We'll take a break from the top of the chart soon, and and please feel free to comment on these "top-25 issues" because I am not sure if it's my system or my perception that's flawed. Are some of these "against the grain" rankings telling us something about certain players being overrated/underrated, or is the system just mssing something important?
Rated About Right
Dan Marino
Before getting into the other guys, I'll start by somewhat "defending" the Marino ranking. He typically goes between 16-22. The rankings consistently uphold the wisdom that Marino is greatest quarterback to never win a Super Bowl, and it goes through all eras for this; only Tarkenton occassionally pops up higher. Considering Marino's passer rating isn't such a standout and the award index has been largely faded, I think the system is doing its job. Marino is one of the 20 best quarterbacks ever, which is pretty darn good. When you fade out the pre-1967 players, Marino is #9.
I could also get into the Terry Bradshaw issue, but think of it as the exact same problem as Marino but with opposite strengths and weaknesses.
Underrated
Norm Van Brocklin
Van Brocklin has always been well-regarded as a top player of the 1960s, but he's been coming up as a top-15/18 overall player due to high very high era-adjusted passer rating.
Bob Griese
The rankings have been consistenly showing Griese is underrated by common rankings, as he's often top 20. His era adjusted passer rating is actually slightly higher than Marino, and his peak year is not even that much lower either. I think he suffers, by fan perception, of not having the gaudy big yard, big TD passing numbers or any "wow" factor, but all of the credentials are quietly there.
Ken Stabler
He's a lot like Griese; he has the numbers and the accolades (All-Decade too) but seems to lack the fanfare of a really big name. He's typically scratching the top 25.
Arnie Herber
If we era-adjust, I belive Herber is the third best Packers quarterback and ahead of Favre. He won four championships in the 1930s, had already won two before Don Hutso arrived, and paved the way for Baugh and Luckman. He was replaced by Isbell, retired, then came out of retirement to play for the Giants. He made it back to the championship game (lost to his former team, the Packers.)
Ed Danowski
It's hard to get too fired up about a guy from the 1930s, but by all available information, Danowski was basically the lite version of Arnie Herber and the class below Sammy Baugh and Sid Luckman. He played in four championships games, winning two. Other players form that era are in the Hall, and other players with that resume are in the Hall.
Ben Roethlisberger
I can't say I really like this one...that the system constantly ranks Roethlisberger in the top 20 all-time, usually in the 18-22 range. I think Roethlisberger is really good but shouldn't sniff the top 20 either, if we're going back 100 years. I've considered adjusting the championship points based on some type of awards index multiplier (the adjustment would be slight but would make a difference.)
Len Dawson
I think most of us realize that Dawson is great and belongs very high, but to a casual fan, he's never talked about in the same sentence as Starr, Unitas, Staubach, etc. Most people know he won a Super Bowl but don't realize he also won two AFL Championships and was one of the best quarterbacks in the country for years.
Overrated
Troy Aikman
Aikman is often in the top 10/top 12 in Super Bowl era rankings, but he'stypically around 25 here. Everything besides his three Super Bowl wins is substandard for that level. Perhaps it's because the system doesn't account for actual Super Bowl/postseason play but just the results. But you also have to ask if a guy with an overall career .57 winning pct and a roughly average era-adjusted rating was very much a...gulp...system quarterback.
John Elway
Elway was coming in around 12-15, but the addition of peak score has been unkind to him as he's now coming in 18-22. I think that the system is off on this one, and like with Roethlisberger, I'm looking for missing gaps in the criteria. Bottom line is Elway should never rank below Roethlsiberger, or even very close. The all-star/all-pro index shows that Elway is a star player while Roethlisberger is not; while I don't think that should give out more raw points, it should be able to identify and correct this problem with a multiplier.
Sammy Baugh
This is certainly a "first world problem" as you have a top-10 guy from the 1930s who is being considered overrated. But this is more about Baugh's ranking compared to Luckman. Baugh is widely regarded as the best QB of that era, but Luckman beats him almost every category, including passer rating and peak rating. Most signifcantly, Luckman was 4-1 in championships and Baugh was 2-3. Three of those games were against each other: the 73-0 game to begin the rivalry, the shocking upset of the 1942 Bears (maybe the greatest team ever), and a Luckman win in the rubber match. Incredible matchups; if you flip that head-to-head championship count, it would flip their rankings. And that's exactly how it should be.
Bobby Layne
Layne was initially showing up as top 10-15 before more precise era-adjustment rankings and peak score were added. Layne is a paradox who needs to be studied further. It's possible that he's closer to Troy Aikman than he is to Otto Graham, but he's probably in the middle. His career stats are not impressive and roughly average for his era, but he does have the pro bowls/all pros. In 1957, Layne got injured, and the Lions still marched to a championship with a career journeyman, Tobin Rote., dominating the Graham-less Browns 59-14.