Effective sprint training



bgoetz

Active Member
Nov 25, 2010
816
38
18
I have never really done much sprint training, it has always just kinda come natural. Last season I could pop off 1500+ watt max efforts almost on command. After losing 7kg and doing a massive amount of SST and other steady state type work on the trainer and outdoors, I find myself really struggling to hit 1300 watts. Granted I have not really tried on a day when my legs are particularly fresh, but it is apparent I will need to work on my sprint a bit. So I know Dave advocates sprint @ 100% every time which I agree with, what I am wondering is does this mean 100% PE, 100% power, or both. IOW, do I go out and give a full sprint effort, recover 10min and go again, maximizing PE and wattage. OR do I go find a nice hill or overpass and do hill sprints with shorter recovery intervals @ 100% PE, recognizing that the shorter recovery will minimize power output?
 
Originally Posted by bgoetz .

...So I know Dave advocates sprint @ 100% every time which I agree with, what I am wondering is does this mean 100% PE, 100% power, or both. IOW, do I go out and give a full sprint effort, recover 10min and go again, maximizing PE and wattage. OR do I go find a nice hill or overpass and do hill sprints with shorter recovery intervals @ 100% PE, recognizing that the shorter recovery will minimize power output?
It means focus on sprints from full or nearly full recovery as track sprinters do, not sprints from intentional partial recovery hoping to turn it into some form of AWC workout instead of a largely neuromuscular workout. That's if you're working on peak speed and peak power as well as sustainable power over the duration of the actual sprint.

Sure you can take the race specificity view and do sprints after a hard minute interval or something like that to emulate the way you'll hit the sprint fatigued in actual races but IME that's not the best way to improve your actual sprint. Also realize that even with five or ten minutes of easy spinning between sprints each one is increasingly less dependent on phosphocreatine and purely anaerobic processes and each one draws more heavily on aerobic processes. This has been demonstrated with track sprinters in as few as three consecutive sprints with long rest periods between them and is almost certainly true out on the road with some riding around between each sprint. IOW, you'll get some cumulative fatigue between each sprint even with long rest periods.

What I would suggest based on a lot of recent success with a variety of racers that were struggling with sprints is to work both the traditional short intense speed burst sprint and the longer 'drag race' sprints. They're different and both very useful but a lot of folks focus on the peak power, peak jump speed variety which works best when someone else has strung out the race, either though an intentional lead out or by jumping first and you were able to grab their wheel. In those situations a big power/speed burst to jump clear, get separation, and nail the final hundred or so meters to the line is the key.

But in many road racing and crit situations you can do much better and own your own destiny better by launching early and long in a bigger gear but without the massive power/speed peak but instead rising slowly out of the saddle, rocking that bike up to speed and continuing to accelerate all the way to the line. You've got to figure out your ideal range for this move because if your speed plateaus much before the finish line or worse fades as you blow up, you'll have just turned your sprint into a great leadout for someone else. But when you get the distance right such that your cadence and speed continues to ramp all the way in it can be surprisingly difficult for someone to come around at the end. These should be something like 20 to even 30 second sprints and can change your range from 200 meters to up to maybe 400+ meters in good conditions (tailwind, fast roads, flat or slightly descending finishes might stretch it longer, uphills, headwinds, slow roads will shorten that range a bit). Practice some of these in training to figure out your best range but you may be surprised how well you can do when you take control of the race and turn a pure speed sprinter's game into a power rider's game and force others to drag race against you into the line.

But even in this case I'd strongly suggest being very well recovered between each sprint so you can give your all and really focus on the neuromuscular and anaerobic work and not do these part way and try to make them into AWC and lactate tolerance work, save that for another day.

BTW, from a power perspective in your files look for a short and very tall power burst at the start of the short speed sprints, cadence should be fairly high even before you jump as in 90 - 95 rpm, then a very tall power spike that drops off rapidly but your cadence and speed continues to climb for the 8 to 15 seconds or so of short intense speed.

In the long drag race sprint the starting cadence is typically very low as in 70 or less, you rise up and start rocking the bike and spend the first half of the 'sprint' getting on top of the gear and up to speed before you wind it out at the end. The power curve is much broader and not as sharp or as tall a peak and you likely won't hit your highest power numbers in these until the halfway point or later but the power stays relatively high for a good 15 seconds or more. Your cadence and speed should ideally ramp up all the way to the finish. There's lot's of variations on this theme like ramping the speed in controlled steps or controlling a rider from the front by just using enough to hold him on your wheel till you either hit your ideal final burst range or he jumps and you have to go. That's track tactics stuff but if you work on your long controlled game you can use it on the road in some situations as well.

So yeah, I'd still say sprint as fresh as you can when you sprint but there are different kinds of sprints and for many roadies the long sprint can be a very powerful weapon or at least a good tool to have in their tool kits.

-Dave
 
Thanks Dave, I may try to incorporate these into just my easy L2 days as I don't think a couple sprints will add much with regards to recovery demand for subsequent workouts, or am I underestimating what a couple hard nuromuscluar efforts can do to recovery? Also I like your description of each type of sprint and how it should look and feel, this is real helpful and may actually provide insight as to why my #s are low, not having sprinted in some time I think my efforts are a mix of the two. I does seem that I have had my success from further out by making a gap and holding it to the line, while my max power is pretty high, I have had some high 30-45 second efforts. In fact the first 45 seconds of any of my 1 min tests seem to be lights out, with a big drop after.
 
A couple of suggestions from my experience. I am not a strong sprinter (max 5s power ~1200W), but I have done a lot of experimenting with sprint training and have some observations (some of which were rather surprising).

I do sprint training in a couple of different ways. On just about every ride, I toss in 2-3 full-power 5s sprints during the last 10-15mins of the ride when I am basically cooling down. During this period, I am doing only ~150W AP, so I'll do a 5s full-power sprint, recover at 150W for 5mins, repeat until I'm home. The second time I do sprints is a dedicated L7 workout. My standard procedure is warm-up for 15mins, then 10x full-power 5s sprints with 5min recovery at 150W, with 10-15min cool-down. I do this workout both on the road and on my trainer. To my surprise, I can do the same max power on my trainer as on the road. This revelation is what caused me to question the value of rocking the bike from side to side.

I was quite curious about the effect of cadence on my max 5s power, so I did a lot of experiments with cadence. I tested every initial cadence from 50 to 150. I found that my max 5s power really plummets if my initial cadence is less than about 90, and my best results are when my initial cadence is about 100. I lose power >150, so starting at 100 and staying in the same gear allows me to accelerate my bike speed by 50% before I start to lose power due to cadence. Basically, I make my power with cadence, not torque. I strongly encourage you to do this sort of testing because as you approach a sprint you want to be in the correct gear for your max power.

BTW, out of curiosity I have done some sprints on my trainer with no upper body involvement. Basically, I just sit straight upright and don't even hold the bars. I just wanted to isolate the power added by my upper body. I was kind of surprised to see how well I could do with no upper body involvement. It was within a couple of hundred watts of my max power with upper body involvement.

So, what was the net effect of all this experimentation. I increased my untrained 5s power from about 800W to 1200W. I have hit 1250W a time or two, but when I do I always wonder if it is real because I can't regularly repeat it.
 
RDO, I second those observations and approaches to sprinting. No real worries throwing the short burst sprints in at the tail end of many rides but keep them short and focus on leg speed and peak power. The long drag race variety can impact recovery especially if you're really hammering big gears or doing them into a headwind or a false flat uphill so save those for days in your schedule when you can afford the extra recovery time.

I also agree with the cadence observations. I've never seen anyone hit their best peak 1 to 5 second power if they jumped with a cadence below 90 or so and definitely not if they're grinding down at 80 or less as many do when they shift a couple of cogs in preparation for a sprint and jump over geared. That's exactly what you want to do for the long drag race but that's not a peak power spike effort that's an extended period of high but not quite peak power. If you want to hit your best 5 second power then gear to spin fast BEFORE you jump and trust in your leg speed to take you to the line or shift up a cog or two after you've gotten that initial separation.

Understand that you'll basically lose a half bike length or more to riders not making those mid sprint shifts each time you shift so it's much better to have the leg speed to wind it out. Most of the folks I've worked with who haven't done a lot of track riding top out around 130 to 140 rpm unless they really work some leg speed drills but even that is a lot. If you're spinning 95 prm pre-jump and can wind it to say 135 and you launch that jump at 27 mph like you might for a crit prime, you'll break 38 mph on leg speed alone without any shifts (assuming you have the power to pull that off).

Interesting observations on upper body involvement or lack thereof. I definitely generate more power with a firm grip on the bars and my hands deep in the hooks. I focus on pulling the bars back towards my abdomen, not up towards my shoulders. Sprint form is also really important in part because aerodynamics plays such a role at typical sprint speeds. So focus on good form with your shoulders right over your hands, arms deeply flexed but firm to hold your shoulders and head low and butt just a bit in front of the saddle but not so far forward that you start skipping your rear wheel. Here's my favorite video demonstrating awesome out of the saddle sprint form: https://www.youtube.com/watch?v=iCUD0CPKJSM

Renshaw basically executes a perfect drag race sprint from about five hundred meters after Hincapie's pull to get them in position and if Cav wasn't on his wheel he wins this race as he took second even after the big leadout. And if you look at the start of Renshaw's pull just before he totally shuts down the inside attack on the final corner, he's not bursting at light speed or peak power, he's winding it up and looking like he's simply pulling the entire field to the line but he just keeps opening it up all the way till Cav launches. Look at their upper body positions in the side view shots and how low and tight they stay while pounding out of the saddle and look at how Cav seals the deal without a shift or bigger gears but with leg speed.

-Dave
 
For the cadence vs torque the I think the ideal condition for peak power development is that you start fresh from high enough speed so that (as stated above) your cadence is high enough for starters to achieve the optimum range before fatigue kicks in (which happens in a few seconds) and on the other hand don't spin out too quickly (which tends to happen if you start from low speed with small gear). That would in practise mean that start the sprint after a downhill where your speed comes up to at least 40-45kmh.

High peak power doesn't win races but it's quite hard to put out 1kW for 30sec unless you can hit that 1500w first so I think peak power development until some point is necessary base for sprinting.

If you can afford to have a dedicated day (after a rest day) in the program for a few weeks I think it is well worth while and on training day quality before quantity.
 
Just an observation reguarding ability to put out 1000w for 30's by being able to hit a peak of 1500w at the start, I've never hit 1500w but can routinely hold 900w -1100w for 20-30's without ever going much over 1200w, I also don't hit these numbers for those durations at high cadences either, in fact I never go above 100rpm.
As far as hitting peak 5's power in training at ~1350w well that usually happens at ~ 110-125rpm, but that is never the case in races, I'm never fresh enough to hit those types of rpms in a sprint unless seated, much like what you see in track sprinting where they might stand for a few sec's to get on top of the gear then sit down and wind the gear right out... and their race only lasts seconds not hours like most of ours,
I don't think Cav's cadence is all that high and certainly no where near 140rpm, he also doesn't have to try and get the jump on his competitors like what we amatures tend to do and see alot in our races, Cav "almost" always gets led out in the sprint at warp speed and gear choice would never be a big concern, either 11 or 12 tooth, not really a deal breaker and hard to fark that up,
but I will agree his technique and position is as good as it gets, he is basicly licking the tyres clean with his tounge and something I have adopted also when sprinting.... I like to have clean tyres too.
 
Originally Posted by bubsy . Just an observation reguarding ability to put out 1000w for 30's by being able to hit a peak of 1500w at the start, I've never hit 1500w but can routinely hold 900w -1100w for 20-30's without ever going much over 1200w, I also don't hit these numbers for those durations at high cadences either, in fact I never go above 100rpm.
I can do 900-1000 for 20-30s too again without going much over 1200 (actual all time peak is over 1300 but if I open with that in a 20-30sec effort I just fade more). However I do have to use high cadences to do it - and also have been freewheeling immediately before and doing low watts for previous minutes, so pretty much completely useless in any race finish.
 
Originally Posted by bgoetz .

Granted I have not really tried on a day when my legs are particularly fresh
Most of the time a sprint comes when you are dead tired. So however you train a measure of the effectiveness of your training is how well you sprint when you are tired.
 
The hardest parts of most races are when you are dead tired, but it would be pointless to attempt a VO2 workout with totally skunked legs (trust me I have tried). Sorry AOG, I think I will go with Dave on this and do dedicated sprint work fresh. Sure I may continue to throw in a sprint at the end of a hard workout, but it is really just a single effort and not what I would consider a workout.
 
Originally Posted by bubsy .

Just an observation reguarding ability to put out 1000w for 30's by being able to hit a peak of 1500w at the start, I've never hit 1500w but can routinely hold 900w -1100w for 20-30's without ever going much over 1200w, I also don't hit these numbers for those durations at high cadences either, in fact I never go above 100rpm.
As far as hitting peak 5's power in training at ~1350w well that usually happens at ~ 110-125rpm, but that is never the case in races, I'm never fresh enough to hit those types of rpms in a sprint unless seated, much like what you see in track sprinting where they might stand for a few sec's to get on top of the gear then sit down and wind the gear right out... and their race only lasts seconds not hours like most of ours,
I don't think Cav's cadence is all that high and certainly no where near 140rpm, he also doesn't have to try and get the jump on his competitors like what we amatures tend to do and see alot in our races, Cav "almost" always gets led out in the sprint at warp speed and gear choice would never be a big concern, either 11 or 12 tooth, not really a deal breaker and hard to fark that up,
but I will agree his technique and position is as good as it gets, he is basicly licking the tyres clean with his tounge and something I have adopted also when sprinting.... I like to have clean tyres too.
Not necessarily hitting 1500w at the start of 30s sprint, but because the decline in power is extremely steep at that part of power-duration curve it is unusual that you could maintain 1000w for 30s if your peak power were "only" say 1200w. I don't say impossible but if that indeed were the case your pm might need some check or you simply don't know "how to sprint big numbers" (no offense meant).

You have to make some distinction between training and racing: Cav is not trying to put out the highest power number, he's trying to win a race. To win a race you use whatever strategy/tactic that gives a competitive edge. Training on the other hand I think should go from general to specific so that you start by developing the essential properties (like peak power for sprinting) and proceed to more specific exercises (add anaerobic capacity by lengthening the sprint, try to improve maximum speed, etc).
 
I don't question Busby's PM calibration at all. I've seen similar things in terms of peak power vs 20-30 second power from several athletes, my numbers are a bit lower but pretty similar with a peak power less than 1400, 5 second less than 1300 and 20-30 seconds in the 900-1000 watt range and I'm pretty anal about calibrating and torque zeroing my PMs.

I've worked with several folks that show similar roll off in their MMP curves over short time frames but one key is these longer drag race sprints are not approached like the short speed burst sprints. The longer sprints don't have the characteristic huge and steep power spike at the start where the rider jumps super hard on those opening pedal strokes but instead are more of a ramped power curve with a broad bell shaped section that's not as high but sustained much longer. The speed curve also ramps up more steadily but ideally continues to ramp all the way to the end of these long sprints or they turn from sprints into leadouts.

They're pretty much different beasts than the short super punchy speed sprints for whipping around a rider in the final hundred meters or so. I'm sure if you approached the longer sprints with the same power profile and instant jump of the short efforts it would take a huge top end in your NM power to get a longer effort anywhere near the numbers we're talking about but these really shouldn't be approached in the same way any more than you'd start a 40 K time trial with the same pacing as a 4 K pursuit. They're both fast individual time trials but approached very differently from a power pacing perspective.

-Dave
 
Don't really want to argue, but if you read it very carefully: "but because the decline in power is extremely steep at that part of power-duration curve it is unusual that you could maintain 1000w for 30s if your peak power were "only" say 1200w." And 20s is different from 30s.

Bupsy said his 5s is ~1350w which would probably make his peak power closer to 1500w.


Edit: to get back to the track of the thread and where this sub-discussion started from, my approach would be to work on peak power for a short while first and then start take it to more race specific increasing the length of the sprint, watching speed more carefully than just power and maybe even do sprints fatigued like in a road race.
Could be that I look too much on the numbers and the peak power part is useless and one can jump right away to longer efforts.

The technical and tactical part then a whole another story.
 
Fair enough, I was actually responding to your original statement "...but it's quite hard to put out 1kW for 30sec unless you can hit that 1500w first..." which hasn't been my experience (in a literal as well as power scaled way) and from the responses doesn't seem to be the experience of others.

But yes, one way or another it pays to work on both the short intense peak power sprinting and the longer drawn out drag race sprinting and they're both useful. Hard to say which to work first, I expect it changes from rider to rider. I've worked with some folks who have pretty decent sprint form including body position, upper body engagement and the like but seem to struggle with a quick and powerful jump. These folks get a lot out of short sprint work focusing on that peak power burst. Others really struggle with body position and skip their rear wheels by jumping too far forward or stand much too tall as if they're out of the saddle climbing instead of sprinting. For these folks a good dose of longer ramped drag race sprints can help them focus on form and body position, bike rock and other things before they put that all together in some short explosive sprinting.

My main point in chiming in as that a lot of folks focus almost exclusively on the short huge power burst sprint with it's high peak and 5 second power and ignore the long drawn out drag race but IME that wins a lot more amateur races where lot's of riders hesitate waiting for someone else to initiate. We all know how the aerodynamics work and how it can pay off to stay sheltered before jumping off a fast wheel but it's surprising how often the race goes to the first guy to jump and that usually means going long. Good idea to train for both situations.

-Dave
 
Originally Posted by frost .


Not necessarily hitting 1500w at the start of 30s sprint, but because the decline in power is extremely steep at that part of power-duration curve it is unusual that you could maintain 1000w for 30s if your peak power were "only" say 1200w. I don't say impossible but if that indeed were the case your pm might need some check or you simply don't know "how to sprint big numbers" (no offense meant).

You have to make some distinction between training and racing: Cav is not trying to put out the highest power number, he's trying to win a race. To win a race you use whatever strategy/tactic that gives a competitive edge. Training on the other hand I think should go from general to specific so that you start by developing the essential properties (like peak power for sprinting) and proceed to more specific exercises (add anaerobic capacity by lengthening the sprint, try to improve maximum speed, etc).
I'll try to stay on topic some what, but first to clarify,

I find it unusual that someone should feel the need for a 1500w kicker at the start of a 1kw 30's effort, that's if the goal was to test 30's power or to do some 30's intervals at max power.

My "peak" power is only ~1420w and if I try and hit that at the start of a 30's all out effort than it results in a much lower 30's avg as Dave has pointed out, infact I'll start and finish just below 1000w and peak in the middle like a nice arch power curve, maybee you don't know how to properly pace a 30's interval or 500m TT (no offense meant)

If i try and hit 1400+ watts at the start of a 30's interval by 20's I'm crawling and both power and speed have faded badly,
If the idea behind the 1500w kicker in training is to get seperation from the pack then it doesn't matter if your crawling at the finish line so long as your the first one across it, but that comes down to tactics and were not talking that here we are trying to get the OP's power up.

I used to be a terible sprinter, only cracked 1000w a hand full of times and peak 5's was only ~985w and that was with short high end sprint training, I just did not respond at all well to that type of sprint work and still don't,

It wasn't untill someone pointed out that in order to be a successful road sprinter (FTP aside) one needed to have a highly developed AWC so ~3years on with 3 - 4mths a year of 1 - 2 days p/wk of nothing but 12's - 30's of L6/7 and just some "maintainence" work every 7 - 10 days during race season my 30's power is now what my old 5's power used to be.
5's peak power now rises with little to no specific work and stays just ahead of 10 - 20 and 30's power as I continue to improve in that area.

For me the focusing on loads and loads of longer sprint intervals and standing or rolling starts 12 -30's in length with "complete" recovery is what stimulates addaption.
but I'm not a natural or gifted sprinter so a true fast twitcher may not need to train the way I do.

I hope this will help the OP or anyone looking for something to improve their peak power.
 
I was never talking about 1500w kicker at the 30s "test" but the decline of the power-duration curve, which your example of having "closer to 1500w peak power" (1420w is quite close) very well demonstrates.

Edit: an analogy to make it clearer: It wouldn't be wise put out your best 5min at the beginning 1 hour TT but you would have hard time ever reaching 300w if your 5 min power is 320w.
 
Originally Posted by frost .
...Edit: an analogy to make it clearer: It wouldn't be wise put out your best 5min at the beginning 1 hour TT but you would have hard time ever reaching 300w if your 5 min power is 320w.
No argument with the VO2 to FTP example, I think the only push back you were getting above was the assertion that a to hit a certain power for 30 seconds you first must have a peak power of 150% of that value. If your example was that you needed a 5 minute power of 450 watts to hold 300 for an hour there would have been push back as well.

I really think Busby's point is key. The trick to a solid long drag race style road sprint is tied more closely to AWC development than pure NM power which is why a lot of time spent working on that high peak power may not translate to holding a very but sub-peak power for 20 to 30 seconds. Sure it'll never hurt to have more peak power at your disposal but if you believe that's a prerequisite and spend most of your sprint training time working on that aspect you may not get there and for the wrong reasons.

I'm really not trying to sustain an internet pissing match as I know you were just illustrating an idea that your peak better be above whatever you're targeting for a long sprint. That makes perfect sense but the point that some of us were trying to make is that even though both peak short burst power (e.g. 5 to 7 seconds) and longer sub-peak long sprint power (e.g. 20 to 30 seconds) might both be used in the closing minute of a race and both be called sprints they're really different things and though there's obviously overlap they really draw on different systems and are built though somewhat different training.

From a practical perspective this can be really important to the very common situation where a fit roadie says 'but I'm just not a sprinter and don't have that kind of jump...' It might be true that a fit and smaller road racer might never develop that 1500 watt peak power but if they believe that they must first accomplish that goal before working on a long drag race sprint they may not even work on the latter which is likely their best bet for winning in group sprinting situations. Sure they definitely should spend time working on both with the reality check that not everyone is Boonen or Cav' and the short kick and peak power is still often needed when you don't get the early jump or are lucky enough to come to the line following a fast wheel. But if you understand that the two sprints are different and draw from slightly different systems it can open up doors to folks who don't think they'll ever be able to bunch sprint.

-Dave
 
Hi, my first post on this forum, but certainly not the first time I've visited Cyclingforums!

Although I have only been on the road since September (probably only ridden about 500km or so, the cycling season in Norway is not very long, still snow here /img/vbsmilies/smilies/frown.gif ), I have realized the thing about cycling that I like the most, is sprinting. Every thread that has something to do with sprinting has my interest, and there have been many good posts in this one. I will now share my experiences (the little I have /img/vbsmilies/smilies/smile.gif ) and thoughts on sprint training, please feel free to comment.

- When training sprint, it is very important to get full rest between efforts. For me that means a minimum of 5min, often more. A typical sprint session for me takes about 90 minutes, where 15-20 of them are warm up, and 10-15 just cooling down. The main session is about 60 minutes, with 8 sprints @ 6-12 seconds, so a huge percentage of my sprint workouts are rest. In order to train speed, you have to be rested to perform at a level that makes you improve. It is true that you also must train your sprint while tired, I usually end all my rides with a 10-15 second all out sprint, no matter how tired I am. When I say end I mean 5-10 minutes before I am home, to let my muscles cool down a bit.

- My favourite sprint session is to simulate a lead out, using a downhill to build up my speed to about 50kph, and then go all out on the following flat for 200m. Other workouts I find good are overspeed (sprinting on slight declines etc), short acceleration sprints (4-5 seconds from a cruising pace @ 90-100rpms in one gear, typically done on longer rides) and power sprints (overgeared, uphill etc).

- I am also a firm believer that strength training does not help road sprinters. The main reason being that a road sprint typically starts at high speeds, ie 50+ kph (30-35 mph), so max strength is not a factor. For track sprinters I do believe it helps on standing starts and sudden accelerations from low speed in a match sprint, but for roadies, nope. Its the same thing as running sprints, gym work and heavy squats will help your initial acceleration, the first 10-20meters or so of a 100m dash, but it will do nothing for your top end. That has to be trained in the track (or on the bike in our case).

Those are a few of my opinions, I am not talking in absolutes here.

Phew, a long first post this one /img/vbsmilies/smilies/redface.gif