NFL 2005 Conference Championship Results

Well, Denver sure got beat. As good as Denver’s defense is, they were dominated by Pittsburgh’s offense for a huge portion of the game, and I’m not really even sure how Denver needs to improve their defense to not let that happen again. Pittsburgh’s third down conversions just killed them. That said, how they reacted in the second half showed me some glimpses of just how good Denver really is (was), that I think had been taken for granted. If not for some impatience, which I think came from Denver’s lack of experience in playing from behind, Denver could have had more of a shot. Those two interceptions were forced passes when they didn’t have to be. I hope Denver gets some more experience next season in generating success from players imposing their will, rather than just from great coaching schemes.

I live in Seattle territory, but I’m having a hard time becoming a fan – I might find myself cheering for Pittsburgh at the Super Bowl.

Here’s the beatpath graph for after the conclusion of the games. It’s much the same, except for Pittsburgh shedding its last beatloss.


Under the normal beatpath scheme, Pittsburgh would have no shot at getting the #1 slot from winning the super bowl. But, this is where the beatflukes variant appears more accurate, once again. For those who are unfamiliar, here’s how the beatflukes variant is different: If a beatloop exists, but a portion of the beatloop is contradicted by one team having a beatpath to the other, the beatloop is broken there, and the resultant segment is applied back into the graph. For example, if A=>B=>C=>A, but C has a beatpath to B, then C=>A=>B becomes a valid beatpath.

The reason the beatflukes variant has been in beta all season is because there is still some question on the ordering of evaluating beatloops – but it has been slightly more accurate all season long. Here is what the beatflukes variant of the graph looks like after the conclusion of the Conference Championship games:


In the beatflukes variant, Pittsburgh is ranked ahead of Denver, but not Indianapolis. But if Pittsburgh beats Seattle, they’ll be ranked #1.

12 Responses to NFL 2005 Conference Championship Results

  1. BillB says:

    Root for the Steelers? Are you nuts?

  2. Israel says:

    “But if Pittsburgh beats Seattle, they’ll be ranked #1.”

    I would hope so!

  3. Tim says:

    Have you thought about going back over the season and seeing how accurate your beatpath method is of predicting who wins using only data from past games? ie only using data from weeks 1-4 to predict the week 5 games. Would be interesting.

  4. ThunderThumbs says:

    re #3: Unless I’m not understanding your comment, that is how it’s currently done. When I backtest a season, I come up with the “predictions” only using the results from the games before that point.

    What I haven’t done is used the season’s *final* beatpath graph to see how it predicted the game results over the previous season. Actually, that is a great idea, because it seems that that would pretty much be the ceiling of how accurate a picking system could be for that season, at least in terms of judging by wins and losses.

    And, hey there Israel. Yeah, this was a touchy season – in general, you’d expect the super bowl winner to end up atop the beatpath graph, but it’s definitely possible for that to not be the case, especially if there are a lot of upsets on both sides. The beatpaths seem to respect the “any given sunday” reality of football, which is why defeating an opponent isn’t always enough to ensure they’ll immediately be ranked ahead of them – witness Pittsburgh still being ranked behind Indianapolis. If the Super Bowl had ended up being Pittsburgh against Carolina, then I think the odds would have been low that either team would have been #1. But Seattle’s good enough that Pittsburgh beating them will finally allow them to take the top spot – at least in the beatflukes variant.

  5. ThunderThumbs says:

    re #1: I’ve always been confused about this. If your team loses, do you cheer for the team that beat you because then you can rationalize that at least you might have been second best? Or do you hope your conqueror loses for revenge, even if it means your team is by definition that much worse?

  6. Tim says:

    What I meant was that it would be interesting to see how many games the beatpath method would predict correctly, ie “120 out of 180 correct” or something like that. Have all the predictions from the season totalled up to compare with other people’s picks for the season.

    The idea of using all of the season data to predict past season games is also interesting. You could compare it to how many correct predictions that method makes vs predictions made during the season.

  7. doktarr says:

    TT, I think it depends. In the absence of any bad blood, I think you root for the team that beat you. As a Redskins fan first, Indy fan second (the divisional weekend was no fun for me!), I am glad to see the two teams that beat my teams meeting in the Superbowl. But if it had been the Cowboys and the Patriots in stead of the Seahawks and Steelers, then it would bring me no joy.

    And as I’m living in Denver, I have to admit I’m a little happy that I don’t have to deal with Bronco-mania for the next couple weeks. I take no pleasure in the sadness of the Denver fans, but I’m glad to be free of the hoopla.

  8. BillB says:

    I’m a Hawk fan, so I’m biased, but I think living in the area and rooting against that team after all their fans have been through over the years is a bit…well…tacky. Hawk fans define the term “long suffering.”

    But as I said, I’m biased.

  9. Israel says:

    #4 – I know. I just couldn’t help walking through that door you left wide open.

  10. Ken says:

    Re #4 & #8,

    It is hard to figure out exactly how to root in those situations. It probably matters how you lost, and what you think of your team.

    I’m a Chicago fan, and I was very happy to see Carolina get smacked, even though it makes my Bears look weaker by comparison. The only reason I can give is that the game was close, we had a shot to win, and I honestly think that our team is as good or better than the Panthers (I know, the evidence says otherwise). So I feel confident that my team is good, even if the Panthers were to lose 100 games in a row with the exact same team.

    In comparison, in 2001-2 I wanted the Eagles to go all the way after they beat the Bears. That’s because I was not entirely convinced the team was as good as their record said, and also that the Eagles were CLEARLY the better team. So I wanted more evidence that the Bears were a good team, or at least more evidence that they weren’t a bad team (of course, I didn’t get it).

    I guess my point is, if you believe that the Broncos are a good team no matter what happened last Sunday, root for the Seahawks. If last Sunday made you think that maybe the Broncos weren’t so good, root for the Steelers.

  11. BillB says:

    RE: #10

    The Bears lost due to a rather pathetic defensive gameplan. It’s not a question of talent, it’s how it was deployed on tha t particular day.

  12. Ken says:

    RE: #11

    I know that. Believe me, I know that.

    If we had used the same gameplan Seattle did against Carolina, I think it would have been us in that game.

    But my point is that the game was still close, and the biggest difference was because of a coaches decision, which is hardly something that can’t be overcome. In the future. I hope.

Leave a Reply

Your email address will not be published. Required fields are marked *