Salazar's 2:08:13, which was probably 2:08:40-45 on an accurately measured course, was run in New York. It gets overlooked because of the reports of the course being short, but it may have been a performance to rival Dinsamo's. Today we would laugh at someone attempting a world record in New York, so I suspect it would have been well under Clayton's world record on a course more suited to fast times.
This post by HRE summarizes many of the reasons - really the lack of prize money, particularly for time bonuses, led to many top athletes not running all out if the win was in hand. In both 1977 and 1978 Rodgers won NYC in "modest" times (2:11 and 2:12) knowing that Fukuoka was right around the corner - in 1977 he won that too. The only time Rodgers ran a pseudo-paced marathon was the 2:08 in Japan against relay teams on a course that was found to be 200m short, plus that random 1 hour race during the summer on the track where he would have set an AR in the half marathon had he simply run one more lap (He did 12.9 miles in an hour). As HRE stated, Greg Meyer has been very frank that he held back at Boston in 1983 despite extremely favorable conditions.
Yes, it was very weird running USATF-NE XC in the early 2000s and seeing Dinsamo there with the "Athletics East" green singlet.
The typical marathon world record gets lowered every 1-3 years or so. This applies before 1988 and after 1998.
The exception was Belayneh Dinsamo who held the record at 2.06.50 for 10 years between these dates.
Presumably this is a function of introducing out of competition random drug testing.
But why this event impacted to this extent? Did it take that much time to develop the next range of tools to beat the testers?
$$$$
Appearance fee$ and prize monie$ upped the ante. Before that, you didn't $TART running marathon$ until you were in your mid 30$ and moved off the track.
As time went on and no one ran faster than Clayton there was more and more speculation that the Antwerp course was short. Pretty sure in Ron Hill's autobiography he says something about beating Clayton who was the world record holder in the marathon. Many years later I read an interview where he said he, Ron, was probably the world record holder in the marathon at one time because Clayton's time came on a short course.
That would mean that the record before Clayton's 2:08 was his 2:09:37 from 1967, then Ron's 2:09:28, then Ian Thompson's 2:09:12 in 1974, then Gerard Nijboer's 2:09:01 in 1980. So we have a bit more than a minute and a half improvement over 22 years and then about three minutes over the next eight years, then as the OP says, nothing for the next ten years and the next improvement is well into the post EPO years.
So in the 27 years since DeCosta broke Densamo's record (thanks to whomever did the spelling correction) the record has come down about twice as much as in the 31 years between Clayton and Densamo. A lot is different in the marathon now than it was much before DeCoasta's record and all of those differences have pushed times faster. I would never put all the improvement down to PEDs but I think it would be incredibly naive to deny they are a big factor.
You're missing Shigeru So's 2:09:06 at Beppu in 1978. That should have been the WR until Nijboer's run.
There was an excellent investigation of the route using modern methods of measurement, including using older maps for intersections that have since been changed by an internet sleuth that I cannot find at the moment, but reading the methodology convinced me that the course was indeed short and his 2:08:33 should not count. If Alberto's 2:08:13 doesn't count as well, then really the progression from Clayton through these great runners, down from 2:09:37 to 2:09:01 from 1967 to 1980 makes a ton of sense and makes that next jump, to DeCastella's 2:08:18 not only seem reasonable but the next logical step. Shoe technology was getting better and better from the late 1970's on, and as the money moved in, the jumps to Jones and Lopes also seem natural and inevitable.
As time went on and no one ran faster than Clayton there was more and more speculation that the Antwerp course was short. Pretty sure in Ron Hill's autobiography he says something about beating Clayton who was the world record holder in the marathon. Many years later I read an interview where he said he, Ron, was probably the world record holder in the marathon at one time because Clayton's time came on a short course.
That would mean that the record before Clayton's 2:08 was his 2:09:37 from 1967, then Ron's 2:09:28, then Ian Thompson's 2:09:12 in 1974, then Gerard Nijboer's 2:09:01 in 1980. So we have a bit more than a minute and a half improvement over 22 years and then about three minutes over the next eight years, then as the OP says, nothing for the next ten years and the next improvement is well into the post EPO years.
So in the 27 years since DeCosta broke Densamo's record (thanks to whomever did the spelling correction) the record has come down about twice as much as in the 31 years between Clayton and Densamo. A lot is different in the marathon now than it was much before DeCoasta's record and all of those differences have pushed times faster. I would never put all the improvement down to PEDs but I think it would be incredibly naive to deny they are a big factor.
You're missing Shigeru So's 2:09:06 at Beppu in 1978. That should have been the WR until Nijboer's run.
There was an excellent investigation of the route using modern methods of measurement, including using older maps for intersections that have since been changed by an internet sleuth that I cannot find at the moment, but reading the methodology convinced me that the course was indeed short and his 2:08:33 should not count. If Alberto's 2:08:13 doesn't count as well, then really the progression from Clayton through these great runners, down from 2:09:37 to 2:09:01 from 1967 to 1980 makes a ton of sense and makes that next jump, to DeCastella's 2:08:18 not only seem reasonable but the next logical step. Shoe technology was getting better and better from the late 1970's on, and as the money moved in, the jumps to Jones and Lopes also seem natural and inevitable.
Someone else pointed out that I missed So. But thanks.
I think we read the same article about using modern measurement methods to remeasure the Antwerp course and finding it short. I just don't know enough about how those methods work and how reliable they are but it made a very good case. On the other hand Clayton has said the course was measured again a day or two after his race and found to be about 100 yards long. If true I'd be more inclined to believe a measurement done within a couple days of the race over the actual course than an analysis done fifty years later on possibly the actual course.
But yes, I'm a little more inclined to believe the course was short than I had been but one thing that made me believe it was accurate originally was looking at complete results. I can't find them again but no one else in the race ran a big PB. In fact I'm not sure anyone got a PB but Clayton. I've been in races on short courses and you usually get a LOT of PBs.