zzz@mit-eddie.UUCP (Mike Konopik) (11/08/83)
Didja ever notice that when, say, the #3 ranked team beats on the #8 ranked team, all the polls drop the latter to, oh, about #15 or so? Why is this? I would think that the latter is SUPPOSED to lose to the former, and that the loss only strengthens the credibility that one team is better than another. Why does it necessarily make the latter team worse than other teams ranked lower than it? The whole thing seems like trying to sort numbers and saying "3 is before 8, and 3 < 8, so move 8 down some"... I'm sure there are "reasons" for this behaviour -- I'd like to hear some comments/opinions on it, tho. -- -Mike genrad!mit-eddie!zzz (UUCP) ZZZ%MIT-OZ@MIT-MC (ARPA)
djb@cbosgd.UUCP (David J. Bryant) (11/08/83)
I've noticed the same thing. As I see it, teams that lose almost always drop in the ratings. The number of positions they drop is related to the ranking of the team they lost to. If #3 beats #8, then #8 is probably not going to drop more than 2 or 3 positions. If #8 is beaten by an unranked team, then #8 can expect to fall quite a distance. This situation is aggravated if #8 was #4 last week and lost a game they weren't supposed to. This can produce a rapid plummet, perhaps out of the rankings altogether (similar to what happened with West Virginia and North Carolina this season). I suppose you must consider that #9 and #10 are close in strength to #8, and that since they won their games they should move up at #8's expense. The paradox is as you pointed out - you expect #3 to beat #8, so why should #8 lose ground? It is as if the #8 team has a feast or famine future. If they beat the #3 team, they jump up several spots, if they lose, they drop several. No maintaining the status quo. There are all kinds of examples of this phenomenon, with interesting variations depending on how closly ranked the teams were, whether they were in the first or second ten, etc. It's a common enough occurrance that most pollsters must just do reflexively. Still, it's curious... Another more unusual case is when a team loses and moves up. This happened this past week with E. Carolina. They played Miami (Florida) a great game and lost in the last few minutes. As a result (plus based on their strong showings earlier in the season) got them rated #25 in this week's USA TODAY Top 25, where they were unranked the week before. David Bryant Bell Labs Columbus, OH (614) 860-4516 (cbosg!djb)
beth@umcp-cs.UUCP (11/10/83)
Let's make this a more concrete discussion. Auburn (#3) beat Maryland (#7) by 12 points on a last few seconds touchdown on a fumble recovery. Maryland had shown on a previous drive that they could march 99 yards for a touchdown. It was Auburn's homecoming. Maryland is now 7-2. In the coaches' poll, Maryland was ranked #17 after that loss. The sportswriters ranked them #11. Seems that the coaches looked at the final score and/or don't take Bobby Ross (the Maryland coach) very seriously. Maryland is ranked just above North Carolina (who plays a teribly weak schedule) and behind West Virginia (who beat Maryland) and Pittsburg (beaten by Maryland when Maryland was using their second string quarterback). Now if Maryland loses to Clemson, maybe the coaches were justified, but I think they aren't really paying much attention to how other teams, particularly those that don't have national exposure, are playing. --Beth Katz