It’s that time again.
Things to know about these projections and postings:
First, the Limper is a number based projection model, not a prediction machine – which is NOT a distinction without a difference. As bettors we all make predictions, based on what works for us – eye-test, hunches, favorite teams, etc. Number based projections are just another guide to our predictions.
The Limper model is based on an algorithm using 76 player and team past-performance variables, specific game matchups, as well as other factors like weather, officiating, key-player injuries, etc. All variables are, of course, objectively determined, but are, inevitably, subjectively weighted. (I spend most of the off-season adjusting variable weights, mostly because of rule changes but, generally speaking, the impact of such changes is minimal.)
Key-player availability, naturally, is a major factor, and I run a separate program each week to determine the impact of player injuries on game outcomes.
The first posting each week will be based on the past week’s team starters, and next-game match-ups. The second posting will be based on the run of the Injury Module, which accounts for current starter availability. Further postings – Final Projections - will be made in advance of specific game days, and will include up-to-date player availability.
As the season progresses, I’ll be posting weekly Standings, 2 types of Power Rankings as well as team unit rankings – rushing and passing offenses and defenses, special teams, and quarterbacks, all of which are determined by their own unique algorithms.
The model requires 3 weeks of data to be at all reliable, meaning that through the first 3 weeks of the current season, data from last season – with a few tweaks – is used. Current player and team numbers are integrated each week, but old data does have an impact, so cave aleator.
The key to a profitable use of the Limper’s projections is to focus on the Straight-Up (SU) pick and Margin of Victory (MOV), which is the sole value-target of the algorithm. The ATS Pick will vary depending on the Vegas line, and is, pretty much, disregarded in the model’s calculations.
Example: the first game shown is DAL @ PHI, with the SU favorite PHI winning by an MOV of 7.9 points. Use the MOV against whatever line you’re given to determine the likelihood of covering. In this case, model does a simple calculation – MOV is .4 points greater than the Line of PHI -7.5 – and determines a provisional or Current ATS Pick of PHI -7.5. A not very easy cover.
Generally speaking, an MOV / LINE difference greater than 6 points means the model projects an easy cover. Closer to a 3-point difference or less, and the model’s saying “although I like the projected SU pick - flip a coin on the ATS winner”.
GLTA
PS – Incredibly, I’ve been to known screw up here and there, so don’t hesitate to point out errors. And all suggestions are welcomed.
Things to know about these projections and postings:
First, the Limper is a number based projection model, not a prediction machine – which is NOT a distinction without a difference. As bettors we all make predictions, based on what works for us – eye-test, hunches, favorite teams, etc. Number based projections are just another guide to our predictions.
The Limper model is based on an algorithm using 76 player and team past-performance variables, specific game matchups, as well as other factors like weather, officiating, key-player injuries, etc. All variables are, of course, objectively determined, but are, inevitably, subjectively weighted. (I spend most of the off-season adjusting variable weights, mostly because of rule changes but, generally speaking, the impact of such changes is minimal.)
Key-player availability, naturally, is a major factor, and I run a separate program each week to determine the impact of player injuries on game outcomes.
The first posting each week will be based on the past week’s team starters, and next-game match-ups. The second posting will be based on the run of the Injury Module, which accounts for current starter availability. Further postings – Final Projections - will be made in advance of specific game days, and will include up-to-date player availability.
As the season progresses, I’ll be posting weekly Standings, 2 types of Power Rankings as well as team unit rankings – rushing and passing offenses and defenses, special teams, and quarterbacks, all of which are determined by their own unique algorithms.
The model requires 3 weeks of data to be at all reliable, meaning that through the first 3 weeks of the current season, data from last season – with a few tweaks – is used. Current player and team numbers are integrated each week, but old data does have an impact, so cave aleator.
************************************
The key to a profitable use of the Limper’s projections is to focus on the Straight-Up (SU) pick and Margin of Victory (MOV), which is the sole value-target of the algorithm. The ATS Pick will vary depending on the Vegas line, and is, pretty much, disregarded in the model’s calculations.
Example: the first game shown is DAL @ PHI, with the SU favorite PHI winning by an MOV of 7.9 points. Use the MOV against whatever line you’re given to determine the likelihood of covering. In this case, model does a simple calculation – MOV is .4 points greater than the Line of PHI -7.5 – and determines a provisional or Current ATS Pick of PHI -7.5. A not very easy cover.
Generally speaking, an MOV / LINE difference greater than 6 points means the model projects an easy cover. Closer to a 3-point difference or less, and the model’s saying “although I like the projected SU pick - flip a coin on the ATS winner”.
GLTA
PS – Incredibly, I’ve been to known screw up here and there, so don’t hesitate to point out errors. And all suggestions are welcomed.
