≡ Menu

Worst Coaching Regimes

With another ugly loss, Dennis Allen’s record as head coach of the Raiders has dropped to 8-28. But does this mean Allen’s tenure as Oakland head coach has been one of the worst 10 coaching regimes since the merger?

Not exactly. For starters, we should remember that Allen was dealt a terrible hand. The year before Allen’s arrival, 2011, Oakland didn’t have a first round pick. He inherited one of the worst rosters in the NFL, and didn’t have a first or a second round pick in his first year. In 2013, the Raiders spent only $67M on the players on their roster, courtesy of $50M of dead money on the team’s salary cap. So an 8-28 record, while perhaps not even good considering the circumstances, is hardly all Allen’s fault.

That said, I thought it would be fun to just compare Allen’s record to that of other regimes since the merger, regardless of circumstances. The most common way to do this would be to use straight winning percentage, but that would put Allen behind say, Cam Cameron, who went 1-15 as the Dolphins head coach.

Another method could be to use games under .500 — Cameron would therefore be 14 games below .500, while Allen would be 20 games below. But Jim Schwartz finished 22 games below .500 with the Lions, courtesy of a 29-51 record.  Your mileage may vary, but to me, an 8-28 record is worse than 1-15 and 29-51; the former could be disregarded as just one terrible year, while the latter was much better on a per-game basis.

One solution would be to add a number of games of 0.500 football to each coach’s record. The devil is in the details, here — what number should we pick? — but let’s say we added 20 wins and 20 losses to each team’s record. That would make Cameron 21-35 (0.375), Schwartz 49-71 (0.408), and Allen 28-48 (0.368). That “works” if the goal is to put Allen below Cameron and Schwartz, and I think as a general rule, adding 40 games of 0.500 football works well enough for these purposes.1

By this methodology, Allen’s tenure in Oakland checks in as the 13th worst since the merger. Here’s how to read the table below, topped by Leeman Bennett and Rich Kotite. Bennett coached the Bucs in ’85 and ’86, while Kotite coached the Jets ten years later. Both went 4-28, for a 0.125 winning percentage, which comes in at 24 games below 0.500. If you add 20 wins and 20 losses to each of their records, you get adjusted winning percentages of 0.333, the metric by which the table is sorted.

RkCoachTmFirst YrLast YrWLTWin%GmU.500Adj Win%
1Leeman Bennetttam1985198642800.125-240.333
1Rich Kotitenyj1995199642800.125-240.333
3Rod Marinellidet20062008103800.208-280.341
3Steve Spagnuoloram20092011103800.208-280.341
5Marty Mornhinwegdet2001200252700.156-220.347
5Chris Palmercle1999200052700.156-220.347
7David Shulacin19921996195200.268-330.351
8Bill Petersonoti1972197311800.053-170.356
9Bill Arnspargernyg1974197672800.2-210.36
10Rod Dowhowerclt1985198652400.172-190.362
11Jim Ringobuf1976197732000.13-170.365
12Dom Capershtx20022005184600.281-280.365
13Dennis Allenrai2012201382800.222-200.368
14John McKaytam19761984448810.335-440.373
15Marion Campbellatl219871989113200.256-210.373
16Cam Cameronmia2007200711500.063-140.375
16Rod Rustnwe1990199011500.063-140.375
18Dick LeBeaucin20002002123300.267-210.376
19J.D. Robertsnor1970197272530.243-180.38
20Dave McGinniscrd20002003174000.298-230.381
21Abe Gibronchi19721974113010.274-190.384
22Joe Bugelcrd19901993204400.313-240.385
23Tom Floressea19921994143400.292-200.386
23Mike Rileysdg19992001143400.292-200.386
25Darryl Rogersdet19851988184000.31-220.388
26Harvey Johnsonbuf21971197111300.071-120.389
26Dick MacPhersonnwe1991199282400.25-160.389
28Ray Perkinstam19871990194100.317-220.39
29Mike Mularkeyjax2012201221400.125-120.393
29Art Shellrai22006200621400.125-120.393
31Hank Bulloughbuf1985198641700.19-130.393
32Frank Kushclt19821984112810.288-170.394
33Ed Bilesoti1981198382300.258-150.394
34Kay Stephensonbuf19831985102600.278-160.395
35Hank Stramnor1976197772100.25-140.397
36Dave Campodal20002002153300.313-180.398
36Mike Ditkanor19971999153300.313-180.398
36Herman Edwardskan20062008153300.313-180.398
36Chuck Knoxram219921994153300.313-180.398
40Marion Campbelltl11974197661900.24-130.4
40Hugh Campbelloti1984198582200.267-140.4
40Mike Nolansfo20052008183700.327-190.4
43Frank Ganszkan1987198882210.274-140.401
44Dennis Ericksonsfo2003200492300.281-140.403
44Mike McCormackclt1980198192300.281-140.403
44Pat Shurmurcle2011201292300.281-140.403
44Norv Turnerrai2004200592300.281-140.403
48Romeo Crennelkan2011201241500.211-110.407
48Richard Williamsontam1990199141500.211-110.407
50Scott Linehanram20062008112500.306-140.408
  • While it’s easy to quibble with the methodology, I think the results are pretty good. It’s hard to top Bennett’s and Kotite’s 4-28 record for pure futility. Marinelli and Spagnuolo essentially went 4-28 with a 6-10 season added on to that, placing them tied for third worst on the list. Mornhinweg and Palmer simply turned one of those losses into wins, so 5-27 records puts them tied for 5th worst. But I’ve included the full data, so you can come up with whatever methodology you desire.
  • You probably won’t be surprised to see Tampa Bay’s original coach, John McKay, finish the most games under .500. A 44-88-1 record left him an unmatched 44 games below 0.500; only David Shula finished more 30 or more games below 0.500 since the merger.
  • Cameron and Rod Rust are the only coaches to ever go 1-15 in their first and only season with a team. That’s good enough for only 16th on this list, though. And it’s not even the worst winning percentage of any coach on the list! In 1971, Bill Peterson went 3-7-1 as the head coach at Rice. Somehow, that got him a job with the local pro football team, and Peterson went 1-13 with the Oilers in 1972. He was fired after an 0-5 start in ’73, giving him a 0.053 career winning percentage in the NFL.2
  • Note that this post looks at coaching regimes, which means continuous coaching eras. So Campbell makes the list after going 11-32 with the Falcons from ’87 to ’89, but he also made the list for his earlier tenure with Atlanta, when he went 6-19 from ’74 to ’76.  If you are wondering how Campbell could possibly get rehired by the Falcons, you are not alone.
  • Two other coaches made the list based on their second stop in a city:
    • Joe Collier was the Bills coach in the late ’60s, but he was fired after the 2nd week of the 1968 season. Harvey Johnson replaced him, and went 1-10-1 as interim coach. Buffalo then hired John Rauch, who coached the team for two years before Buffalo hired…. Johnson to replace him.  In Johnson’s second time circling the wagons in Buffalo, the team went an ugly 1-13, which landed him at #26 on the above list. That career 2-23-1 record, tho.
    • Chuck Knox went 10-4 or better and won the NFC West in each year from 1973 to 1977 with the Rams. But due to conflicts with owner Carroll Rosenbloom, Knox was gone after five years in California. When Knox returned in 1992 at the age of 60, he led Los Angeles to 6-10, 5-11, and 4-12 records before retiring for good. His work in his second tenure with the Rams placed him at #36 on the list.

What stands out to you in the above table? Other than a pair of washed-up Saints coaches appearing at #35 and #36, of course.

  1. What would be a fun post is trying to derive what the right number actually is; we can put that on the offseason to-do list, or perhaps a guest post? []
  2. His one win? That came in a 26-20 victory against the Jets. Classic let down by New York following the historic win at Baltimore a week earlier. []
  • James

    “What would be a fun post is trying to derive what the right number actually is; we can put that on the offseason to-do list, or perhaps a guest post?”

    Tsk tsk tsk, Chase. Not only did Neil Paine already show it’s 11 games of 0.500 ball (so add 5.5 wins and 5.5 losses), but it was in a guest post on this site!

    http://www.footballperspective.com/estimating-nfl-win-probabilities-for-matchups-between-teams-of-various-records/

    • Chase Stuart

      Heh, I didn’t think Neil’s study worked for these purposes for two reasons.

      1) When I used 11 games, it just didn’t make enough of a difference. I would prefer to give more weight to coaches for failed for longer stretches. For example, I’d say 8-24 is worse than 3-11; using 20 games, that is the case, but it does not hold if we drop it to 11 games.

      2) On a more theoretical level, a coach is only responsible for a portion of a team’s record, of course. So while at the team level, it may take “only” 11 games of 0.500 ball to make things equivalent, records say more about the team than they do the coach. As a result, you need to introduce a larger variable. If anything, 20 may be too small.

      • James

        That’s a particularly good point about the difference between teams and coaches, I was overlooking that.

  • JeremyDe

    You know you’ve been to this site too many times when you see the title of this article, click on the title, and immediately do a Ctrl-F for Campbell before reading a single word.

    • Yes, it makes me pretty happy to see the Swamp Fox getting mentioned in a coaching post. He needs to find his way into non-coaching posts, too. Posts on either the 1960 Eagles or high fashion would work.

    • I just cackled madly. And threw a spoon against a window.

  • Richie

    There was at least 1 bad regime in each season between 1970 and 2013.

    1985 had the most bad regimes at the same time – with 6. Bennett, Dowhower, Rogers, Bullough, Stephenson and Campbell.

  • RustyHilgerReborn

    Not one, but two Detroit coaches in the top(bottom?) 5. Millen’ed!!!

    The Detroit sports media still talks about how Darryl Rodgers, after starting 2-9 in the ’88 season, said out loud during the postgame press conference: “What does it take for a guy to get fired around here?”

  • RustyHilgerReborn

    On Football Outsiders a few years ago, Mike Tanier had a fascinating article on group of coaches he dubbed “The Jauron Gang”. Coaches that had one winning season that made people think they were good, but then sucked the rest of their (usually short) careers.
    It was written right after Todd Haley was fired from Kansas City, and when Raheem Morris was about to be fired by Tampa Bay. I think that would be a fascinating study for the next article (The Lions rear their ugly heads once again, as Jim Schwartz is the latest member of the Jauron Gang).

  • I think it would be interesting to do a similar study but based on SRS, Pyth2, DVOA (Andreas’ EDVOA for older teams), or some other granular measure. Maybe even find a way to measure wins relative to expectation, so you can try to mollify the Parcells Effect (taking advantage of regression to the mean). Andrew might already be working on this…not sure.

    • Chase Stuart
      • I suppose the problem with having nearly 1000 articles is that I am bound to forget a few.

        Speaking of Tony Dungy: does it (should it) hurt his Hall of Fame chances that he only won one title with Peyton Manning as his quarterback? I have never heard of any reasonable person claiming John Fox should be in Canton, but he has turned 2.1875 years with Manning into a 28-7 record and a Super Bowl loss. Even Jim Caldwell made it to the big game with Manning at the helm.

        This reminds me of Adam’s view of Eddie DeBartolo, Jr. making the HOF. Just replace “Hired Bill Walsh,” with “Coached Peyton Manning.”

        I am neither anti-Christian nor a racist. (Just in case anyone from the ESPN message boards found their way to your site).

        • Chase Stuart

          It’s an interesting point, but then again, consider Andrew Healy’s work:

          http://www.footballperspective.com/ranking-the-almost-dynasties/

          http://www.footballperspective.com/ranking-the-almost-dynasties-part-ii/

          Those articles suggest Indianapolis overachieved by even winning 1 Super Bowl! And that includes a couple of post-Dungy years. Is 1 title in 7 years underachieving? Recall also the Schottenheimer Index: http://www.footballperspective.com/the-schottenheimer-index/

          That said that over the course of Dungy’s playoff career, based on the point spreads of each game, he had 1 win fewer than average. And my guess is that if you look just as his Indy time, it would be positive.

          • There’s a lot of chicken and egg there, and I’m not sure how to account for that. Sounds like an off-season project to me.

        • Kibbles

          Bryan, I don’t think you can boil Tony Dungy’s credentials entirely down to “won a title with Peyton Manning” like you can with DeBartolo’s and “hired Bill Walsh”. It’s hard to stress enough just how atrocious Tampa Bay was before they hired Dungy. They had two winning seasons in their entire history before Dungy came in and brought them four more in his six years. He made the NFCCG with Shaun King. While there would be some debate over the provenance, he and Kiffin popularized a major defensive scheme. And Peyton Manning or no, 6 straight 12-win seasons is an amazing accomplishment.

          Plus, Tony Dungy suffers from a bit of a double-standard. For most great coaches, it is widely accepted that their system is responsible for some sort of residual success even after they leave. George Seifert won two Super Bowls with San Francisco, but many give Bill Walsh much (most) of the credit for them. Same with Barry Switzer, Jimmy Johnson, and the Dallas Cowboys. And John Gruden got all kinds of credit for the success of his Oakland Raiders in 2002 under Bill Callahan. But somehow the fact that Dungy’s Bucs won a Super Bowl the year after he left stands as a mark against him, rather than a bullet point on his resume.

          Additionally, it’s pretty universally acknowledged that coaches take a little bit of time to put their stamp on a team. Bill Parcells is famous for what he did in year 2 with franchises, for example.

          In fact, since the conventional wisdom seems to be that it takes a year for a coach to really get his system in place, and that system hangs around for a year after the coach leaves, we could create some sort of “1-year-lagged coaching record”. Setting aside for a moment the appropriateness of such a measure, and just considering what it might look like… Tony Dungy’s “1-year-lagged” record is absolutely insane. In Tampa, Dungy was “responsible” for a 60-36 record, an NFCCG appearance, and a Super Bowl victory. In Indy, Dungy was responsible for an 89-23 record (!!!), seven straight 12-win seasons, a second SB victory, and a third SB appearance. Overall, he was responsible for 12 playoff appearances in 13 years and just the fourth 70+% winning percentage by a coach with 100 games (joining HoFers Madden, Lombardi, and Allen). “One-year-lagged Tony Dungy” would be a slam-dunk Hall of Famer and, in my opinion, one of the top-5 coaches of all time.

          Again, not really discussing the appropriateness of using a simple one-year-lagged record. Just illustrating the Tony Dungy double-standard.

          • All great points, and I am fascinated by the idea of a staggered record. The hope of getting ideas like this is why I ask the questions I ask. I honestly didn’t realize that people used that double standard on Dungy. Everyone I know always says Gruden won a Super Bowl with Dungy’s team. I do find it interesting that Tampa’s best year defensively was the year right after Dungy left, similar to how Montana’s best year came the year after Walsh left. Non sequitur, I know, but interesting nonetheless.

    • I am working on this! I’m going to write up some more about this stuff pretty soon and then maybe the next post will be the complete model.

      Love the “Parcells effect,” by the way. Nobody was better at taking over the hopeless than him. The coaches he replaced all appear on Chase’s list: Ray Perkins (#28), Dick MacPherson (#26), Rich Kotite (#1), and Dave Campo (#36).

  • Joe Collier has one of the all-time weird coaching careers. He became a head coach at 34 never having served in a higher capacity than a position coach before, had that awe-inspiring three-year run as the head coach in Buffalo, and then spent the next two decades with the Broncos. He was the defensive backs coach for three years before taking over as the defensive coordinator for another 17, working under FIVE different head coaches.

    And after that, he actually got fired–he had been with the team for 20 years and got fired. He had been the Broncos defensive coordinator for six years longer than Dick LeBeau has been in that position with the Steelers, and he got fired. And then he didn’t go away! After two years out of the league, he returned as the Patriots defensive coordinator for two years.

  • With a 1-24 record as Browns head coach, Hue Jackson would be 21-44 if you added 20 wins and 20 losses to his record. That’s a .323 winning percentage, which would set a new record for the worst regime ever.