Yes you are correct, if you NEVER bet a line that is more than -3, then that eliminates the psychology. Unfortuantely it takes a great deal of commitment to look at a team you think should be -10, and see them at -5.5. I have to think that may happen in 5-10 games a year. THAT is the psychological factor I am speaking of. If those games that you see that truly do stick out, cover, then what good does following that line of thinking do for you? You basically gave away 5 to 10 winners in a sport where 5-10 winners can most definately mean the difference between winning or losing.
But it isn't a criticism, it is more of an observation. Any line of thinking that is broadband and basically based on a blind assertion doesn't make sense. You can bet a team that moves to -2.5, but refuse to bet them if they are -3? Even if you bet the 3, NOT betting them at 3.5 is the same difference. Letting an arbitrary line made up by people who's sole job is to confuse, and moved by people that don't really have a clue more than half the time doesn't make sense.
NFL is maybe the sport where experience and finding small edges on a team by team basis, makes the most difference. So when you eliminate more than 2/3s the games on the schedule, at least on one side, it is a huge detriment.
To be successful you are basically relying on dogs to cover. Other than that you have 4-5 games a week with -3 faves or less, and basically you have about a 50/50 shot of having them win and cover as well.
It just makes your plays go from around 512 possible teams to maybe 300. That is a lot of sides to eliminate on a rule like this.
IMO it should be based on a game by game team by team basis. Hope you make enough right decsions to overcome the wrong ones. If I can hit 54% regualrly, then I will still hit 54% by betting the "bad" numbers. I doubt that eliminating plays will make my win percentage high enough to over compensate the volme.