NO GARBAGE STATS

There are lies, damn lies, and then there are “statistics”

What is a Garbage Statistic?

While there may be disagreement on which statistics are garbage, it is generally agreed that garbage stats are those which are of no assistance in predicting the outcome of games. When the statistics are remote in time, relate to a single team or single matchup, cover different time frames and involve statistically insignificant numbers, they are unreliable. The fatal flaw in most statistics employed by handicappers is that either the statistic or trend is based on too few events to be statistically significant or in order to amass enough events to be of statistical significance, many of the events are so remote in time that they have no bearing on the game at hand.

To illustrate—if team A covered against team B in both their meetings last year and their first meeting this year, does that mean they will cover in the second game this year? 3-0 is not statistically significant. Would you conclude that a flipped coin will always land on heads based upon 3 tosses that came up heads? Of course not. On the other hand, suppose team A and B are divisional rivals that have played each other twice a year for 30 years, so that there are 60 events in the statistical analysis. Do you think what happened 10, 20 or 30 years ago when the only similarities between those teams and the ones playing today are the colors of their jerseys and designs on their helmets ( if that ) matter?

The single cornerstone statistic we utilize has held true across the board for all teams, each and every year for thirty-six consecutive years (over 9000 games). It is statistically significant based upon 9000 events and is unaffected by remoteness, because it does not focus on matchups between individual teams, but applies to all games, every year, regardless of the teams involved.

Consider the following examples:

EXAMPLE 1 (Hypothetical)

It is week 11 . Team A is favored on the road against a Division Rival, is Service A’s 10 star pick of the week and boasts the following Stats:

  • 4-0 Straight Up (SU) and Against The Spread (ATS) in its last 4 games

  • 7-2 ATS in its last 9 visits to this Divisional Rival

  • 7-3 ATS Divisional Road Favorite in November since 1985

Team B is coming off a bye and at home against a Division Rival, is Service B’s Upset of the year and boasts the following Stats:

  • 12-5 ATS coming off a Bye over the past 17 years

  • 18-4 ATS all time in first of 3 straight home games.

  • 4-0 ATS in its last 4 meetings with this week’s opponent

All of these Stats sound impressive initially and a bettor receiving this information might want to play on teams A and B in their respective games. A review of these Stats, shows why they are garbage and should not be considered.

Team A is 4-0 SU and ATS in its last 4 games. Each game is an independent event, much like a coin flip, so that no matter how many heads came up in a row, the odds on the next flip are still 50/50. In fact, the line is likely inflated for any team that has just covered 4 straight, making it even harder to win the following week.

Team A is 7-2 ATS in its last 9 visits to this Divisional Rival. What happened 9 years ago between these 2 teams, when most likely each team had a different coach and hardly any of the same players is of no impact on how the current teams will fare against each other and the spread this week.

Team A is 7-3 ATS Divisional Road Favorite in November since 1985. See above. If what happened 9 years ago between 2 teams is meaningless, what does that say about something that happened to 1 team 35 years ago?

Team B is 12-5 ATS coming off a Bye over the past 17 years. In addition to rationale stated above, you don’t know the sequence of the record. Did they win 12 straight off a bye and then lose the last 5, or vice versa? The answer is it doesn’t make a difference, it’s a garbage stat anyway. Covering the spread is a matter of exceeding expectations in a particular contest and the higher the expectations, the more difficult it becomes to exceed them and cover the bet.

Team B is 18-4 ATS all time in first of 3 straight home games. See above. By now, you should see the worthlessness of this stat on your own.

Team B is 4-0 ATS in its last 4 meetings with this week’s opponent. Well, I saved the best for last. Here is an apparent indication that over the past 2 years, when the rosters and coaches were largely the same as now, one team seems to have the other’s number. Once again, we don’t know what the expectations were that this team exceeded. But more importantly, this gives me an opportunity to point out how statistics can lie. Although Team B is 4-0 ATS in its last 4 games against this opponent, it is only 2-7 at home ATS over the past 9 years against this opponent. THAT’S RIGHT —- TEAM A AND TEAM B ARE PLAYING EACH OTHER. Garbage stats are not necessarily mutually exclusive—Team A is 7-2 ATS in its last 9 visits against Team B (they were 7-0 until losing to the spread the past 2 years). Combining these past 2 years losses ATS on the road by Team A, with 2 losses ATS at home by Team A yields Team B’s record of 4-0 ATS in the last 4 games between these Division Rivals.

EXAMPLE 2 (Actual)

It is Week 3 of the 2009 season. Baltimore is 13.5 point Home Favorite over Cleveland. Cleveland lost 27-6 to Denver the previous week and has Cincinnati on deck. Baltimore upset San Diego 31-26 the previous week and has New England on deck.

Consider:

  • Cleveland is 8-1 ATS off a double digit SU loss in Sept

  • Cleveland Head Coach (Mangini) is 5-0 ATS as an underdog off a double digit ATS loss

  • Cleveland Head Coach (Mangini) is 8-2 ATS away v. the Division

  • Baltimore is 1-4 ATS v. conference opponent off a double digit SU loss

Looks like the stats really favor Cleveland; however…………………

  • Cleveland is 1-8 ATS before facing the Bengals

  • Cleveland is 1-4 ATS after scoring 7 or less points

  • Baltimore is 8-1 ATS home off a SU underdog win

  • Baltimore is 8-2 ATS as a double digit home favorite v. an opponent off a SU loss

So what do you do?

(A) reconcile these stats and choose which are true indicators and which are garbage

(B) skip this game and look for one where all the stats favor the same side

(C) ignore all these stats as garbage and look for a more reliable method to select games.

If you chose (A) and selected Cleveland, give yourself a loss. If you chose (A) and selected Baltimore you either cheated by looking at the result, got lucky, or are smarter than I and don’t need my help. If you chose (B) you are missing the point, and by the way—good luck finding such a game. If you chose ( C) the GREEN MACHINE is for you.