SITE NEWS: We are moving all of our site and company news into a single blog for Sports-Reference.com. We'll tag all PFR content, so you can quickly and easily find the content you want.
Also, our existing PFR blog rss feed will be redirected to the new site's feed.
Pro-Football-Reference.com » Sports Reference
For more from Chase and Jason, check out their work at Football Perspective and The Big Lead.
ShareThisGreatest Coaching Records of All Time
As evidenced by the title, this is not a list of the greatest coaches of all time. I don't feel equipped to answer the question of which coach was the best of all time, much less which coach should rank #9 or #42. That said, I feel pretty comfortable in ranking coaching win-loss records.
Generally, discussions about coaches center around three numbers: championships won, regular season wins, and winning percentage. All three of those metrics have some merit, but all are obviously flawed in other respects. Championships won't doesn't help us decide who was better, Bill Cowher or Tony Dungy. Wins are nice, but are obviously heavily weighted towards coaches with more games. Winning percentage works in theory but it tends to underrate two groups of coaches -- those who have coached for a long time (and therefore lowered their career winning percentage) and those who took over bad teams (and were bad at first but ultimately built those teams into top contenders). Further, it can overrate coaches who haven't been around for very long.
My solution is a formula that incorporates all of those things. We start with nets wins -- each coach gets credit for wins minus losses. A 14-2 season is +12, an 8-8 season is +0. Basic, simple stuff.
If a coach won his conference, he gets +5. If he then wins the Super Bowl, he gets another +7, for a total of +12 in the playoffs. I didn't spend forever deciding those weights, but I did spend a nontrivial amount of time. Obviously there must be a big weight towards winning the Super Bowl, but it can't be overpowering. A 9-7 season with a SB championship would be +14; that is equal to a 15-1 season with no Super Bowl appearance. That seems pretty fair to me. As far as other playoff bonuses, I decided something must be given for a Super Bowl appearance, but I don't know that any other playoff bonus merits anything. I decided against such a bonus, or a division championship bonus -- should we really care that much that you won your division if you didn't get to the Super Bowl? Reasonable people could certainly disagree here, and perhaps I'll be persuaded as such in the comments.
So a 14-2 Super Bowl Championship season is +24; Mike Tomlin gets +20 for his work in 2008. Bill Belichick gets +21 for his 16-0 season in 2007 that ended with a Super Bowl loss.
So Super Bowl winners get +12; Super Bowl losers get +5. In the pre-Super Bowl era, I gave +8 to all NFL Champions and +6 to the six AFL Champions (counting the '66-'69 AFL Champs as part of the SB era). The reduced weight was designed to reflect the fewer teams in those leagues. And that's pretty much the scoring system.
For coaches in the non-16 game era, they get their "wins over losses" number pro-rated. As an example, when Joe Gibbs went 8-1 in 1982, he was +7; a straight pro-rating (of 7*(16/9)) would put Gibbs at +12.4, which would be better than a 14-2 season. That seems too high to me, so I split the difference. I pro-rated short seasons by the average of 16 and the number of games played, divided by the number of games played. So for Gibbs' 1982 season we'd multiply 7 * (12.5/9) to get +9.7. This puts him right below a 13-3 season, which seems more appropriate to me. Paul Brown in 1955 (9-2-1) would be at +7 wins, and get that number pro-rated to 8.2. Since he won the NFL Championship that season, he gets +16.2 for 1955.
That enabled me to grade every coach, in every season, in NFL history. From there, we just need to get a career ranking. I used the familiar 100/95/90 drop-off rate approach; coaches get 100% credit for their best seasons, 95% for their second best, and so on. For example, here's how John Madden's ten year career with the Raiders looks:
year raw wt final nflg w-l-t 1976 24.9 100 24.9 14 13-1-0 1969 11.2 95 11.8 14 12-1-1 1974 9.6 90 10.7 14 12-2-0 1977 7.3 85 8.6 14 11-3-0 1975 6.9 80 8.6 14 11-3-0 1972 5.6 75 7.5 14 10-3-1 1973 3.8 70 5.4 14 9-4-1 1971 2.8 65 4.3 14 8-4-2 1970 2.6 60 4.3 14 8-4-2 1978 1.1 55 2.0 16 9-7-0 87.9
So in Madden's best year, he was 12 games over .500 in the fourteen game season; that gets pro-rated by 15/14, for a result of +12.9. The Raiders won the Super Bowl that year, so he finished the season with a +24.9 score. In his 10th best season the Raiders were just 2 games over .500 (and in a 16 game season); he gets a raw score of +2 but for calculating his career grade, he gets just just a little over one point.
Here's the list. The coaches are ranked by their score, as calculated above. The "Only+" column eliminates all seasons where the coach finished with a below average records; it treats all of those seasons as .500 seasons. So if you want to give Bill Walsh a mulligan when he went 2-14 his first season, you can do that. I've also included for each coach the number of seasons he was a HC and his score per season for our Vince Lombardi fans. Note that score per season is slightly misleading here, as our built-in scoring system gives less weigh on each successive good year you have.
rk Coach Score Only+ #Sea Score/Sea 1 Don Shula 141.7 141.7 33.0 4.3 2 George Halas 139.4 139.4 39.5 3.5 3 Curly Lambeau 128.9 128.9 32.8 3.9 4 Tom Landry 119.2 119.2 29.0 4.1 5 Chuck Noll 106.4 107.5 23.0 4.6 6 Vince Lombardi 99.8 99.8 10.0 10.0 7 Joe Gibbs 95.8 99.2 16.0 6.0 8 Paul Brown 95.6 95.6 21.0 4.6 9 Bill Belichick 91.5 99.9 14.0 6.5 10 Bud Grant 80.2 82.5 18.0 4.5 11 Steve Owen 78.4 79.0 23.0 3.4 12 Bill Parcells 78.1 82.6 19.0 4.1 13 John Madden 75.7 75.7 10.0 7.6 14 George Seifert 74.1 82.2 11.0 6.7 15 Guy Chamberlin 72.6 76.3 6.0 12.1 16 Dan Reeves 71.2 74.9 22.8 3.1 17 Mike Shanahan 70.5 73.6 15.3 4.6 18 Tony Dungy 70.4 72.0 13.0 5.4 19 Mike Holmgren 69.3 72.5 17.0 4.1 20 Bill Cowher 69.2 72.6 15.0 4.6 21 Bill Walsh 69.1 80.8 10.0 6.9 22 Hank Stram 66.1 72.3 17.0 3.9 23 Marty Schottenheimer 65.0 65.1 20.5 3.2 24 George Allen 63.4 63.4 12.0 5.3 25 Marv Levy 57.9 65.0 16.4 3.5 26 Chuck Knox 55.7 58.0 22.0 2.5 27 Jimmy Conzelman 53.2 62.2 15.0 3.5 28 Weeb Ewbank 53.0 59.6 20.0 2.7 29 Ray Flaherty 52.0 52.0 7.0 7.4 30 Buddy Parker 51.8 56.8 15.0 3.5 31 Mike Ditka 48.5 62.8 14.0 3.5 32 Blanton Collier 47.2 47.2 8.0 5.9 33 Sid Gillman 45.7 52.7 17.0 2.7 34 Tom Flores 45.1 59.0 12.0 3.8 35 Jimmy Johnson 43.2 52.9 9.0 4.8 36 Greasy Neale 42.9 52.0 10.0 4.3 37 Dick Vermeil 41.6 56.1 15.0 2.8 38 Jeff Fisher 40.1 48.1 14.4 2.8 39 Tom Coughlin 40.0 50.4 13.0 3.1 40 Andy Reid 39.1 44.8 10.0 3.9 41 Jim Lee Howell 35.4 35.4 7.0 5.1 42 Don Coryell 33.7 40.7 13.3 2.5 43 Potsy Clark 32.3 36.5 10.0 3.2 44 Jon Gruden 31.0 39.5 11.0 2.8 45 Brian Billick 30.5 38.1 9.0 3.4 46 Jim Mora 29.7 38.7 14.5 2.1 47 Barry Switzer 27.8 31.2 4.0 7.0 48 Lou Saban 26.7 40.3 14.4 1.9 49 Buck Shaw 26.6 33.7 8.0 3.3 50 Dennis Green 26.1 35.9 12.9 2.0 51 Mike Martz 25.9 28.2 5.3 4.9 52 Roy Andrews 24.8 33.5 7.0 3.5 53 Don McCafferty 24.1 27.8 3.4 7.2 54 Mike Tomlin 23.8 23.8 2.0 11.9 55 Elgie Tobin 23.4 23.4 2.0 11.7 56 Hunk Anderson 23.2 27.6 3.5 6.5 57 Luke Johnsos 23.2 27.6 3.5 6.5 58 Red Miller 23.1 23.1 4.0 5.8 59 Ralph Jones 22.5 22.5 3.0 7.5 60 Tommy Hughitt 21.8 21.8 5.0 4.4 61 Wade Phillips 20.9 23.4 7.4 2.8 62 Adam Walsh 20.7 20.7 2.0 10.4 63 John Fox 19.5 24.0 7.0 2.8 64 John Rauch 19.5 30.9 5.0 3.9 65 Mike Sherman 18.2 24.2 6.0 3.0 66 Bobby Ross 17.4 21.0 8.6 2.0 67 Paddy Driscoll 16.8 18.7 5.0 3.4 68 Norm Barry 16.5 17.6 2.0 8.3 69 Lovie Smith 16.0 22.5 5.0 3.2 70 Jack Pardee 15.7 25.7 10.6 1.5 71 Raymond Berry 14.9 19.4 5.5 2.7 72 Hampton Pool 14.8 14.8 2.9 5.1 73 Wally Lemm 14.5 30.2 9.6 1.5 74 Jock Sutherland 13.9 13.9 4.0 3.5 75 Jim Fassel 12.7 21.4 7.0 1.8 76 Allie Sherman 11.9 25.6 8.0 1.5 77 Bum Phillips 11.6 22.6 10.8 1.1 78 Dudley DeGroot 11.5 11.5 2.0 5.8 79 John Robinson 11.3 23.4 9.0 1.3 80 Steve Mariucci 11.3 28.2 8.7 1.3 81 Dick Rauch 11.1 17.5 5.0 2.2 82 Mike Holovak 10.4 22.9 7.7 1.4 83 Joe Schmidt 10.4 15.3 6.0 1.7 84 Lou Rymkus 10.4 12.4 1.4 7.7 85 Jerry Burns 9.5 12.5 6.0 1.6 86 Chuck Fairbanks 9.2 19.1 5.9 1.6 87 Earl Potteiger 8.4 19.2 3.0 2.8 88 Clark Shaughnessy 8.1 8.1 2.0 4.1 89 Ray Malavasi 8.0 20.1 5.9 1.4 90 Ed Weir 8.0 8.0 1.6 5.1 91 Al Davis 7.6 9.5 3.0 2.5 92 Forrest Gregg 7.5 23.5 10.9 0.7 93 Art Shell 7.0 16.9 6.8 1.0 94 Ken Whisenhunt 7.0 7.0 2.0 3.5 95 Mike McCarthy 6.4 10.0 3.0 2.1 96 Ron Meyer 6.3 10.6 7.0 0.9 97 Jack Del Rio 6.2 15.5 6.0 1.0 98 John Harbaugh 6.0 6.0 1.0 6.0 99 Mike Smith 6.0 6.0 1.0 6.0 100 Tony Sparano 6.0 6.0 1.0 6.0
This entry was posted on Wednesday, May 27th, 2009 at 7:14 am and is filed under Best/Worst Ever. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.
You seem to have forgotten Dick Jauron.
Chase,
With your 100/95/90 system for players,that's okay--most GREAT players don't last 15 years. With the coaches at the top, they have 30+ yrs. Do they get ZERO credit for those years, or does the drop-off tail off toward the end?
Either way, I think that it is a great disservice toward these coaches. If you stay around for 20+ yrs. as a coach--IN THE SAME PLACE, NO LESS--you are a GREAT coach.
Please redo this list with something like this percentage drop-off, as coaches do not lose "athletic" ability like a player would, nor do they suffer debilitating injuries that can cut a career short.
My %= 5yrs at 100%, then cut off 2% for every subsequent yr.
Re: playoffs/div. champ/SB's: I would say +1 for a division champ--as it means the coach's team DIDN'T make the playoffs as a WILDCARD. In the bygone days where teams went directly to the championship game, they would get no credit because they already get some for advancing to the championship game.
Love the entry--I just think that when Shula makes 4.3 points per season and Lombardi has 10 pps, something looks wrong. Guy Chamberlin (WHO????) should not be so high. I mean, obviously he was a pretty good HC for 6 yrs, but he is a HUUUUGE outlier amongst your top 26. If you showed me a list and said, "I've got a metric that shows these 25 guys (in no particular order) are the best 25 guys of all time," (your top-26 MINUS Chamberlin), I would reply, "That looks like a pretty good list to me."
To sum up--your weighting system has SEVERELY overrated a few guys who had a few good seasons compared to men who were good/great for a LOOONG time.
http://www.pro-football-reference.com/coaches/ChamGu0.htm
Guy Chamberlain coached in the '20s, so I'm sure almost no one has heard of him. But he won four championships in six seasons with multiple teams. And he's in the HOF.
Personally, I think points per season is not a good metric; I just knew someone would ask for it in the comments, so I preemptively put it in there. But Shula ranks #1 on the list, so I don't think you can argue that he doesn't look good by my metric.
I think if we gave out points for division titles, it would overrate some of the current/more recent coaches; division titles are beginning to become more meaningless since we now have more 9-7 or even 8-8 teams finishing on top of their respective divisions. Why reward mediocrity? If anything, I think it is more fair to give credit for 10+ or 11+ wins in a year. Either that, or incorporate some kind of "strength-of-division" or "strength-of-conference" rating/coefficient; i.e. Belichick deserves more credit for 2008's 11-win team that missed the playoffs than Norv Turner does for an 8-win team that won the atrocious AFC West.
As for the final list, it does a good job of reminding us how good Bud Grant was.
Could you put up a ranking that doesn't decrease the value of successive seasons? I'm curious to where Dungy ranks without the decreasing values, he seems like the kind of guy that would lose the most in that system, lots of similar, but very successful seasons, only one losing season.
I refuse to believe any list that doesn't have Wayne Fontes in it.
Ok, now that I've stopped laughing.... shouldn't this system be based on some sort of power rating formula that takes into account the natural regression to the mean over time? Let's say that in Year Y+1 a team's normal power rating is 0.65 * Rating(Year Y). So a coach could be graded against his expectations.
If team A is worth +10 in Year Y then they should normally be worth +6.5 in Year Y+1. If they perform at a +10, the coach would get a rating of +3.5.
Ratings like this would help put guys like Bill Walsh in perspective. And you could use it to examine the guys at the other end of the scale. Who is the worst coach? Rich Kotite? Les Steckel? Rod Marinelli?
Follow-up questions regarding Mr. Chamberlin:
1--Why didn't he coach anymore? He was really young.
2--Is he really getting 8 points (times 4) for those NFL championships? No wonder he ranks so high. Personally, I think you have to devalue some of those championships. I mean, IIRC, they didn't even have CHAMPIONSHIP GAMES!!! If you only give him 4 ppchamp, it looks like he would drop to around #25--which would still give him his due.
I agree with #5 Shake--what would unweighted totals look like (or, if you choose, my weighted version)?
Add me to the list of those who doesn't feel a need for the 100/95/90 factor. Since you are using net wins, I don't think we really need to worry about guys climbing the list due to being compilers. You have to be over .500 to improve your score, so if you are able to do that for 20+ years, I think that is a reflection of your coaching ability - your ability to adapt to changing rules and trends.
I also think that the +7/+5 bonus for Super Bowls is a little high. Look at a guy like Andy Reid. He's been to something like 6 NFC championship games, but only won one of those so only ends up with a +5 for his career.
What if instead, you went with net wins during the playoffs and then multiply it by a factor. Not sure if that should be 2, 3 or 4.
Joseph, before 1958, I think a lot of people associated with the NFL were basically weekend warriors. It was a hobby they did, but was not a career. Not sure if it applied to coaches the same way it did to players. In the movie "Leatherheads" there was a total of one good scene, and it was one where somebody asked the star college player (in the 30's) about playing pro football and he laughed as if to imply that pro football was "beneath" a star football player.
A guy like Chamberlin could very easily have decided that he was wasting his years as an NFL coach, and needed to get on with a real life.
Anyway, just speculation on my part.
Wikipedia says Chamberlin retired at age 34 and he became a farmer, state livestock inspector, and businessman.
He was a player-coach for 6 seasons and then a coach only for his 7th season.
I'm guessing he made more money as a farmer than a player-coach.
sorry for the multi-posts, but I love learning about the real old days of the NFL. Also from Wikipedia:
"The Yellow Jackets had a hand in the 1925 NFL Championship controversy. A dispute arose over a game that the nearby Pottsville Maroons had played against the Notre Dame All-Stars in Philadelphia; the Yellow Jackets asserted that their nearby rivals had infringed on their territorial rights by playing the game (against a non-league opponent, no less) in Philadelphia. The league agreed and suspended the Maroons, allowing the Chicago Cardinals to win the 1925 title. Frankford won the 1926 title the next year with a 14-1-2 record."
Poster #6: I did that sort of thing here.
Numbers don't lie, but Walsh semms to be awfully low... Maybe we've just romantified him for comming up with that passheavy attack that's so "in" right now...
WOOHOO!!! Shula's at the top and Sparano's already in the top 100 after just one season! That makes this list good enough for me
DD-f, Walsh is in his proper place because he was a head coach for 10 years (a branch of the Paul Brown coaching tree, too), while others on the list held the position of head coach for way longer (Norv Turner is about to enter Year 11 as a head coach if not mistaken). Walsh made the most of his time by installing a "finesse" West Coast offense in a time when power/smashmouth was the supposed standard and won three Super Bowls; there is something to be said for that. Today, I would say 60-70% of the teams use some version of Walsh's offense.
All of that praise for Walsh aside, I still believe he is in his right place (though I do question Seifert's standing a bit.
Seifert > Walsh! WTF!? I don't care what you have to do to the formula to fix that anomaly but it needs to be done before I can take it seriously.
Maybe you can factor in how much better you were than your immediate predecessor. Don't know how you work that for the guys who were the first head coach of a franchise, maybe take the previous record as .500 or a shade less
That would boost the coaches that took over losers & made them win while penalizing the coaches that inherited a winner & made them lose.
Megamanic,
This doesn't say Seifert > walsh. It says Seifert has a better record than Walsh. And considering Seifert won 77% of his games in San Francisco while Walsh won "only" 61% of his games, that's a pretty reasonable position.
Now you could say that Super Bowls should be weighted heavier, but I think they're getting a pretty big weight here. And I think outside of the Super Bowls, it's pretty obvious that Seifert has the better coaching record than Walsh. That said, nowhere in this post did I indicate that every coach with a better coaching record than another coach indicates that the former coach was better. In fact, I'm positive that is not the case. Seifert and Walsh is a good example of that.
I think just about any rating system is going to wind up overrating Seifert, unless he gets severely penalized for his time in Carolina. I mean, the guy won 10+ games in every season in San Francisco. He won 10+ games with leading rushers such as Dexter Carter, Terry Kirby, Derrick Loville and Keith Henderson. Yeah, he was the beneficiary of a system and culture that Bill Walsh installed, and the beneficiary of an all-timer like Jerry Rice, but he must have been doing something right.
For Richie above--not sure why 1958 is the cutoff year for Pro Football becoming more of a F/T profession for the players (a "hobby" as you put it). Its good to note that in the distant past it was probably more of an avocation for players than a genuine career as we understand it now. I would say starting after WWII, Pro Football (the NFL and the AAFC) rapidly made being a player more of a career as the game's posterity rapidly increased. Certainly players back then did not train year round, but much of that did not happen until the end of the 60's.
For all readers, remember, Chase is listing the best coaching records, not the best coaches. You can have a great record (see Seifert) and not necessarily be considered great (see Seifert again).
I get it w/Seifert...it is just funny to see him so high up when (to me) it just didn't feel that way.
Tim, 1958 is likely Richie's cutoff because the '58 title game (referred to as the Greatest Game Ever Played) was pivotal in reaching out to the casual fan. Because of that game, football on television laid the groundwork for the players and coaches to get more exposure and therefore bigger paychecks. Without that game being as tight as it was, going into OT, the NFL would have been on life support and not the juggernaut that it is today. Read Mark Bowden's book about this game and how important it was to the league...it is compelling stuff and it made me appreciate the evolution of the game more. (Sorry for the tangent...)
Tim, I was born in 1972, so I what I know about the NFL before about 1980 is what I've read or heard. I used 1958 as a rough cutoff for the modern game, because that is just the sense I have acquired for when the NFL became more of a big time sport. I think the Bal-NY championship game in 1958 really changed the sport in the eyes of the public. As I understand it, the top 4 spectator sports in the early 50's were: baseball, College Football, Boxing and Horse Racing. But I think by the 1960's that changed quite a bit.
Art Donovan played from 1950-1961, and I think I remember hearing him say that he usually had another job in the offseason.
How are you counting Paul Brown's AAFC wins? The same way as AFL wins?
AFL seasons are counted fully (except for their championship game); AAFC seasons are excluded. This mirrors the official stance the NFL takes.
"It says Seifert has a better record than Walsh. And considering Seifert won 77% of his games in San Francisco while Walsh won “only” 61% of his games, that’s a pretty reasonable position."
Sure, it's a reasonable position if you just want to use winning percentage. But since your article was trying to avoid that, using winning percentage is an awfully strange response.
Bill Walsh never lost 15 straight games and didn't go 14-2 and win a SB with the team he inherited in his first year or help run a franchise into the ground at the end of his career, unlike Seifert. So I'd say it's a pretty reasonable position to say he didn't have actually a better record than Walsh. Walsh took over an awful team with little talent, so of course his winning percentage isn't going to be extremely high. Seifert took over a SB winning team that was stocked with HOFers, so of course he was going to have an extremely high winning percentage. Perhaps a fairer comparison is in order.
After winning their first SB, SF record only:
Walsh: .683
Seifert: .697
Chase,
I would like to repeat my request for the raw (unadjusted) numbers. Sure, it will reward "compilers", but the reason they have so many seasons is that they were really good coaches.
Joseph,
Unfortunately, I'm very busy this week but I will try to respond to yours and all other comments this weekend. Thanks for the reminder.
Chase
I'm very sorry for my tardiness in replying. It was a busy week. Let's get started.
Joseph, yes, I was giving coaches zero credit for every positive season after their 20 best seasons. So if Shula had 21 seasons with a winning record (and of course he did), he doesn't get credit for those. I understand you're upset about Shula/Halas/Lambeau/Landry, saying that they are GREAT coaches, but they rank 1-2-3-4 in my system. My system certainly recognizes their records as great. And if you're unhappy with their per season marks, I understand that, but once again -- I would not rank these coaches by their per-season grades.
That said, here are the unadjusted numbers meaning each season is given 100% weight:
The first line says it all...
Greatest coaching records is the same as greatest coaches.
A coach with a crappy record is not gonna be great and a coach with heaps of wins is not gonna be crappy.
You can spin it anyway you want (sort of like Bill James and sterilize the game), but wins equal greatness no matter how you want to stat it out...
Dave, all wins are not created equal.
"Winning percentage works in theory but it tends to underrate ... those who took over bad teams (and were bad at first but ultimately built those teams into top contenders) ... My solution is a formula that incorporates all of those things..."
I hate to cheaply carp when you guys do such great work pulling these analyses together, but I don't see how your formula compensates for the coach taking over a great/terrible team. Two examples:
Weeb Ewbank took over two awful organizations, expansion quality teams (the Jets were arguably sub-expansion by NFL standards, being the inherited ruins of the collapsed AFL Titans franchise). He built them both into world champions -- did it *twice* -- and won arguably the *two* most historically important championship games in NFL history, one with each team.
But he finished with a barely .500 career winning record because his early records with the (lack of) talent he inherited
offset his later-year records, which clearly drops him down in your rankings. You have him ranked below one of his assistants, Chuck Knox, who never won anything important but was plus 30 W-L over 22 years with good organizations. Chuck was entirely competent, but he's not going to the HoF where Weeb is.
George Seifert is ranked ahead of Bill Walsh. Walsh took over a league bottom team and built it into a historic championship organization. Seifert inherited that organization when it was at its top. Who thinks Seifert was better than Walsh?
Now I know you say you are ranking W-L records, not coaches by quality, and that's fine. By W-L Weeb was .500, Chuck was >.500 and Seifert was higher than Walsh.
BUT you also say your formula adjusts for "those who took over bad teams, (and were bad at first but ultimately built those teams into top contenders)", and presumably the inverse -- and that, by these examples and others, I do not see.
Don't feel back for criticizing, Jim. We always need more of that.
As for how my formula helps compensate for building a team, consider two examples:
1) A guy takes over a good team, and goes 10-6 for four straight years.
2) A guy takes over a crap team, goes 4-12, then 7-9, then 10-6, then 13-3.
If we just added the numbers up, the first guy would be +8 for the four years; the second guy would be -8, -2, +4, +10. He would be +4 for the four years. But by using the descending weights and ranking seasons from best to worst, that 13-3 season will always be worth a lot and that 4-12 season won't be worth much if the guy coaches ten years. Assume that both coaches go 8-8 over the next six seasons. The lucky coach (who took over the good team) will have a weight adjusted score of 7.4; the unlucky coach who took over the bad team will have a weight adjusted score of 8.2
As for Weeb, even as a Jets fan, I recognize that he's not exactly in the pantheon of great coaches. Yes, he won three titles, so if you want to put him over everyone who never won a title, I won't argue with you. But he got fired from Baltimore after going just 21-19 over a three year stretch on a very talented team. As someone who has studied Ewbank's career a bit, I feel safe in saying no one ever considered him a coaching great; he was not a genius behind the curtain. Who knew when to get out of the way (a valuable trait that some coaches never learn) and he had those sorts of intangibles that a coach needs to win. But I think his "building of two championships" is mostly because he got surrounded by terrific talent. Of course, maybe that's because he's a great coach -- I don't know. As I said in the beginnning, I don't feel competent in ranking the coaches. Ranking coaches is really hard.
Walsh doesn't rank too high because he only coached ten seasons. I don't think Seifert is inappropriately ranked.
Tony, I agree that giving out division title bonuses would lead us down a bad path. A coach who gets to 10 wins should get more credit, IMO, than one who gets to 9 wins, even if only the latter won the division title.
I think incorporating a SOS into this would be interesting, and perhaps do-able.
Shake,
You're right. As you can see, Dungy moved up to #12 on that list (post #27).
Chase, thanks for the list. Personally, I like the second list better than the first. I don't agree with the per-season ranking either--"small sample size", esp. for Chamberlin & Lombardi, would make them look like they were head and shoulders above the rest. They were in their era, but not overall. Probably the "only + seasons" list would work for Jim Glass (#30). If you wanted to post a revised list with the factors I listed in post #2 or something similar, I would be fine with that also. I just think that Shula (for example) ought to get some credit for ALL of his winning seasons, not just his 20 best. Your method helps those who had about 17-18 winning seasons (with more than 20 overall), but hurts the top 4. In fact, the + yrs. only might be the best list of all, as it discounts the years some of the best coaches took to pull a bad team up. (Landry, Noll, Walsh, & Stram, for examples.) Maybe a list where you discount the coach's first 2/3 yrs of his career, IF AND ONLY IF he was pulling a bad team up to respectability. If he has a 1-15 season AFTER his first couple of years (cough-Seifert-cough), he has to take full credit (blame) for it. In Lombardi's case, for example, he would still get all his credit due. Just another idea.
Thanks, Dungy has had a HOF career and the records reflect that. Not just the amazing run with the Colts, or his contribution to racial diversity in coaching, but those two, PLUS turning the Bucs from perennial laughingstocks with 3 winning seasons in their 20 year history and none in 13 years before him into a regular playoff contender after just one losing season (his only losing season as a HC, at 6-10, which was actually still higher than the Bucs franchise win% at the time).
"As for Weeb, even as a Jets fan, I recognize that he’s not exactly in the pantheon of great coaches.... I feel safe in saying no one ever considered him a coaching great"
Ewbank is in the Hall of Fame, so quite a few people considered him a coaching great and he actually *is* in the pantheon of great coaches. Whether or not he should be there is a different argument.
"Walsh doesn’t rank too high because he only coached ten seasons."
So did Lombardi and Madden. Of course Lombardi had 5 titles, so he's going to rank higher on this, as well he should. Madden had a great regular season record, but he took over a very good team and 'only' won one title.
BTW Seifert coached for 11 years. So I guess 1 more year would have made the difference for Walsh. Pity.
Chase, thank you for the additional info. It looks like the Top 10 mostly remains the same, except for Madden making the leap into the 10-slot, knocking Bud Grant down to 11. Madden was a really young coach for that era and somehow meshed a lot of flaky personalities into a consistently good team; he deserves a lot of credit.
Eagle: BTW Seifert coached for 11 years. So I guess 1 more year would have made the difference for Walsh. Pity.
Considering the year he would have gotten (14-2, Super Bowl Champion), that seems self-evident.
Walsh isn't the one who suffers most from bad early years. That'd be Parcells, who has taken over crappy teams and made them good on multiple occasions.