Rankings Philosophy for the HSFB100

The expansion in coverage of high school football across the internet has flooded the market with national rankings. There are multiple human polls as well as two respected computer rating systems, and determining which is valuable becomes the job assigned to each consumer of the information.

At NationalHSFootball.com it was deemed important to explain the rankings philosophy that goes into the production of this site’s signature product — the HSFB100 — as an aid in that decision making.

WHERE DO THE TEAMS COME FROM?: It is the belief of this site that high school football is — at its core — a local sport. When determining which teams belong inside a national HSFB100 poll, the base will be built from the ground up. Each and every week that is where the process begins — by ranking the relevant classes across the country. After each relevant class is ranked, the move is made upward to a statewide compilation regardless of school size, before regional and finally national. The phrase “nationally ranked, locally spanked,” had become too prevalent as rankings often ignored local polls after preseason evaluations had taken place and that is something that was vowed to be remedied within the HSFB100. This is not to say there will not be upsets but the goal is to make sure that when losses do occur locally, they are to teams that are also nationally relevant and not off the radar.

MOVEMENT IN THE RANKINGS: There is a two-part answer for why there is such sudden movement within the HSFB100 and it is the belief of NationalHSFootball.com that each is extremely important in making a quality product. First is an immediate connection with the belief that everything starts locally. If a program moves from second to fourth in its own state then it will move down in the regional as well as national rankings. Those local movements can come off of a high scoring victory and simply winning does not guarantee placement in the HSFB100 the next week. The second point to this movement mentality is that each poll is a separate entity from one week to the next and no spot is guaranteed. The question often arises that Team A was ranked at a certain spot in one week, won its game, and then dropped in the rankings. There is nothing that ties Team A to its spot from the previous week and any evaluation point can alter its placement whether positive or negative.

WHERE ARE THESE EVALUATIONS MADE?: NationalHSFootball.com maintains relationships with coaches, reporters, and recruiting analysts across the country that provide feedback from in-person evaluations or film study. The reason for this system is that in-person reporting is simply the best way to determine which teams are playing well and which are not. There are games that the margin of victory does not tell how dominant the performance was — much like instances where a team can play poorly but open up a wide point spread on the scoreboard. Seeing what goes into the box score is more important than just seeing the final score. Simply put, there are times when a 17-0 final can be more a more impressive performance than a 41-10 score.

HOW MUCH IMPACT DOES DIVISION I ATHLETES PLAY IN EVALUATIONS?: The simplest answer is a significant enough role to address it publicly. The baseline belief of NationalHSFootball.com is that there is much more to a team evaluation than the final tally of Division I athletes — including coaching and surrounding cast — but having superior talent is preferred by both analysts and coaches. Additionally, it has been seen time and again that a team of very good high school football players will beat a team reliant on a star player with little else on his supporting cast. Coaching plays a major role in our evaluations as well but there is a percentage of the team evaluation that does factor in raw talent. Teams loaded with college prospects have proven to be historically formidable and while some have also underperformed, those tend to be the exception and not the rule.

UNTIL THE LAST POLL, EACH WEEK IS INCORRECT: This is something that NationalHSFootball.com has no issue in admitting. The HSFB100 is based on the information currently available and that information changes each week — including injuries, wins and losses, as well as unexpected performances. The most accurate poll will always be the final one of the season because that is when all information is available and nothing is left to learn of the teams. NationalHSFootball.com has no issue with facing that fact each week and wants to address it out front by saying that the journey is more important than the destination. If total accuracy was the sole concern, there would not be a weekly poll — but that simply isn’t fun. NationalHSFootball.com does not run from being wrong; rather, it embraces the opportunity to gather more information and address any incorrect decisions and why they were made.

ARE COMPUTER RANKINGS RELEVANT IN NATIONAL RANKING?: This is a human-based poll and as such NationalHSFootball.com believes that human rankings are a better example of reliable information for ranking a select group of teams — especially at the highest end. The argument against humans is bias and inconsistent evaluations and, while those are certainly valid concerns, it is the belief of NationalHSFootball.com that those concerns are more easily excused than the inherit flaws with a computer rating process. The first point is that computers are not devoid of human error as they are programmed by humans. Computers are also not able to determine what happened in a game as well as humans as its data is based on final scores and not football fundamentals. Why that is important is that not all 38-30 games are the same. One contest could be 30-30 into the fourth quarter and won on the final drive while another may be 38-0 with a running clock and the backup players allowing 30 points late in the game. Those two games would be weighted equally by computers but not humans. Additionally, football is not the same across the country and an algorithm that is laid over all areas will produce outlier results that are unreliable.

WHAT RANKINGS SAY … AND WHAT THEY IMPLY: The reason why NationalHSFootball.com uses 100 teams for its rankings is that it allows for a much more accurate picture to be painted of high school football across the county. There are more than 11,000 teams playing 11-man football. Even by ranking 100 teams, it is still only representing the top 0.7 percent of all those in America. What that means to a consumer of information is that when limited to a top 25, there is a temptation to sacrifice accuracy for an increased scope and wider audience awareness. What a ranking says is that a team is a certain number nationally, but what that implies is that is it better than many other teams in other states. For example, if there were a team from a state like Alaska at No. 25, it may very well be the best team in that state but it could imply that said team would be ranked inside the Top 5 of states like Texas, Florida, Ohio, Georgia … and atop the rankings of many other states. That is a philosophy that NationalHSFootball.com does not subscribe to. The goal of the HSFB100 is accuracy and if that means evaluations place six teams from the same state inside the top 25, that is where they will be ranked. It will not look like that on a poll that is limited to just 25 teams as it would be a disproportionate amount that would exclude too many other areas of interest. NationalHSFootball.com believes the expanded rankings allows for a more accurate product and is always welcome to questions as to its individual state rankings to show where its national poll is derived.

WHAT BEARING DOES THE PRESEASON POLL HAVE GOING FORWARD?: This is an issue within high school football — and most every poll — that a preseason poll sets expectations for going forward. Many polls in existence are reluctant to move teams from their preseason rankings unless that program suffers a loss at the risk of looking like there was an error made. NationalHSFootball.com is fundamentally opposed to that operation and would prefer to not do a preseason poll versus remain beholden to evaluations done in the summer. Mathematically it is unsound when dealing with such a large volume of teams and fundamentally it is unsound as evaluations for preseason rankings are based off of speculation. A poll should be immediately changed once teams are on the field. If it were possible for every team ranked in the preseason to go undefeated, the HSFB100 final poll would not look like the preseason poll as it is statistically improbable for each team to have met expectations. If evaluations are not ongoing, a poll is doing a disservice to its followers.

HOW MUCH DOES THE FINAL SCORE PLAY IN A WEEKLY RANKING?: That is a hard question to answer because it is situational. The perception exists that humans are often overvaluing of blowout victories and critical of close contests. The HSFB100 has tried to stay away from that philosophy and by looking at what actually happened in each game as well as critically evaluating the level of competition. Combining those factors has been a good way to set expectations and make determinations based on meeting, exceeding or falling short of those expectations. Teams that are ranked in the HSFB100 are mathematically inside of the top 0.7 percent of all teams nationally and as such there is an expectation to dominate weak competition. That does not mean there is a magic number — often referred to as “style points” — to meet each week, but to simply dominate teams that have been dominated by others on its schedule. Often the ranking result for teams playing lesser competition is a drop in the rankings and not upward movement, as winning by a lot is the expectation and struggling to do so is not. Likewise, playing a competitive game against a program that is nationally ranked or considered nationally relevant will often produce more of an upward movement. Winning does not guarantee upward movement and losing does not require a drop in the rankings, especially when two teams inside the HSFB100 meet on the field.

WHAT IS LOOKED AT FOR EVALUATIONS: There is a not an exact formula for what goes into a national ranking. That sentence alone has become part of the reason why some consumers have elected to move towards a computer model as the theory that an error applied to every team is better than an error applied selectively. NationalHSFootball.com understand that stance but believes that the strengths that humans hold over computers outweighs any misgivings that may happen as a result of inequity in ranking application. For its HSFB100, NationalHSFootball.com looks at many factors for determining its preseason rankings as well as what is taken into account on a weekly basis. In general terms preseason rankings are determined off of the success a team had the prior season, combined with what it returning to the team, and what talent is moving up or into the program. This is one of the few spots where prior success plays a role in the rankings process. The only other, major area in which past consideration is taken is with coaching. NationalHSFootball.com does not factor in program history to its week-to-week movement but the evaluations of coaches are made and applied to all teams as it is one of the more consistent factors to a team. In other general terms the evaluations are based on perceived strengths at positions groups as opposed to individual talent — excluding quarterback — and given different weights. NationalHSFootball.com places the most weight for game impact on quarterback before moving to defensive line, offensive line, defensive secondary, running back, special teams, linebackers, and finally wide receivers. The other factors for what is considered in the rankings are detailed above but it was important to allow the consumers of this information as much access as to how the rankings are made, what positions are most considered, and where the information is gathered.

Hopefully this post will make for a more educated public. Please post any questions as transparency helps all processes.