Based on four factors — preseason poll, number of voters, number of teams ranked, and final poll coming after bowl games — 1968 could be considered the start of the 'modern' era of the AP Poll.
The first preseason poll was in 1950. Although its usefulness or accuracy is often questioned, it does allow for a more complete view of a team's rise or fall during a given season.
Prior to 1960, voting was open to any TV, radio, or newspaper sports editor. The average number of voters per poll was 158 and varied greatly with as many as 409 for one poll in 1954. Panel voting began in 1960, where a set number of football experts from around the country were selected before each season.
From 1936 to 1960, there were 20 teams ranked in the poll. That changed from 1961 to 1967 where the polls had just 10 teams. This affected the ability to appear in the poll and continue appearance streaks compared to previous seasons. In 1968 the poll went back to 20 teams, and then to 25 in 1989.
And finally, 1968 is when releasing the final poll after the bowl games became permanent (the first and only other instance was in 1965). In previous seasons the final poll came at the end of the regular season, nullifying any effect postseason games between highly-ranked teams could have had. This actually helped preserve the National Championship for five teams who lost their bowl game.
Considering these factors, the years since 1968 have been the most complete and consistent for the AP Poll since its inception.
Based on samples from each season (where available), here are how points have been awarded over the years.
|Seasons||# Teams||Point System|
|1968-77||Top 20||20-18-16-14-12-10-9-8-7-6-5-4-3-2-1 (15 teams)|
Note: Perhaps due to the third split National Title of the 1990s (Michigan/Nebraska in 1997) and/or the start of the BCS in 1998, AP voters have apparently not been allowed to split their ballots among multiple teams - sometimes as many as five prior to 1980 - since then.