Mel’s Pulling Through Stroke

Mel's pulling through

Mel Krieger’s concept of a “pulling” stroke as opposed to a “pushing” one wasn’t intended as a way of differentiating two different casting styles, it was coined to describe any good casting stroke whatever the style used. So you should “pull through” when using an “elbow up-down” style, as much as when using an “elbow backward-forward” one.

Pulling was the term Mel chose to convey the idea of leaving for the end of the stroke as much of the rod rotation as possible. In summary (and using two other Mel’s concepts) “pulling” isn’t a style, it is substance; “pushing” isn’t a style, it is a fault.

The gif above shows the paradox posed by the interpretation of “pulling” as just an stylistic issue: Mel himself is showing what he meant by “pulling through” while performing a cast that some instructors would still define as “pushing”.

Better to let Mel explain himself about the “pulling” concept, and what the gif above is showing (bold is mine):

In the pull through casting stroke, the casting hand precedes the rod tip through most of the casting stroke and the turnover and stop take place only at the end of the casting stroke… Lay out 70 or so feet of fly line on a lawn behind you, fly rod pointing to the fly, and throw a javelin, turning the rod over only at the very end of the throw. You may be pleasantly surprised with this extreme pull through casting motion.

Relax, Balance, Control

balanced reel

We usually say about masters in any discipline that “they make it look easy”. Probably it isn’t just that it looks easy, I have the conviction that they look so relaxed because… well, they are totally relaxed.

As my fly casting mentor likes to say, a long cast starts in the tip of the toes and ends in the finger tips. Less extreme casts have less body parts in play, but share some important trait in common: good form requires to be as effortless as possible.

In everyday trout fishing casts, our main engine is a combination of arm, forearm and hand. That is why fly casting manuals recommend to apply force with the hand only at the end of the stroke: just around the stop.

Closing your rod hand tight only at the right moment, and for a brief time, makes a lot of sense. Open and close your rod hand while holding your forearm with your other hand, and feel how many muscles get tensioned by that simple exercise. This offers an explanation for that ache starting in the neck and going down the back, so many of us feel after a long day on the river.

I have no idea about biomechanics, but my gut feeling tells me that muscle tension and accuracy are rather incompatible. Just another good reason to train the closing-opening of the hand during the casting stroke.

But truly relaxing the hand is pretty much impossible if we departure from an inadequate starting point. Every time I see someone casting with his reel parallel to the casting plane, I know that things aren’t as effortless as they could be. They say that a picture is worth a thousand words, so if it is a video probably even more.

I must confess that just making a few casts with the rim of the reel facing the target felt totally uncomfortable, as you have to constantly force the rod to keep the reel in place. The same goes for an overhead cast, as we normally make them with a slight inclination to the side.

One day this Autumn I handed my rod to my guide and asked him to show me his long nymphing approach in a particularly deep run. I immediately noticed how he got the reel in the “hanging”, relaxed, position. I asked him why; “balance” he said. A much more concise way of conveying the same idea.

Focus Shift

Nowhere in the world of sending a fly out there with a line you can find rod load being more glorified than in the spey casting scene. Everything seems to gravitate around that. If the cast is good it is because the rod was properly loaded. If it went wrong… well, sure it is due to the rod not having enough load or unloading prematurely.

Sometimes it is possible to get more clues from the analysis of a bad cast than from a perfect one. That is the case with the casts depicted here. The video and pics show a pretty common occurrence that will be used to point out some keys of spey casting mechanics. Take them just as a brief introduction to following articles which will get deeper into that subject (slow motion clips and some not-that-heavy-physics included).

The scenario is the forward cast of a spey characterized by some kind of V-Loop that we will call 7-Loop (thanks to Simon Gawesworth). An extreme 7 for that matter.

Let’s say that you set a nice V-Loop, make the cast and present the fly on target.
On the next cast you manage to get a 7-Loop and the fly falls short of the target on top of a heap of line and leader. That is just one of the possible outcomes of that loop configuration -as it is a fat loop, a tailing loop or even the three of them combined-  if the caster doesn’t compensate his stroke to adapt. Even if he modifies his stroke successfully the 7-Loop is still inefficient due to the amount of wasted energy.

The following gif made from a couple of pics from a still camera will shed some additional light.


We could look for an explanation to the inefficiency of that 7-Loop in the gif above in the usual way, basing our analysis in the behaviour showed by the rod. It would go along the following lines.

What happened to this cast?

  • Hmm, probably the rod didn’t get loaded… but I see a pretty good bend, though!
  • However, is that a properly loaded rod? Who knows? I, for one, don’t have a clue nor have met anybody capable of quantifying whether a given load is enough for a given cast or not. If only because you can make a cast to the same distance with the same rod and line with rather different loads.
  • Well -you say to yourself- it could be that the rod got unloaded prematurely due to the anchor slipping… but the loop’s rod leg looks pretty straight, whereas the rising tip of an early unloading would have set a wave in it!
    As shown here:

Hmm, we are not getting very far with that approach.

So let’s address the issue from a different standpoint, forgetting the rod and putting the accent in the line.

From the gif above we can quickly draw some visual clues:

  • A totally ineffective anchor due to the angle of attack.
    Only part of the leader is in contact with the water, moreover the angle at which the loop pulls on the anchor makes the latter specially prone to slipping. See how at the end of the stroke the apex of the loop still hasn’t moved forward due to the slipping anchor! In fact it moves a little bit backward! It is a perfect case of a loop propagating but not traveling (but that will be the subject of another article).
  • The amount of line in the rod leg of the 7-Loop.
    The length of line in the rod leg at the start of the stroke is very very short. The longer the piece of line we propel during the casting stroke (the “live line” so to speak) the more efficient the cast:

The reason for that inefficiency has already been covered in this previous article. You can also relate it to the case when we rush the forward stroke of an overhead cast and start it with the line still half its way backwards.

So, compared to a proper V-Loop configuration, for presenting the fly at the same distance a 7-Loop:

  • Will ask for a higher rod butt acceleration, as a way to give enough momentum to the comparatively shorter length of “live line” we propel during the stroke, to carry a comparatively longer length of “dead line”. That is the reason why the  “dead line” to “live line” ratio is key regarding efficiency: the longer the live line the better.
  • Anchor slipping wastes energy: the line moving backwards goes in the opposite direction of the target, and it doesn’t move by itself so it detracts from the energy needed to move the line forward.

Obviously the longer the cast the higher the impulse you need, which may result in a bigger load, but load is a byproduct of our force application to give the line enough momentum. It really isn’t our goal.

Given that the function of the casting stroke is to give enough velocity to the line in the right direction, it is better then to shift our focus from the rod -which says very little- to the line -which speaks volumes.

Why is a Jump Roll More Efficient than a Static One?

That is a rather usual question and also a very interesting one. It has been asked to me again recently via the internet by a couple of fellow casting instructors. Let’s go for it.

Let’s take a look first at the Roll vs. Overhead video used in a previous article:

I will describe the scenario:
Just one rod rigged with two lines: Royal Wulff #7 and Rio Tournament #6.
The TT has the ideal taper for roll casting; the Rio is designed for long casts overhead.
The Rio is unrolled behind the caster; the TT set in a roll cast configuration with its leader anchored by means of a screwdriver stuck in the ground (is there a more solid anchor than planet Earth itself? If there is let me know).

In this way the very same casting stroke applies force to both lines. What the video shows, however, is that the overhead line reaches its target whereas the rolled one falls short. It is an interesting experiment that any caster can try by himself (just in case somebody doubts of the result).
Anyway the result is that, if you use a stroke with the amount of energy needed for the overhead line just to straighten, the roll cast line falls short of the target. Always. Why?

I have said that the very same casting stroke applies force to both lines simultaneously, but is it the same amount of force for both? No, it isn’t.
What our stroke is doing is applying the same acceleration to the rod, and the rod to both lines, but the length of line actually accelerated in one case is much longer than in the other: we accelerate the whole length of the overhead line whereas only a short piece of the roll line is subjected to acceleration.

Do you remember the basic formula of Force?
F = m.a.
For a refresher this is a nice and easy source Force

Acceleration is the same for both lines but mass isn’t. So the force we apply to the roll cast line is much less than that we apply to the overhead line. For the same, identical, casting stroke getting the same distance with much less force sounds impossible, right?

If you prefer we can address the problem from the standpoint of Energy.
By applying force to the lines over a given distance we are doing Work on them. Do you remember the formula for Work?
W = F.d
For a refresher this is a nice and easy source Work

The work done on an object amounts to the energy transferred to it; that is, more Work = more Energy transferred.
But we have already discovered that the roll line has been subjected to less force than the overhead one; less force amounts to less work done on the roll line and, consequently, less energy in it.
How do you expect to get the same distance with less energy?

So what we have is that the casting stroke that works for an overhead cast isn’t good for a roll cast. By the same token the stroke good for a jump roll isn’t enough for a static roll.

Some may say that, after all, that rod load shown in the video is just enough for propelling one line but not both. Another of the blindfolds that the casting paradigm based in load puts over our eyes.
Yes, only one line is propelled all the way forward but, how is it that it is always the one with the longer piece of “live line”?

So, as a corollary, IMO we have to think in terms of how long is the piece of line I am accelerating directly to the target, and not of the mythical “load” and “anchor loading” or the impossibility of a “highly energized V loop”.


The key to understand this issue lies in the fact that the rod in the video isn’t applying the same force to both lines, just the same acceleration. It seems that it isn’t easy to grasp so an additional (more graphic) example to clarify this is in order:

Let’s say I have a video showing three model railway coaches.
Two are connected together and laying on a straight rail. Parallel to them there is another rail with the third coach laying on it.

In the front part of both convoys we have a string, one string for each convoy.
At a given time the strings get taught and the two convoys start moving with exactly the same acceleration.

Since the track has a ruler alongside, by means of Tracker or any other application we are able of easily calculate the magnitude of that acceleration.

We also know the respective masses of each convoy: isn’t a surprise that mass A has a value X and mass B has a value of X/2.

By solving the F = m.a equation we get that the force applied to mass A is double the force applied to mass B. Right?

Well, after that first video I show you a second one with a general view of the scene. Now we can see that there is a guy pulling on both convoys at once by holding both strings in one hand.
So now, out of a sudden, we discover that he is applying the same force to both convoys? Not al all, the force exerted on each of them is different, one being exactly half the value of the other.

A Little Exercise

As an attachment to the previous articles on tailing loops (here and here) now an exercise on diagnosing a common casting fault. You are a casting instructor and your student is getting a recurrent tailing tendency. I shot this clip yesterday, playing as student and instructor at the same time. After dozens of plays I still can’t say what the origin of the problem is, even seeing when it is produced (watching carefully you can see the slight rise of the rod tip and the subsequent wave in the line).

What I know is that I was playing with the haul, trying to release the line just at loop formation (wherever that is). That could have resulted in a premature end of the acceleration of the hauling hand and the immediate tip rise. But, honestly, I don’t know and find incredibly difficult to diagnose and cure this kind of things.

Mysterious Creature Rides Again

Tailing loops. So frequent and still so puzzling.

As I wrote on the first post in this series we already have a pretty good idea of how tails are formed; getting rid of them is another matter entirely. I truly admire the insights of instructors from yesterday: reaching the conclusion that tailing loops come from a concave tip path of the rod tip wouldn’t come easily, specially if we take into account that there wasn’t high speed video available at the time. Today’s technology effortlessly shows that, in fact, it is a dip/rise of the rod tip what creates the dreaded tail. And this evidence renews my admiration for the amazing observation skills of those pioneers of casting studies, for although that dip/rise is somewhat a “concave path of the rod tip” it has nothing to do with those big bowl shaped tip paths so many drawings depict. For years those bowl shaped explanations were to me as perplexing as the tailing loops themselves: however much I looked whenever I saw a tail in someone’s casting I couldn’t see that big concave path everybody was writing about. Not even on the casting videos available. Reality is much much more subtle, so subtle that seeing with the naked eye the expected anomaly in the tip path -even knowing what to look for- is really hard. Here we have a tailing loop in full glory. It is played at a slower pace than real speed. The tail could be used to illustrate a casting handbook; can you see the “bowled rod tip” anywhere?: Tailing loop backcast a bit slower than real Better to use a gif at 100 frames per second, that is 1/4 of the actual speed: Tailing loop backcast slow Observe how even at a pace three times slower than reality we just can catch a glimpse of some anomaly in the tip path. So let’s use a visual aid to see what is exactly happening with the rod tip: Tailing loop path This has cleared things up a little bit. Mainly two things come to my mind. First is that to get a tailing loop, even a huge one like that shown above, you only need to mess up a relatively short piece of the casting stroke. Second is a consequence of the previous observation and my main point so far: that this problem is so recurring due to the fact that a very small error, for just an instant, results in a surprinsingly big effect. It isn’t easy to feel, and then correct, things that happen in an instant, is it? It isn’t easy to detect for the caster himself nor for anyone else. The tailing loop depicted above is really huge. Let’s watch carefully another good one of more moderate size. Can you detect where in the stroke does the error happens even in slow motion? I can’t. The only way is playing the original video frame by frame to discover a veeery subtle dip and rise of the rod tip: Tailing loop backcast small tail Dip/Rise of the rod tip. It is worth to emphasize the “Rise” part since that motion is key in the formation of the transverse wave in the fly leg that we commonly call tailing loop. But that, together with some considerations about what is the ultimate cause of tails, is the stuff for a next article. P.S. The tailing loops shown here are real ones, nothing staged for the camera but involuntarily produced. The caster is a really fine one who drove from 400 km away for a course to improve his technique (I felt flattered and, at the same time, worried: would I deliver as expected?) His hauled casts were really nice. Then I took the camera and asked him to cast with the rod hand only. Removing the haul wreaks havoc with line control, but it is a fantastic exercise to educate our rod hand.

Distance casting? What for?

It is useless, most of the fish are caught within 12 meters. I have lost count of the number of times I have read and heard that kind of statement. Being a 99.9 % dry fly fisher myself I, almost, agree. There is a lot of truth in that reasoning. Anyway, if we don’t catch as many fish further than 12 meters away it could also be because we don’t cast to them, couldn’t it? Admittedly getting a dead drift with a long cast is some sort of mission impossible, although there are nymphing techniques for which distance isn’t a problem: if there is a fish lie out there… out there my nymph goes. There is another type of very special nymphing that asks for being able to cast as long as possible: nymphing for sea trout. Are you kidding? Sea trout on nymphs? Yes, big sea run brown brown trout on small nymphs. Only in Southern Patagonia I must add. For instance on Río Gallegos. Size #10 nymphs, like that in the following pic: Rubber legs

I will commit to training more specifically for distance with the double handed rod before traveling to Río Gallegos again. Not all the lies are very far away (although some of them ask for 30+ meters casts) the real problem is the relentless wind; if it wasn’t so cold one would say it comes directly from hell. Here is the result of a sideways breeze (an a mild one by the river standards) on a cast with a 500 grains skagit head (correction from César -the caster himself: isn’t a skagit but a Rage Compact, something like an embrutished Windcutter :-):Wind-1

Sometimes frustration is a word that falls short of explaining some feelings. I will never forget what Loro, our guide, told us the last day on the river: I have guided people who after a couple of hours fishing thrown the rod away and sat angrily on the bank. So I want to thank you for understanding how things are here. Fortunately great prizes await those who persevere:

Yes, definitely distance casting practice isn’t a strange proposition.