Search the Community

Hello everyone, it's been a while since I did this. I figured I had a good thing going when I did the data for 2004-2017, and I wanted to do something for other eras as well. So I'm looking back at the timespan from 1990-2003 to figure out the answers to some important questions. How did the long term starters of the era fare when their defense was above par, and when their defense was below par. I used the same methodology as the last time I did this, finding the point where the average defense lies in my selected time span. From 1990-2003, the average defense gave up about 20.3 points, about 2 points lower than modern defenses do today, so I set the breaking point at 21 points. If your defense allowed less points than that, they had a good game, if they allowed more, they had a bad game. The null hypothesis is that the average QB will win every game his defense allows less than the average amount of points, and will lose every game the defense allows greater than the average amount of points. Obviously, the data shows that this isn't true due to games where both teams score 20 or less points, and games where both teams score 21 or more points. The real percentages over this time span show that the average QB can be expected to win 74.23% of the time his defense holds up, and a mere 22.86% of the time his defense crumbles. With those numbers gathered, I looked for every QB who had at least 90 starts and put them onto the list for analysis. I added one exception who was fairly close to the amount of starts simply because he was one of the most impactful QB's of the era, and it felt wrong to leave him off the list. The reason for 90 was just because I wanted a round number, and 96 would be 6 full seasons of games, so going a little under due to injuries and such felt better than going a little over. I managed to get a list of 25 guys. Some played their entire career in the time span, some only had the tail end of their careers in there, and a few guys were just starting up. Neither Tom Brady nor Joe Montana made the list due to lack of starts in the era. I curated some games, removing starts where players only played a few snaps before getting hurt, or adding starts where they came in relief, because the place I get my data from just counts starts as the QB who plays the first snap. Anyway, let's get right into the data. First let's look at the winningest QB's of the era: https://i.imgur.com/tP5DoKU.png As you can see, there's a lot of names you'd expect at the top, Steve Young, Jim Kelly, John Elway and Brett Favre. Meanwhile the bottom is about what you'd expect as well, with Jim Everett, Jeff George, Jeff Blake and the like at the bottom. It's a good indicator, but it doesn't tell us how their defenses held up, and it's really just a blase look at things, so let's keep exploring, and see which of these guys had the best defenses. Sorted by Good Defense percentage: https://i.imgur.com/9PGxixv.png So this was something I found interesting. I was not expecting Troy Aikman to have the best defenses of the bunch, but as it turns out he just narrowly edged out Steve Young, who I was expecting to be really high up there in that regard. Despite only playing 12 games for the Ravens, Trent Dilfer has the third best defenses of the bunch, bolstered by his time in Tampa with Tony Dungy and their Tampa 2 defenses. On the other end, you've got the guys whose defenses gave up a lot of points. Jake Plummer, Jeff George, Jim Everett, and Peyton Manning who unlike those other guys managed to pull out a winning record despite his porous defenses. Jeff Blake sitting at the bottom makes a ton of sense when you realize he played for the Bengals, Saints and Cardinals, all before they got good (if they ever did). Continuing on, let's sort these guys by how they performed when the chips were down, and their defenses gave up 21+ points a game. Remember that the average win percentage in these games is a hair below 23%, so if the percentages look low, that's why. Anything above 23% is basically a passing grade in this curve. https://i.imgur.com/Shltje6.png This would be about what you'd expect. Guys like Young, Kelly and Elway cut their teeth on winning in tough situations, winning over 40% of the time. Add in Favre, Cunningham, Rich Gannon, and Steve McNair who all won over a third of the time and you have a very impressive group overall. Peyton is right behind them as well, but he boasts a 50% rate post 2004 which is insanity only matched or surpassed by one Thomas Brady. Meanwhile the bottom of the list is anchored by everyone's favourite bust, Jeff George. And lastly, what you've all been waiting for, the metric that sorts all of these numbers out nice and evenly, my derived Wins over Average stat. https://i.imgur.com/bVWC8Ae.png This was probably the biggest surprise of all for me. I was expecting this to be Steve Young when I first started this, with about a 20% it'd be Favre somehow. However Jim Kelly is the man up top, averaging an entire 2.82 wins more than the average QB would in his situation. If you want to know why the Bills went to 4 super bowls in the 90's, you don't need to look much farther than Kelly. Though he doesn't get the accolades that Marino and Elway did, there is no question that he's the superior member in the 90's. Steve Young finished a little bit behind with 2.7, Brett Favre had a very strong 2.34, and even outside of his prime years, Peyton still managed to be 2.07 wins over average. Troy Aikman may not be thought of as a good QB, since he was carried by some great teams, but he still managed to be a win above average, which places him with contemporaries like Joe Flacco and Matt Schaub. Dan Marino didn't do too hot, but considering he lost both of his pro bowl receivers in 92, it doesn't really surprise me that he's not dominating this section of the list. His prime was definitely in the 80's. Moving down the list, Trent Dilfer managed to be a shade above average, netting about .17 more wins a season than the average QB would. It's nothing impressive, and his closest contemporary in the modern era would be Alex Smith, another QB who did alright with great defenses, but Smith never lucked into a ring. Jeff Blake is our closest to average QB, netting a mere 0.01 more losses a season than the typical QB, which is virtually 0 at that point. So if you want to say you're a decent starter in this league you need to be better than Jeff Blake. Let's look at guys who are worst than Jeff Blake. Drew Bledsoe, who does pretty well in high scoring games, but vanishes in a lot of defensive struggles. Vinny Testaverde and Chris Chandler, career journeymen with long careers and stormy injury history. Jake Plummer and Kerry Collins, about half a game worse per season than average starters, started early on bad teams and have their best years cut off by the range of this sample size. Then you have the trash. Something interesting about this data set is that there are a lot of QB's with enough starts to qualify who fall well below average. In the modern era list, the only guy below average was Ryan Fitzpatrick, who journeymaned his way to starts on lots of teams. The guys here did it much the same way he did, 90's QBs had a lot looser leash it seems. Jim Harbaugh flamed out with a good Bears defense, and depsite having some of the best defenses out of all the guys on the list, he couldn't manage a .500 record. Boomer was old and miserable, and got his prime years in the 80's cut off by the sample size. His performance when his defense showed up was pathetic, even when he managed to be average when his defense didn't. Jeff George was a colossal bust and his numbers are even buoyed because he got to throw to Jerry Rice, Tim Brown, and Randy Moss in his career. Lastly the worst of the worst was Jim Everett, a guy who could barely win a third of his games. His 21-18 mark when his defense actually showed up is by far the worst I've seen thus far. He may have been on some bad teams, but that is beyond pathetic. The -1.38 WAR is also the worst mark I've seen thus far, below Ryan Fitzpatrick. --- Anyway, this was a lot of fun to make up, and I may do some more in the future, especially since certain QB's aren't represented well, or got their primes cut off by the era. The next data set would be 1978-1989, the start of the 16 game era, as well as the introduction of the Mel Blount rule. If you have any questions, want any trivia or whatever, feel free to let me know. And here is the link to my data sheet. Thanks for reading guys. https://docs.google.com/spreadsheets/d/1nAyFHNzY2KESR7XK3w2IipuFW2Zc3FJKXkc7rodbIVs/edit?usp=sharing