From Scrimmage

The source for college football, recruiting news

Double Click to select a Photo
No. 17 LSU lost to No. 20 Ole Miss.::Rogelio V. Solis/AP

By Hugh Falk,

When two teams' records are the same, the winner of the head-to-head match-up should be ranked higher.  I call this the "head-to-head rule."  Based on feedback, it seems some people still aren't sold on the idea.   Let's start by way of example:

Nineteen AP voters ranked LSU (8-3) over Mississippi (8-3) this week after the Tigers lost to the Rebels. What's interesting is that even though 41 out of 60 voters ranked Mississippi higher, LSU still came out ahead in the poll (No. 17 to Mississippi's No. 20).  That's because LSU voters heavily favored the Tigers while about 30 of the Ole Miss voters only ranked the Rebels one spot above the Tigers.

I actually thought we would see more than 19 voters rank LSU higher because last week the ranking gap was huge between these two teams in LSU's favor.  So while I'm pointing out that 19 voters still think that LSU is a better team after the loss, I'm also giving kudos to the other 41 voters for making "significant changes in your ballot from week to week" as the AP voter guidelines state.  Too many voters continue to use the "slide" rule of moving teams up a limited number of spots after a win and sliding down teams a limited number of spots after a loss.  When voters do this, it is tough for lower ranked teams to make headway, and it is the reason that the AP included the "significant changes" guideline. The truth is that some of those 19 voters might think LSU is still a better team, but I'm confident that some are just using their own slide rule.  And as we all know, slide rules were replaced by computers years ago.

On that note, the computers think LSU is better than Ole Miss, too.  So there is obviously merit to ranking LSU higher if you don't believe in the head-to-head rule.  But that is one of the things that separate the humans from the computers.  Humans put a lot of faith in head-to-head results; computers don't.  (Except Billingsley, which has a one-week, head-to-head rule.)

Do people really support the head-to-head rule?   I've heard plenty of arguments from people about why head-to-head isn't that important, but for some reason they usually come from fans of teams that recently lost a game.  However, when talking objectively (like before the season starts) I hear far fewer arguments.  In fact, I hear many calls for a playoff system, which is the most unforgiving head-to-head rule of them all - you lose, you're out. It doesn't matter how good Team A usually is or how bad Team B usually is. The popularity of playoffs is the biggest evidence that humans accept head-to-head results for determining their champions.

The head-to-head rule is far more forgiving than a real playoff. At some point, however, you have to say, "Sorry, you lost. Better luck next year." If fans and coaches were allowed to rationalize who progresses in a real playoff, there would be bloodshed. Obviously, this season-long playoff isn't a formal rule, and I don't expect it to be. However, I do think it is a fair method of ranking teams with the same record. Voters who don't follow it should certainly be checked for ignorance (i.e. forgot), bias or worse.

To be clear, I don't think anybody should follow the head-to-head rule blindly.  Every rule has exceptions, but I do think it should be the standard until there is a good reason to do otherwise. The decision to rank a loser over a winner should be made objectively and intentionally.   Personally, it would take a special situation for me to do so, but it is possible.

Keep in mind that one of the goals of Pollspeak is to help weed out bias and even corruption in college sports. So it is important that voters are able to justify their rankings. Otherwise voters could be unreasonably biased or (and I know it sounds paranoid) they could be bribed or coerced. That is why it is important that the process be transparent (no secret ballots) and why voters be questioned about what seem like irrational choices. Going against head-to-head results is one way to help identify bias, but it isn't definitive. Voters can have valid reasons for going against the head-to-head rule. Obviously, it also isn't the only indication of bias.  Pollspeak will be tracking voters over the season and over the years to identify patterns of bias. Does somebody vote Team X higher because they think they are a better team that week, or do they just usually rank Team X higher?

I also think the head-to-head rule holds more weight earlier in the season. For example, if Boise State beats Oregon in week one, why wouldn't you rank the Broncos higher that week?  In week one, who is to say one team is better than another based on anything other than the game they played against each other? As the season progresses, the sample size is bigger, but don't be fooled. The football season is way too short to put more faith in transitive speculation than what actually happened when the teams meet on the field. There is already so much subjectivity in the BCS (i.e. which is the best undefeated team). That is the price of having so few games in the season.  We don't need to create more subjectivity by valuing biased opinions over actual results when we are lucky enough to have them.

So should Ole Miss be ranked over LSU?  An LSU fan could argue that the Tigers lost to better competition:  Florida instead of South Carolina for example.  Their point is that if Ole Miss lost to inferior competition, the Rebels must be inferior.  However, LSU lost to...well...Ole Miss.  How bad is Ole Miss?  The Rebels can't be so bad that they make LSU look like they lost to inferior competition, but if they're not inferior, then the Rebels should be ranked higher.   Also LSU hasn't beaten a currently ranked team, while Mississippi beat...LSU.  Shouldn't who you beat count as much as to whom you lost?  It's a catch 22 that doesn't really work in LSU's favor no matter how you look at it.

Even if you think LSU is better, I don't know how anybody who watched the end of that game can conclude that LSU is MUCH better.  Granted, LSU had a chance to win...on the road.  However, it took an on-side kick and a Hail Mary to get into that position.  Then there was the infamous spike heard round the world, which was a clock management mistake of epic proportions -- a mistake that is indicative of coaching issues beyond a typical fumble or injury problem.  The only slide rules I follow are posted on the playground:  "Always slide down feet first and sitting up, never head first."  I think both of these teams should be closely ranked either way, but based on the head-to-head result, I'd give the edge to Ole Miss.


Remember to keep your posts clean. Profanity will get filtered, and offensive comments will be removed.

Truth & Rumors


  1. 1
    Kerr 'absolutely expects' Knicks offer
  2. 2
    No return timetable for Lightning MVP
  3. 3
    Yankees, Mets, Red Sox among Hanrahan hopefuls
  4. 4
    Niners table talks with Kaepernick
  5. 5
    ... So, L.A. will line up for Love and Durant

SI Photos