≡ Menu

538: Carolina, Denver and Super Bowl Rematches

Today at 538, a look at the history of the Thursday Night Opener and how teams fare in Super Bowl rematches.  Also, in researching for this piece, I found one of my new favorite pieces of trivia:

On January 31, 1993, the Cowboys obliterated the Bills 52-17 in Super Bowl XXVII.  364 days later, Dallas beat Buffalo in the Super Bowl again, 30-13, the only time the same teams have ever met in consecutive Super Bowls.  But in between those historic games, the teams also met in Dallas during Week 2 of the 1993 regular season. The Cowboys, missing Emmitt Smith because of a contract dispute, lost 13-10 on a late field goal. It was a result symbolic of that entire Bills era: in games started by Hall of Fame quarterback Jim Kelly from 1990 to 1993, Buffalo went 14-0 in the regular season against the NFC, including a 4-0 mark against Dallas, New York, and Washington. Against those same teams in the Super Bowl, of course, the Bills went 0-4.

You can read the full article here.

  • sacramento gold miners

    Those early 90s Bills defenses just seemed to get worn down in the Super Bowls, and they didn’t have the physical edge in the secondary other teams seem to have. Safety Mark Kelso had to wear a protective helmet, and one of the defining plays for the Bills in the Super Bowl era was that short pass to Mark Ingram who dodged many defenders for a big first down.

    • Richie

      Are you suggesting Mark Kelso’s soft helmet shell was a reason the Bills lost those 4 Super Bowls?

      • sacramento gold miners

        Kelso wasn’t a physical safety, and the Bills could have used more of those type of players.

  • Richie

    That is a great stat. I lived through the Jim Kelly era, and had no idea that they were undefeated against the NFC during their run of Super Bowls.

    I hate when people say “Buffalo Bills lost 4 straight Super Bowls, they suck, LOL”.

    • JeremyDeShetler

      There’s a typo there somewhere. The Bills went 14-2 against the NFC in the 4 Super Bowl years, losing in Week 17 in 1990 (vs Washington) and 1991 (vs Detroit). They were 4-0 vs the NFC East in 93 (the year Chase is referencing) and also 4-0 vs the NFC West. I believe games were meaningless for the Bills, who had locked up home field, but had value to their opponents (Washington for a wildcard and Detroit for a 1st round bye).

      Seconded on that 2nd comment. I was in high school & college in the 90s and knew a number of people who were merciless idiots when it came to the Bills.

      • Not a typo, I don’t think (or, at least, I hope not!)

        Yeah, the stat I used was 14-0 with Jim Kelly. He sat in those two meaningless week 17 games.

        • JeremyDeShetler

          Doh! That was my bad. My brain crossed a wire there and I misread the clause about games started by Kelly somehow.

    • Tom

      Yeah, don’t get me started. While it’s true that they suffered some horrible blowouts, you don’t just go to 4 Super Bowls IN A ROW without being very, very good.

  • Phil

    off topic:

    is there any literature connecting AV with wins above replacement, or wins in general, or anything like that?

    I’m thinking of the Bradford trade

    If a 1st and a 4th should be expected to be something like (17 + 3 =) 20 AV [if they were picking around 15 http://www.footballperspective.com/draft-value-chart/%5D

    how many wins this year does Bradford have to save them to be worth giving up 20 future AV?

    if he winds up being the difference between 8-8, and making the playoffs at 9-7, is that worth giving up 20 future AV?

    • Richie

      Hmmm…How about a flow of consciousness response?

      AV is primarily based on team points scored per drive (or team points allowed per drive for defense). Obviously, W-L record will be somewhat dependent on how good the other side of the ball is.

      I took a quick look at the past 5 seasons, and there is a fairly strong correlation (.76) between a QB’s AV and games won. (Among QB’s with 300+ pass attempts in a season.)

      The correlation between win differential (wins minus losses) and AV is a little weaker (.62).

      In 2015, Peyton Manning was the biggest positive outlier of the past 5 years. He had an AV of 2, but won 7 games (a +5 differential). Going back to the merger, players with an AV of 3 or less, average winning a little over 3 games.

      In 2005, Kyle Orton won 10 (+5 W/L differential) despite posting an AV of only 2.

      On the other side, Cam Newton in 2011 was the biggest negative outlier. He posted an AV of 19, but only won 6 games. QB’s in the 18-20 AV range usually win about 12 games.

      Over the last 5 years, each point of AV (for Quarterbacks) has been worth about .66 wins. I am going to guess that the correlation between wins and AV is much lower for non-QB’s. (I wish PFR tracked W-L for all positions – not just QB.)

      The AV for non ball-handlers is trickier, because Pro Bowls and All-Pros are a key component of AV, and those awards are sometimes based on team success, so it can get a little circular to turn AV into wins.

      Also, I think the number of players that combine to get the 20 points of AV is relevant. A 20 AV player is harder to come by than two 10 AV players. And ten 2 AV players are WAY WAY WAY easier to come by.

      Additionally, we need to consider replacement value. Who would that 4th rounder replace on the Vikings roster?

      To me, if Sam Bradford is only 1 win better than Shaun Hill, then 20 AV is way too much to give up.

      (Remember that 17 AV first round pick you noted is the net AV above 2 per season for 5 years. So that would be a guy who actually earns 27 AV over his first 5 seasons – a little more than 5 per season.)

      But if we assume that each point of AV for all players is worth .66 wins each, and the marginal AV that Minnesota traded was 20, that would mean Bradford would need to be worth 13 more wins than Shaun Hill. I think one thing to consider is that the Vikings are concerned about Bridgewater’s 2017 season as well. I think there is a chance that Bradford could actually be worth 13 incremental wins to the Vikings over the next 2 seasons.

      • Fantastic post, Richie. Stuff like this is why the comments section here rocks.

        • Richie

          I should probably have my boss bill you for lost time. I was just going to add a quick thought, and instead spent an hour on the Play Finder.

          I wish I was better at turning that data exploration into a coherent posting.

          • Tom

            Man, I hear you. First thing I do when I get to work is check out FP and then think, “Well, I’ll just write this one comment” and then about an hour passes by in which I’ve copy and pasted 10 years of PFR data into a spreadsheet and made one lousy comment…hilarious.

            • Richie

              Haha. Same here on all accounts!

      • Phil

        Thanks for taking the time to write that out, that clarifies my thinking about it considerably


        It’s seem that thinking about that trade in terms of linear value exchange, it’s hard to justify the trade from the Vikings perspective

        But it seems like they were fairly uniquely positioned to get fleeced, so they got fleeced

        Obviously Bridgewaters injury was a terrible break to what looked to be a promising team

        The question then becomes should they take there lumps from that break now, or are they in a finitely available window that they need to maximize now?

        It hard for me to think that looking at this as a unique window is the best way to go, but the Vikings obviously seem to disagree, and overall I’m pretty impressed with the job they’ve done building that team, so maybe I should give them the benefit of the doubt

        (Thanks again for writing that out, that was a really good post)

        • Richie

          I think the Vikings think they have a window this season for a possible Super Bowl. For some reason I’m not as crazy about the Vikings as some people are. I feel like they overachieved a little bit, and then due to the cold weather almost stole a playoff win from Seattle.

          They have some decent, young, pieces on both sides of the ball. But, the window for Adrian Peterson is probably very small. 2016 may be near the end for him, so if they want to ride him to a Super Bowl, upgrading from Shaun Hill may be necessary. Of course, we don’t know for sure if Bradford is an upgrade over Hill. (Though it seems likely he is.)