How the Race Was Lost: Attack! Attack! Attack!

After losing a race to a solo breakaway early in the week, I decided to jump into another race and attempt a breakaway of my own.

Let me say here at the start: I’ve never been a breakaway rider. Not outdoors, and certainly not on Zwift. As much as I admire Eddy Merckx, his “Always attack!” mantra has never worked for me.

It’s unclear if this is due to my power profile, my mentality, or something else. But with Zwift’s latest Pack Dynamics changes, breakaways stand a much better chance of surviving. So I figured it was time to start giving breakaways a go. Today was the day!

I chose stage 1 of the ZRacing series as my event, because looking at the race through the lens of the breakaway made its tiny 10km length seem quite doable. Little did I know what was in store…

The Warmup

The race began at 7:10am, so I was on the bike by 6:30 to spin up the legs with Coco. I hadn’t eaten breakfast, but I figured that was OK since it was a short race. Knowing it would be a short, sharp effort, I chewed some caffeine gum and applied PR lotion to the legs… my typical “hard race” race prep.

I paired up my Zwift Play controllers and put in a few hard, short efforts to get the heart rate up and break the ice. Then I headed for the start pens on my trusty Venge + Disc combo.

The Start

Joining the pens, I noticed the route was set to “Flat Route Reverse.” For some reason I’d been thinking it was the forward version, and had already marked my mental map for great spots to attack (the climb out of the ocean tunnel, the snaking slight climb toward the Villas, or even the ramp after the sprint).

So I had to recalibrate my brain a bit. Where are the smart attack points here? Just after the sprint, when the road ramps up out of the Villas? On the dirt before we head into the ocean tunnel? Or what about a long attack from the ramp out of the tunnel near the finish?

But I didn’t have much time to think about it – soon the clock hit zero and the race was on!

Along with wanting to initiate my own breakaway, I had decided to chase any promising-looking attacks, because with a race this short, a breakaway at any point may be able to stay away. Right at the start the race strung out, with a handful of riders clearly pushing to get into a break. I pushed to join them, but the peloton gobbled us up soon enough and we headed up the short climb into The Esses.

More Attacks, More Chasing

On The Esses one L. Weber attacked, and I chased. It was actually really fun to chase him down with the Play controllers in The Esses, because I could cut corners! Check it out below (I’ve cued the video to the right place):

Everyone else was watching for breakaways too, though. Or maybe they were marking me? (It’s flattering to think so, but I doubt it.) The blob chased their way back to me, and we rode together again for a bit. Up the kicker and down through the sprint, the effort level stayed high as riders pushed on the front – I averaged 345W for the first 4 minutes of the race, which took us through the Italian Villas.

A. Cooper jumped ahead on the snaking descent to the Fishing Village, and after a couple of riders followed I pushed hard to join them. I made it, but also succeeded in pulling most of the peloton with me!

Arriving at the Fishing Village things calmed down as riders caught a breath. Remarkably, despite all the attacks and chasing, we still had 34 riders in the group from the original 40 starters… only one rider had dropped since we entered The Esses.

The Finish

We recovered a bit down into the ocean tunnel, then with 2.3km to go two riders (C. Duponq and A. Cooper) rolled off the front and upped their watts. I had been planning to wait until the ramp out of the tunnel to launch a final breakaway attempt, but I ramped up to join them, rolling to the front to keep the speed up and get a gap on the peloton.

The riders behind just weren’t having it, though. They kept the leash short, and half the group flew past me as I eased up at the worst time – on the ramp out of the tunnel!

Out of the saddle, hammering my way back to the group, I made it to the front and grabbed a wheel. 600 meters to go, and I had an aero powerup.

With 300 meters to go, a rider popped his aero powerup and came around, so I did the same. There were two problems, though:

1. I activated the 15-second powerup too early, so it ran out with 150m to go
2. I didn’t have a sprint left, because the race had been so hard up to this point

I crossed the line in 5th. Congrats to Ramon van Rixtel, who executed a perfectly timed sprint for the win!

See results on ZwiftPower >
See activity on Strava >

Watch My Race

Takeaways

Lots to talk about here.

New race, who dis?

I’ve finished just over 1000 races according to ZwiftPower. Let me say this clearly: this race didn’t feel like a “typical” Zwift race. If you haven’t raced with Pack Dynamics 4.1 yet, you’re in for a surprise in your next race.

(Related: earlier this week I wrote about a race I lost where I rider won with a long solo breakaway. That’s something you never used to see in Zwift, either!)

I went into today’s race thinking I was the sneaky racer who knew a breakaway might work. But clearly, plenty of riders held that same knowledge, and we all had the same idea – attack and get into a breakaway that stays away.

The race was so short that it became a nearly-continuous battle of riders trying to establish a breakaway. But the peloton was strong/smart enough to pull those attacks back, so in the end it came down to a bunch sprint.

If we’d had 20 starters instead of 40, I bet one of the breakaway attempts would have stuck.

I predict this is how more and more races on Zwift will feel moving forward. Before PD4.1 you could let solo attacks go because you knew they couldn’t stay away from the fast-churning pack. That’s not the case anymore, so now we have battles to establish a breakaway, like you see IRL. And once that breakaway is up the road, the peloton will have their hands full trying to organize a chase.

It’s going to get really interesting when team racing enters the picture as ZRL and DRS kick off next week.

Hindsight Is 20/20

With the benefit of hindsight, I could say that, had I raced this differently, I could have podiumed or even won. If I had sat in the wheels and let others chase down the attacks, chances are good others would have chased them down. We still would have entered the final kilometer as a group, but my legs would have been fresher.

But that’s the tradeoff you take in bike racing outdoors, too. When pack physics (IRL or virtual) allow for breakaways to survive, you have to choose between:

• sitting in the pack and risking that the breakaway never comes back, or
• chasing/initiating breakaways, risking burning out your legs and wasting energy because you still get caught by the peloton

Pack Dynamics 4.1 has dropped us into a new realm of Zwift racing, where the possibility of successful breakaways forces us to make strategic mid-race decisions if we want to race smart. I think that’s a very good thing.

Knowledge is Power

I haven’t been researching the competition prior to my races lately, but with breakaways standing a chance, it’s more helpful than ever to have a short list of riders to watch.

Because you can’t chase every attack (I did in this short race, and had nothing left for the finish). You have to decide if an attack is worth following, and that is determined in a big way by how strong the attacking rider is.

Sauce for Zwift has a rider info window that may also be helpful here. I think I’ll dig into that…

Power Curve

This was a tough race for sure – I averaged 334W for the ~14-minute race and hit some 6-week power PRs at various spots along my power curve between 4 and 14 minutes:

Was I wasting watts? A bit, I’m sure. To be honest I wasn’t even thinking about the watt-wasting issue during this race, except at one point when it popped into my mind. (I then eased up a bit. Lo and behold: I stayed in the same place, with lower power! See that moment here, although it’s hard to see the power reduction because I have Sauce for Zwift showing instant power and not 3s average):

Mostly, though, this was a hard race because I chose to make it hard. I chased breakaways and made some attacks myself, because that was the approach I’d chosen before the clock hit zero.

Your Thoughts

Have you tried breakaway away in a recent Zwift race? Have you noticed others are trying the same? Share your thoughts below.

The first race of Zwift Racing League 2023/24 Round 1 happens Tuesday, September 12. We’ll be on Zwift’s newest map (Scotland), riding the first official scratch race in ZRL’s history!

With Pack Dynamics 4.1 in play and ZRL’s unique scratch race setup there’s much to discuss along with our usual topics of bike choice, strategic options, and crucial segments. Let’s go!

Our First ZRL Scratch Race

This is the first time* we’ve had an official scratch race in ZRL. How does it differ from the typical ZRL points race format? Simple: you don’t have any intermediates to contest! The only points available are those you earn based on your finishing position, and those finishing points follow the same finishing points scheme as a ZRL points race.

*We actually had an “unofficial” scratch race when we raced Road to Ruins in March 2022. Officially, this was a “points race,” but since there were no intermediates on that route, the only points you could earn were at the finish line!

Looking at the Route: Scotland’s Rolling Highlands

Each lap of Scotland’s Rolling Highlands route is 9km (5.6 miles), with 77m of climbing. But don’t miss that there’s a long 5.1km lead-in to the start of that first lap, ending with the short Breakaway Brae KOM! All categories are racing three laps for a 31.9km long race with 260m of climbing.

This is a constantly-undulating course, but none of the pitches are particularly steep or long. Our advice: keep your eyes open, mind your pack position, and do a few test laps before Tuesday if you aren’t familiar with the route.

This route has three notable climbs, and everything else is a minor undulation or downhill. First, we have Breakaway Brae – the only climb that is an official segment:

620 meters long at 2.4%, this one starts easy but turns into a bit of a slog over the top through the banner. Try to start near the front, and know there’s a short downhill after the banner where the pack will typically regroup.

The other two climbs are worth mentioning since they provide platforms for hard attacks/pushes:

1. The climb up through the Cliffs area is the longest uphill of the route, but it’s not a continuous grade, nor is it very steep. This climb begins just after you ride through the mystical roundabout.
2. The sneaky “Corkscrew Castle Climb” serves as an excellent attack point due to its punchy steepness and proximity to the finish line (1.5km up the road). This one arrives after the longest descent on the map, so rest your legs and prepare to hammer!

Read more about the Rolling Highlands route >

PowerUp Notes

In a fresh twist on powerups, WTRL tells us riders will be awarded feathers at the banner near the start pens, and burritos at the Breakaway Brae banner. If you haven’t been racing on Zwift much lately, it’s worth noting that Zwift rebalanced several powerups (including the feather and burrito) in May. Make sure you know how they work!

Lightweight (feather) (given at banner near the pens): reduces your weight by 10% for 30 seconds. Use on climbs, when weight matters the most.

Burrito (given at Breakaway Brae banner): turns off the draft effect for riders in a cone-shaped window behind you for 20 seconds. Use when attacking off the front so opponents have to work harder to follow you, or in the peloton to make others work harder.

Bike Frame + Wheel Choice

None of the climbs here are long or steep enough for a true climbing bike to be the sensible choice. A full aero bike will deliver the fastest overall lap time at steady state wattage, but with the course’s rolling hills and uphill finish, the Tron is probably going to be the most popular choice since it’s a bit more nimble on the climbs, where every second will count.

See Speed Tests: Tron Bike vs Top Performers (Scatter Plot) for more nerd-level detail on frame and wheel performance.

More Route Recons

Many recon rides are now being planned each week on the upcoming ZRL route. If you’re unfamiliar with this course, jump into an event and do some recon! Here’s a list of upcoming Rolling Highlands events.

Si Bradeley

Strategic Options

How will the race unfold, and what strategies will riders employ in the first race of the season? Here’s what we predict:

• Breakaways galore: racers are learning that Pack Dynamics 4.1 slows down the peloton and greatly increases the chances of breakaways surviving. Every category should expect repeated attempts to establish a breakaway, and all the team tactics that go along with it.
• Sneak attacks: the small climbs on this loop lend themselves well to a coordinated breakaway attempt from 2-4 teammates. If that happens, everyone else will have to decide between chasing or letting it go. And with Pack Dynamics 4.1, if you let a group of teammates go, you may never see them again…
• Going long on the Corkscrew Castle Climb: the sneaky Corkscrew Climb begins just 2km from the finish line on the last lap. Surely someone will try to go long… for season opener glory!
• Saving it all for the Brae: in the end, every single point will be decided on the final ride up Breakaway Brae. Many racers will simply ride as efficiently as possible apart from this final short climb.

Your Thoughts

Any insights or further thoughts on this race? Share below!

How the Race Was Lost: The Breakaway That Got Away

I’ve heard many stories of solo breakaways winning Zwift races since PD4.1 went live. Here’s a fairly dramatic one, told from the perspective of someone (myself) who was merely pack fodder as the victor stormed to the line in a glorious, long-lasting solo break.

The race was KISS’ regular Wednesday event at 7am local time. I was looking for a longer race, as I’m growing a bit tired of the sub-half-hour ZRacing events. This one would take just about an hour, which sounded lovely. Four laps of New York City’s Park Perimeter Loop. Let’s race!

Lap 1

19 B riders jumped out of NYC’s start pens, and we quickly settled into a rhythm. Everyone seemed to know that the first sparks would fly on the Harlem Hill climb 3km into the lap, as this is the first and largest climb of the Park Perimeter Loop. 670m long at 5.1%, it takes a little under 90 seconds to finish in most B packs.

Someone always gets dropped on Harlem Hill, but there are also three more short climbs just afterward (the Three Sisters) where you can get tailed off the back if someone attacks and the pack follows. So Harlem Hill is really more like a hard 90 seconds that kicks off another ~3 minutes of solid work on the hilly back half of the Park Perimeter Loop.

We dropped 8 riders on the first Harlem Hill effort, and I put in some work on the front to help our group stay away. But our group didn’t want to work hard to stay away, probably because they reasoned that anyone dropped the first time up Harlem Hill wouldn’t survive the next 3 attempts.

So the chasers caught us, and we were a group of 17 as we finished up the lap and joined up with a straggling A rider or two.

1 lap down. 3 to go.

Lap 2

We recovered for a few minutes, then the group hit Harlem Hill a bit harder than the first lap. I was able to hang in with the front, but as we made our way over the Three Sisters, Danish rider “Martin Wang” attacked.

With my legs tired from the hills we were still in the middle of, I didn’t feel like chasing. There was an A rider up the road, and I just hoped he wouldn’t work with Wang to stay away.

Fortunately, Wang rode right past that A rider. Unfortunately… Wang seemed to have really good legs! He kept putting in a steady effort around 4W/kg, and the gap widened to 18s by the end of lap 2.

Lap 3

On the third time up Harlem Hill, I thought my goose was cooked. Twice!

The front group of 6 was 2 seconds ahead of me by the end of the climb, and I figured the front of the race was riding away. But 5 riders from behind joined me, and a few kept the pace up so we stayed within striking distance of the front 6.

Then, just when I thought the two groups were joining… I fell off the back! I knew if I eased up now I wasn’t coming back, so I decided to give it one last go. Shifting to a tougher gear, I got out of the saddle and went hard to catch the group. And made it!

Note to self: Pack Dynamics 4.1 has a shorter draft “fall-off”, meaning you can’t dangle off the back of the pack as easily as before.

Near the end of this lap a couple of riders from our group tried to bridge up to Wang, but nobody made it. Wang was still 17s ahead at the end of lap 3.

Lap 4

After nearly getting dropped (twice) on lap 3, I knew I wasn’t contending for the win today. Instead, my goal became hanging with this front pack to the finish line.

A couple of halfhearted attempts were made to work together to catch Wang, but they came to nothing. Was everyone tired like me? Did they think we would catch Wang automagically, as in Zwift races of yore?

Wang began extending his gap. Every time I looked at his power I felt like it said 4.2W/kg. “He must be cheater,” I said, comforting myself in my glycogen-depleted state. “Surely no legit B rider can stay away for 2.5 laps…”

The last Harlem Hill was fairly tame – I think riders were keeping their powder dry for the finish. With 5k to go, Wang had 26s on the pack. He wasn’t coming back.

With 700 meters left some riders from our group jumped. I just tried to surf the wheels, because I didn’t have much left. (Probably shouldn’t have skipped breakfast). I sat up as I finished last in the front group for 8th place. Wang? He took first by 10 seconds. Well done, sir.

See Strava Activity >
See ZwiftPower Results >

Watch the Video

Takeaways

My first takeaway is: don’t do hourlong fasted races. 20-30 minutes I can handle just fine, provided I had some carbs the day before. But I seem to run out of juice somewhere in that 30-60 minute window when I’m pushing hard in a fasted state.

The second (and bigger) takeaway is that our group’s lap times were remarkably slow. Looking at my Park Perimeter Loop history, our lap times were 30-60s slower than past races. I’ve done social rides here that were faster than this race!

It just proves what we’ve already been saying: Pack Dynamics 4.1 slows packs significantly, to the tune of 1.5-2 kilometers per hour slower. And that’s why Wang could pull off a well-timed attack and stay away for more than 35 minutes to take the solo win!

Wang’s power, in the end: 4.2W/kg for 20 minutes (highest in the race). 3.7W/kg average (302W), also the highest in the race, but not by much. His numbers were well within Zwift’s category enforcement range for B category – he just rode brilliantly and used his strengths well.

It’s a new day in Zwift racing land, and breakaway riders are loving it. Riders in the pack will have to learn to truly work together if they want to pull anything back. Things are going to be interesting!

Questions or Comments

Have you won from a breakaway with PD4.1 yet? Got questions or comments? Post below!