS OPEN VS CLOSE VARIANCE

NCAAF TOTAL OPEN VS CLOSE VARIANCE

NFL ATS OPEN VS CLOSE VARIANCE

NFL TOTAL ATS OPEN VS CLOSE VARIANCE

In the embedded Excel workbook below, you will find the following worksheets:

Twitter Question – In College Football, How Do Ranked Road Dogs Do When Facing Unranked Home Favorites?

We got a question via Twitter from @EltLearn33 who asked, in college football, how do ranked road dogs do when facing unranked home favorites?

To answer this question we used five years of our NCAAF history which included the 2007/2008, 2008/2009, 2009/2010, 2010/2011, and the 2011/2012 seasons.  It is important to note that the rankings we use are our own internal calculations that may or may not , sync up with nationally published ranking data for the given time frames.

The way the information lays out makes it impractical to create a screen capture, so please visit the embedded spreadsheet below for the full breakdown of the data by Season, by Week, and by Conference Match-Up.

Some miscellaneous notes and observations:

• The selection criteria only yielded 64 games which is a very small sample size for analysis purposes, so please keep this in mind as you review the findings
• In 7 of the 64 games the closing line was a pick. Within our system, when a game is a pick the home team is our listed favorite
• In cases such as the PAC-12, which was known as the PAC-10 during part of our history, we use the current name across all time frames
• When analyzing the ML and ATS data by Season no note worthy observations were made other than that for the 2010/2011 Season Ranked Away Dogs were 7-1 ATS
• For all games in all Seasons the Under has been 58.7% in the 64 games in this role
• When analyzing the data by Week no note worthy observations were made
• When analyzing the data by Conference Match-Ups Ranked Away Dogs playing within their Conference from the BIG-10, BIG-12, and the SEC were found to have success both ATS and straight up
• Games played between BIG-10 teams have gone Under the Total 7 times with only 1 Over and a Push
• Games played between BIG-12 teams have gone Under the Total 6 times with only 2 Overs
• Games played between SEC teams have gone Over the Total 8 times with only 2 Unders
• Points do not typically come into play in these games; Excluding 3 ATS pushes in only 5 of 61 games did points enter in to the ATS result

Twitter Question – Do Tired Arms Lead To More Overs As We Approach The All-Star Break?

We got a question via Twitter from @JoeDellAnno who asked, how do MLB totals fare the week before the All-Star Game with the theory being that the bullpens are tired?

To answer this question we went to our MLB history for 2009, 2010, 2011, and season to date for 2012 and extracted data in three groups.

1. Ten Days Prior To The All-Star Break
2. Last Series Before The All-Star Break
3. First Half Of  An MLB Season – Used as a baseline to compare the findings from the other time-frames

In addition to analyzing the totals, we also took a look to see if there was anything of interest regarding money line or run line data for the selected time-frames.

The screenshot below presents the overall findings, which can also be found in the spreadsheet embedded below.

 

Twitter Question – Does The Wind Blowing In Or Out At Wrigley Materially Affect The Total Result?

We got a question from @JeffFogle who asked  us if we had any data to indicate what impact  the wind blowing in or out at Wrigley has had on the Total Result there.

The first thing we told Jeff was that we did not carry any weather information in our archives. He suggested that if we could provide outcome data where the total was >=11, which would indicate a wind blowing out day, and where the total <= 7.5, which would indicate a wind blowing in day,  that it would suffice for his purposes.

So below you will find two screen captures that show the total results data at Wrigley field for the 2009, 2010, 2011, and STD for 2012.  The first screen capture shows the Wrigley total  results compared to all other stadiums in MLB which we used as a baseline.  The second screen capture shows the deviation of the Wrigley total results for totals of 11 or higher and 7.5 and lower versus the mid-range of 8.0-10.5.

Some miscellaneous notes and observations: