Tuesday, 15 February 2011

Some thoughts on race tactics

The photo shows Cavin Woodward at the London end of the London to Brighton route in 1975, the year he won the race in 5hrs 12min 7 secs (52miles 1172 yds). This was one of his 12 Brighton finishes spanning the years 1971 to 1998.
I've always been a firm believer that the best pacing for a race up to a 100k is a negative split (second half quicker than the first half), assuming the course has equally hard halves. But I doubt there's many 100 mile races run with anyone getting a negative split, so what's the best tactic? And should it vary depending on whether you're trying to win or to beat the cut-offs? I'm not sure I have any answer, but here's some of my thoughts, prompted by a very unusual article I read.

Well, I was shown this article on a 100-mile track race in 1975 and it certainly made me think again: http://www.ultralegends.com/the-tipton-100-miles/.

In the race described, Cavin Woodward (who unfortunately passed away in early 2010) set the world best time at that point for 100 miles in 11:38, but it's the way he did it that amazed me. He ran 2:31 for the marathon, 3:01 for 50k, a world best for 50 miles (4:58) and a world best for 100k (6:25)! That would have been fast enough to win the World Championship 100k most years nowadays...easily. And he held on to break the 100 mile record in a top class field where five other guys broke 13 hours. That just sounds like the craziest race I've ever heard of. I mean, imagine if Geoff Roes had got to half way in last year's Western States in 6 hours then held on for the 15:07 course record ("CR"), since that's about the equivalent.

So I wondered whether Cavin Woodward has any lessons for others or if his time was a combination of him being mentally like granite as well as maybe a bit lucky that it didn't all go wrong. Should people just go out hard and assume that the pace will drop off? This is the opposite of the standard advice of going out easy (since the pace will still tend to drop off).

I'd guess that anyone, even at the elite end, who tries to replicate Cavin's tactics would probably not finish their 100 miler or would crawl in near the back of the field. But perhaps going out a bit harder than might seem sensible is actually the best idea for the elite runners (although slower runners are likely to be less well trained and so not have such high levels of endurance to pull this off, in general)?

I say this after having had some emails going back and forth with Eric Clifton, the man whose 15-year old RR100 record I was fortunate enough to break. Eric is well known for being an all-or-nothing runner with many, many CRs which still stand and even told me that he didn't like winning unless he also ran as fast as he thought was possible. As Eric said to me, referring to pacing purely for a win instead of a fast time:

"I have had many, many people tell me 'how you can win [Western States]' and they can't get that is not the way I want to win WS. It would be a hollow victory to me."

I like his mentality and it was this type of tactic which made the 2010 WS race so enthralling. Anton Krupicka and Kilian Journet hammered away at each other for 80 miles before the more evenly paced Geoff Roes overtook for the win (and CR). I'm sure Eric liked the front running, although nobody could say Geoff wasn't motoring too.

So, for those hyper competitive races with hard fought CRs, sometimes it takes this kind of all out effort to push to a new level. But not always...Russian Oleg Kharitonov holds the current world best for a road/track 100 miler (11:28) and he ran evenly (splits of 5:37/5:51). And I'm happy with my even pacing, which has paid off in the shorter ultras and also did at RR100. I don't think I'll be changing it any time soon, except maybe as a one-off experiment at some point, just for fun.

One last thought is to consider the tactics of the legendary Bruce Fordyce, who won Comrades an unprecedented nine times and still holds the world's best time at 50 miles (4:50). He always went off at his own pace and was often a long way behind the leaders, only to come through near the end of the race as the hares slowed. I don't have a quote to hand, but he was known for advising runners at Comrades that if they went out too fast, they'd pay for it later and run a slower overall time. In particular, he said that for every minute a runner goes ahead of their optimal even(ish) pace at half way, they'd lose several minutes in the second half.

I've read Bruce's book and his attitude of running his own race and ignoring the competition is the way I prefer to run. He believed that if he runs his best race personally then it was up to the other runners to beat him. If they started faster and held on, then they deserved the win, but he wouldn't be closer to them if he matched their early moves.


  1. Gee those times of Calvin's make sub 6 hours at Comrades look easy as pie.

    Common Ian lets have a statement that you are going all out at Western States to break the course record. (There is always next year after all if you blow up).

  2. Ian,

    You should read Stuart Mill's blog (http://ultrastu.blogspot.com/). Stuart was the runaway winner of the Lakeland 100 last year; he keeps to the theory that you should run fast while you can. He also stresses the importance of positive thinking, something which appears to be a great strength of yours.

    Interestingly, he too has read the article and writes about it: http://ultrastu.blogspot.com/search/label/100%20mile%20World%20Record


  3. Charlie - much as I'd love to have a perfect WS this year, I don't think the way to do that is to aim for a particular time and spend the day thinking about splits and the CR. I'll do my thing, hope it all holds together and see where I end up compared to the rest of the incredibly talented field. Undoubtedly, there'll be some very fast times.

  4. I think there's a difference depending on whether you're running a flat, non-technical course or a hilly mountain race. Even splits on the flat are, for obvious reasons, much more realistic and you can pace much like you would for a marathon (as demonstrated in Geb's WR http://www.sportsscientists.com/2007/10/haile-gebrselassie-world-record-splits.html or by Kouros in the 24hrs). For a mountain race, I find perceived effort to be a much better gauge of pace since the terrain varies so much it tends to naturally dictate the pace you should run.

  5. Joe, I agree that the terrain/conditions have a big effect, but if the first and second halves are comparable in difficulty, then I'd say it's the same situation as a flat race. If not, then the optimal pace would obviously be adjusted to be slower over the more difficult sections. But should the second half of any 100 miler always be a lot slower than the first half?

  6. I guess what I'm trying to say is that the more tame you make the race environment the easier it is to pace evenly. In a mountain race, you introduce more variables that impact your ability to run a faster second half and which aren't necessarily of concern on a road or track. For instance, its harder to run technical single track during the night than it is during the day but on a road or track it really makes no difference. So I don't necessarily think the second half should always be a lot slower, it just becomes somewhat inevitable the tougher the course.

  7. Joe - I'm not pretending to be an expert on 100s, but have done plenty of shorter ultras where things like daylight don't factor in. I was just publicly musing whether someone should aim to be running (for example) an easy section at mile 10 at a similar pace to an easy section at mile 90. All other things being equal, I'd favor a yes to that scenario, but things are not equal in a typical 100 miler. In fact, since most 100s have very few fast sections, maybe it's worth running those fairly fast because the harder sections will allow a 'rest' of sorts.

  8. Sharmanian. Yes excellant article which can give a lot of information for ultra runners today and how to do things. By the way - thanks for linking to my site. All the best for the future.

    Phil (Ultralegends)

  9. It is interesting looking at Tony's last few hundred milers and trying to determine what tactics he should employ. I have to admit, if I was Tony, I would be very confused.

    Leadville - he goes all out, 2 years in a row, in an effort to break the CR only to wind up semi-concious at 80 miles and DNFing.

    Western States - he goes all out running with, reportedly, the worlds best mountain runner, drops him at 80 miles and goes on to beat the previous course record. Unfortunately he gets run down at 90 miles by someone else and cruely pushed into 2nd place.

    Rocky Racoon - He employs more conservative tactics and runs a sustainable pace with the expectation that all those guys who have gone out really hard will eventually come back, which is what happened, apart from some (virtually) unknown Pom, who blew away the previous course record.

    You have to feel for the guy.

  10. Ian, a very interesting post. I have also read and studied pacing (to a certain extend that is). I have also ready Stuart Mills' blog who seems to be quite successful with going off very fast (unless he gets lost ;-)) since he wins most of his races.
    I know that running a negative split became popular some years ago and it was claimed it was the right way of running a race (I think it was postulated for a Marathon though). But the only reason for that was that often the winner of a Marathon ran a negative split. And even the Marathon world record by Haile Gebrselassie was produced with a negative split. But even that does not provide strong enough material to assume the theory is right.
    For the Marathon world record for example you cannot run a positive split since the pacers would not make it to 35k. And in a city Marathon the negative split is just a natural tactical phenomenon. The faster runner just breaks away eventually (after "waiting" in the pack for as long as needed) and you can of course only break away if you speed up, hence the negative split (not in all cases of course)).

    The problem however is the lack of actual scientific evidence how fast a race should be run (by anyone elite or not).
    The problem with pacing a longer race is that in order to allow a negative (or even or positive) according to plan requires the knowledge of the finishing time in advance. How can split times be calculated otherwise?
    Since no one does have that knowledge (or does one?) we are running on a "believe" base. Mental that is. If you believe you can run the second half faster than the first, then you probably can do that. It is as easy as that.
    Even easier is the other way around of course ;-)

    I know of course that starting a Marathon way too fast (I have tried it out myself) does not pay. But how would we explain that running 1:28/1:32 splits for a 3 hour marathon (for example) is wrong (exercise physiologically). And on the other hand running 1:32/1:28 is right? Maybe 1:30/1:30 is?
    Hypothetically we suggest however, that the runner who just ran 1:32/1:28 would not be able to achieve 3:00 for the Marathon if he started 1:28 for the fist half? How on earth are we going to provide the evidence? No run is ever like the other so we cannot prove our theory?
    I am not saying running a negative split is the wrong way of doing it. You need a race plan though. No question.

    Enought said, I could go on forever. ;-) Interesting subject though.

    Recover well and keep enjoying the races!

  11. Love the post Ian and replies also. It's very timely as I'm just trying to figure out how to run a 12hr race later in the month. Last year I run the same race with a much faster first half and left the field behind who never then caught up despite me slowing down considerably. So going out fast did work. However I think that if I run it more evenly this this that I will overall go even faster. I will report back on what happens in my blog ultradiscostu.blogspot.com

  12. Thomas - I don't think ultras are as precise as marathons, by any means (I consider it a bad race if I drop more than a minute in the 2nd half of a marathon, but that would be hardly any time in an ultra). For up to 100k on a course without any sections that are way harder than the rest of the race, I'd still say that the best pacing is close to even. For Comrades, I'm certain of this and so many guys who are faster than me go off too quickly but then I catch them near the end.

    But for 100 milers, I'd say an even split is within about half an hour for each half. That's still a small proportion of the overall time and reflects that the slow down at the end isn't too significant.

    I just found it fascinating how Cavin could go through intermediate distances in his 100 in world class times. He dropped off significantly, but was still able to jog to the finish. If he'd taken just a bit of fire out of his earlier pace, would he have gone faster overall and still hold the world best for the distance today? I'd guess so, but there's no way to tell.

  13. Great post, and I stand by even pacing. I followed the RR100 via twitter (@crazyrunnerguy) and was floored with your relatively even splits.

    I finished my first 50M last October. It was a 3-loop course - ~16.8/loop. I ran a 2:50; 3:00; 3:14 for a 9:04 finish; I was also at 4:27 at the half-way point. I paced myself on what I thought was possible and based my pace on effort, as discussed above. I set out to run even splits and I feel like I accomplished that goal.

    I stand by even pacing because it feels better to me. I do not have a sudden burst of energy at the start, and then slowly fade to black as the race progresses. This killed me in my first 50K and 50M attempt (finishing the former, DNF'ing the latter). It feels better because my effort is constant.

    Now, I am not a front-pack runner and may never be. But until I am proven otherwise, I will try and run even (or even negative) splits.

  14. Joe, it appears that lack of light at night isn't a factor for Ian even in 100 milers now. :)

  15. It certainly didn't seem common to fear going out too fast in that era. John Tarrant and even Arthur Newton before him were known to put out low 6 hour 50 mile splits in 100 mile races. When Newton ran his first 100 miler he went through the first 10 miles just under 65 minutes. The culture of the acknowledged fade didn't scare these folks. Run strong while you can and try to maintain as best as you can. The distance itself is enough to facilitate slow down over the duration. Newton ran with a ton of confidence in himself while he was running. I don't think he ever led any of the Comrades he won in the first 20 miles or so. His 100 mile efforts were the opposite strategy.

    I always really enjoyed reading about Bruce Fordyce and how he ran. He and guys like Barney Klecker seemed to understand that you could run right on that razors edge and had the confidence and the strength to see it through, rather than worry about what the competition around you was doing.

    In the US, most folks associate Ultrarunning with Trail running, so perceived effort is much more of a consideration rather than doing what Eric has always done. I think part of why he was great is that it was all racing, to him.

  16. I am Cavins son and I can say without fear of contradiction that if he had started slower he would have finished slower. The thing about ultras is that they are more a psychological challenge than marathon or 10k. You have to be able to literally endure. Running at any pace for 6 hours you will experience highs and lows. My father just hit those lows with momentum.

    What would have given him a chance of still holding those records would have been better physical preparation. He was a smoker, home brew beer drinker and trained only 80 miles per week with no stretching, no speed work and no conditioning.

    If you are a mid pack runner then run even splits and enjoy your run. but if you are a racer then race to your limits. Some days you will exceed your limits and crash and burn. But others you will run to them and achieve results that caution would never have achieved.

    1. Ross, thanks for commenting. The reason I used the example of Cavin's record was that it's unusual to see such huge positive splits in a good run in an ultra as most time with that sort of percentage drop-off it'd be a less fit runner at much lower paces.

    2. I'm not aware of many big ultra wins or records with such an aggressive start and such a contrast in pace which makes his run all the more impressive. For example, winners at competitive road ultras usually have to run had enough at the end to break away from the pack, as they have to do in marathons, so starting too hard tends to mean blowing up then a worse finish time, whether elite or back of the pack.