Here's a follow up to yesterday's thoughts on the underperformance of dogs in the NFL......
SPORTS HANDICAPPING REPORT
Thursday, September 25, 2003
NEW: SO, WHAT HAPPENED BY YEAR'S END THE LAST TIME (2001)?
Note: You may want to read the first part of this essay -- "THE DOGS ARE KILLING US." Otherwise, these statistics won't make a lot of sense.
In the previous year where the favorites started out hot, the trend continued for the first four weeks of the season . Then, things rapidly reversed themselves.
As I mentioned previously, 2001 was a fluke year where the favorites were winning way beyond normal expectation in the early part of the year. Coincidentally, that was the worst period I've ever had at MadJacks (other than the slump I'm now in).
Here are the results (W-L record of favorites in parenthesis):
2001 -- Week 1 (7-6)
2001 -- Week 2 (8-5)
2001 -- Week 3 (9-4)
2001 -- Week 4 (9-5)
So, the first four weeks of 2001 produced a record of 33-20, which is about 64 percent winners! Does this sound familiar?
Now, let's look at the next few weeks of that season:
2001 -- Week 5 (2-9)
2001 -- Week 6 (1-14)
That's right. Amazingly, the favorites all crashed and burned the next two weeks. They went a dismal 3-23 in Weeks 5 and 6. It's difficult to know if the linesmakers adjusted the numbers to be more generous to the dogs. I suspect that things just evened out eventually and we saw a normal return to the mean. However, if you go back and compare Week 4 -- versus 5 and 6, we note that many of the lines were around three points. In weeks 5 and 6, many of the lines were up around 6 points. I realize this is subjective reasoning, but I do want to report that the lines INCREASED which might be a factor as to why the dogs suddenly went on a solid run.
By season's end in 2001, the dogs ended up winning 60 percent in the final 13 weeks of the season. In fact, looking at all the data, dogs have NEVER performed worse than 49.75 percent in the last twenty years of line tracking.
What this means is -- dogs are very likely to return to form in the next week or two. Whether or not you want to wait and let Week 4 play out (assuming there's some correlation to 2001) is up to you.
THE DOGS ARE KILLING US!
I had an interesting telephone conversation yesterday with a fellow handicapper who runs a computer program based purely on statistics. The system he uses has never had a losing NFL season in the six years he's worked on it. The system usually produces about 60-70 units profit per year -- using a spread of 1-5 units per game per week. That might not sound like much, but it's proven to be a relatively "risk free" way of producing a profit. I've tracked his system, so I know his numbers are accurate.
What interested me in his system is that it plays dogs mostly -- with assorted weighing factors.
The results of the 2003 NFL season so far have been a disaster for dog players. I wanted to see if my disappointing results are shared by anyone else out there. I was particularly interested hear the feedback from this "dog" handicapper, who plays dogs for statistical-based reasons (I'm oversimplifying his methods to make this easier to understand).
First, let's establish the fact that dogs have performed horribly this season. In all my years as a handicapper, I do not recall favorites ever covering at such a high percentage through three weeks of the season.
Here's the proof. In the early games last Sunday, the favorites went 7-0! That's right -- a perfect 7-0! They faded in the later games, but still ended up going 9-4-1 for the week. That brings the favorites to 28-14-2 for the season, to date. Favorites have covered in 66 percent of NFL games! Amazing!
Why?
An interesting point was raised about possible reasons why the favorites are blistering the board. Here's one theory. My colleague mentioned that the dogs do not seem to be getting as many points, this year -- as before. His theory is that oddsmakers got hammered with dogs last season, and have adjusted the numbers towards parity. In other words, the dogs are currently not getting as many points as in previous years. Our discussion continued. Using a couple of games from last week as examples -- Pittsburgh probably should have been favored by greater than 4.5 points versus dismal Cincinnati (based on the perceptions of these teams). Pittsburgh was a -6 road favorite at Cincy last season -- why just -4.5 this year? And, Super Bowl champion Tampa probably should have given more points to "Vickless" Atlanta. So, there is some justification for this idea.
I decided to go back and compare all the numbers from this season versus last season. For both seasons, there were exactly 46 games played through the first three weeks.
In 2002, the average dog was getting 5.1 points. That means the average spread of all NFL games was -5.1 Contrast this with 2003. The average dog is now getting just 4.5 points. That's a drop of about a half a point per game. Looked at another way, dogs have lost about 12 percent of value based on the "shift" towards favorites by linesmakers.
Of course, we must take into account that the game matchups are different from season to season. (Perceived) Lopsided games produce higher lines than close divisional matchups. Nevertheless, I believe there has been a conspicuous decision on the part of oddsmakers to drive down the numbers and to reduce the value of dogs.
SIDE NOTE AND AN APPARENT CONTRADICTION: Several sportsbooks are getting hammered in the NFL (I know this for a fact). The next question is -- why would the sportsbooks reduce the line value of dogs if they end up losing money? Of course, not even the sportsbooks could have predicted these astounding results. They were so afraid of the big money coming down on the dogs again this season that they intentionally shifted the line values towards the favorites, knowing most of the "sharps" are dog bettors. The had no way of knowing that the lines in most cases wouldn't matter and the favorites would win at such a high rate. So, the public is cleaning up -- at least temporarily.
ANOTHER SIDE NOTE: Most Las Vegas sportsbooks do not post independent lines. They get their numbers from other sources. I am not suggesting that each sportsbook has intentionally made a decision to move the lines towards the favorites. I do believe this decision was made by some of the key offshores and linesmaking services and this has filtered into the Las Vegas sportsbooks. If you walk into any Vegas sportsbook this month, you've seen the $50 bettors in ecstasy.
ANOTHER THOUGHT AND APPARENT CONTRADICTION: Even if we were to add a half point (or a full point) to all the dogs that lost so far this season, the results would pretty much be the same. The favorites would still be winning big. So, it's wrong to suggest this is the sole reason why the dog players and contrarians are losing. I merely want to post a few thoughts that suggest oddsmakers may have to GIVE EXTRA VALUE to dogs to make them more attractive to bettors. Since the public is all over the hot teams and stay away from the bad teams, the lines will eventually have to bear out the fact that there will be value in betting these dogs.
THE BOTTOM LINE: These results cannot continue. If they do, some sportsbooks are going to go bankrupt. What should happen is -- some favorites will be over-valued and thus will be bad plays (I expect Kansas City, Seattle, and Denver will fall into this category). I do want to establish the fact (and I have) that dogs have been getting shafted, both by linesmakers and in the game results. I also want to point out that linesmakers MUST make adjustments for these results, or they will go out of business. Dogs are going to have to be given more points, or the public will continue winning.
SPORTS HANDICAPPING REPORT
Thursday, September 25, 2003
NEW: SO, WHAT HAPPENED BY YEAR'S END THE LAST TIME (2001)?
Note: You may want to read the first part of this essay -- "THE DOGS ARE KILLING US." Otherwise, these statistics won't make a lot of sense.
In the previous year where the favorites started out hot, the trend continued for the first four weeks of the season . Then, things rapidly reversed themselves.
As I mentioned previously, 2001 was a fluke year where the favorites were winning way beyond normal expectation in the early part of the year. Coincidentally, that was the worst period I've ever had at MadJacks (other than the slump I'm now in).
Here are the results (W-L record of favorites in parenthesis):
2001 -- Week 1 (7-6)
2001 -- Week 2 (8-5)
2001 -- Week 3 (9-4)
2001 -- Week 4 (9-5)
So, the first four weeks of 2001 produced a record of 33-20, which is about 64 percent winners! Does this sound familiar?
Now, let's look at the next few weeks of that season:
2001 -- Week 5 (2-9)
2001 -- Week 6 (1-14)
That's right. Amazingly, the favorites all crashed and burned the next two weeks. They went a dismal 3-23 in Weeks 5 and 6. It's difficult to know if the linesmakers adjusted the numbers to be more generous to the dogs. I suspect that things just evened out eventually and we saw a normal return to the mean. However, if you go back and compare Week 4 -- versus 5 and 6, we note that many of the lines were around three points. In weeks 5 and 6, many of the lines were up around 6 points. I realize this is subjective reasoning, but I do want to report that the lines INCREASED which might be a factor as to why the dogs suddenly went on a solid run.
By season's end in 2001, the dogs ended up winning 60 percent in the final 13 weeks of the season. In fact, looking at all the data, dogs have NEVER performed worse than 49.75 percent in the last twenty years of line tracking.
What this means is -- dogs are very likely to return to form in the next week or two. Whether or not you want to wait and let Week 4 play out (assuming there's some correlation to 2001) is up to you.
THE DOGS ARE KILLING US!
I had an interesting telephone conversation yesterday with a fellow handicapper who runs a computer program based purely on statistics. The system he uses has never had a losing NFL season in the six years he's worked on it. The system usually produces about 60-70 units profit per year -- using a spread of 1-5 units per game per week. That might not sound like much, but it's proven to be a relatively "risk free" way of producing a profit. I've tracked his system, so I know his numbers are accurate.
What interested me in his system is that it plays dogs mostly -- with assorted weighing factors.
The results of the 2003 NFL season so far have been a disaster for dog players. I wanted to see if my disappointing results are shared by anyone else out there. I was particularly interested hear the feedback from this "dog" handicapper, who plays dogs for statistical-based reasons (I'm oversimplifying his methods to make this easier to understand).
First, let's establish the fact that dogs have performed horribly this season. In all my years as a handicapper, I do not recall favorites ever covering at such a high percentage through three weeks of the season.
Here's the proof. In the early games last Sunday, the favorites went 7-0! That's right -- a perfect 7-0! They faded in the later games, but still ended up going 9-4-1 for the week. That brings the favorites to 28-14-2 for the season, to date. Favorites have covered in 66 percent of NFL games! Amazing!
Why?
An interesting point was raised about possible reasons why the favorites are blistering the board. Here's one theory. My colleague mentioned that the dogs do not seem to be getting as many points, this year -- as before. His theory is that oddsmakers got hammered with dogs last season, and have adjusted the numbers towards parity. In other words, the dogs are currently not getting as many points as in previous years. Our discussion continued. Using a couple of games from last week as examples -- Pittsburgh probably should have been favored by greater than 4.5 points versus dismal Cincinnati (based on the perceptions of these teams). Pittsburgh was a -6 road favorite at Cincy last season -- why just -4.5 this year? And, Super Bowl champion Tampa probably should have given more points to "Vickless" Atlanta. So, there is some justification for this idea.
I decided to go back and compare all the numbers from this season versus last season. For both seasons, there were exactly 46 games played through the first three weeks.
In 2002, the average dog was getting 5.1 points. That means the average spread of all NFL games was -5.1 Contrast this with 2003. The average dog is now getting just 4.5 points. That's a drop of about a half a point per game. Looked at another way, dogs have lost about 12 percent of value based on the "shift" towards favorites by linesmakers.
Of course, we must take into account that the game matchups are different from season to season. (Perceived) Lopsided games produce higher lines than close divisional matchups. Nevertheless, I believe there has been a conspicuous decision on the part of oddsmakers to drive down the numbers and to reduce the value of dogs.
SIDE NOTE AND AN APPARENT CONTRADICTION: Several sportsbooks are getting hammered in the NFL (I know this for a fact). The next question is -- why would the sportsbooks reduce the line value of dogs if they end up losing money? Of course, not even the sportsbooks could have predicted these astounding results. They were so afraid of the big money coming down on the dogs again this season that they intentionally shifted the line values towards the favorites, knowing most of the "sharps" are dog bettors. The had no way of knowing that the lines in most cases wouldn't matter and the favorites would win at such a high rate. So, the public is cleaning up -- at least temporarily.
ANOTHER SIDE NOTE: Most Las Vegas sportsbooks do not post independent lines. They get their numbers from other sources. I am not suggesting that each sportsbook has intentionally made a decision to move the lines towards the favorites. I do believe this decision was made by some of the key offshores and linesmaking services and this has filtered into the Las Vegas sportsbooks. If you walk into any Vegas sportsbook this month, you've seen the $50 bettors in ecstasy.
ANOTHER THOUGHT AND APPARENT CONTRADICTION: Even if we were to add a half point (or a full point) to all the dogs that lost so far this season, the results would pretty much be the same. The favorites would still be winning big. So, it's wrong to suggest this is the sole reason why the dog players and contrarians are losing. I merely want to post a few thoughts that suggest oddsmakers may have to GIVE EXTRA VALUE to dogs to make them more attractive to bettors. Since the public is all over the hot teams and stay away from the bad teams, the lines will eventually have to bear out the fact that there will be value in betting these dogs.
THE BOTTOM LINE: These results cannot continue. If they do, some sportsbooks are going to go bankrupt. What should happen is -- some favorites will be over-valued and thus will be bad plays (I expect Kansas City, Seattle, and Denver will fall into this category). I do want to establish the fact (and I have) that dogs have been getting shafted, both by linesmakers and in the game results. I also want to point out that linesmakers MUST make adjustments for these results, or they will go out of business. Dogs are going to have to be given more points, or the public will continue winning.
