Playing a Flamme Rouge Grand Tour

In May 2026, I completed an eleven-stage Grand Tour in Flamme Rouge to evaluate the new Specialist Riders across diverse terrains and to run a live trial of my custom-built companion app. I wanted to observe exactly how asymmetric rider abilities alter the peloton’s dynamics and test whether the software effectively streamlines the race management.

I played as the white Visma-Lease a Bike team (Jonas Vingegaard and Wout van Aert), competing against a full peloton of five bot teams.

Here is my full review, including the match breakdown, the race results, an analysis and my observations.

This text is part of a three-part series:

The Game

My App

The Grand Tour

The Specialist Riders

Sprinteur Jasper Philipsen (Super Sprinteur): Recognized as one of the fastest men globally in mass sprints.

Rouleur Mathieu van der Poel (Flandrien): The premier specialist for cobblestones and demanding one-day races.

Sprinteur Mads Pedersen (Flahute): A power sprinter who excels particularly under adverse weather conditions.

Rouleur Lennard Kämna (Grimpeur): Utilizes his strengths as a climber and a calculated breakaway specialist.

Sprinteur Sam Welsford (Descender): Noted for his aerodynamic sprint tuck; is regarded as an elite descender.

Rouleur Filippo Ganna (Super Rouleur): Multi-time Time Trial World Champion.

Sprinteur Wout van Aert (Polyvalent): All-rounder capable of winning sprints, time trials and providing support in the mountains.

Rouleur Jonas Vingegaard (Grimpeur): A specialist dedicated to securing the General Classification.

Sprinteur Fabio Jakobsen (Super Sprinteur): A traditional finisher designed for flat stage arrivals.

Rouleur John Degenkolb (Flandrien): An experienced pavé specialist who frequently operates as the squad’s road captain.

Sprinteur Marius Mayrhofer (Polyvalent): Widely considered the top German talent for fast finishes.

Rouleur Julian Alaphilippe (Puncheur): The quintessential explosive attacker for short, steep climbs.

The Teams

Alpecin-Premier Tech

This squad presents one of the most formidable combinations for flat finishes and classic arrivals.

Lidl-Trek

Through strategic new signings, this team has evolved into a versatile force across all terrains, with a pronounced focus on its German contingent.

INEOS Grenadiers

The British outfit relies on a synergy of exceptional raw speed and technical precision.

Visma | Lease a Bike

The Dutch equipe merges multi-disciplinary versatility with absolute dominance in the high mountains.

Team Picnic PostNL

This roster brings together two veterans who perfectly embody a classic tactical division of labor.

Tudor Pro Cycling Team

The Swiss project has seen significant reinforcement for the 2026 season, specifically targeting the classics specialist department.

The Grand Tour 2026 Teardown

General Classification

Stage Results

Stage 1

Stage 1 Results (Team Time Trial) + General Classification:

Stage 2

Stage 2 Results:

General Classification after stage 2:

Stage 3

Stage 3 Results:

General Classification after stage 3:

Stage 4

Stage 4 Results:

General Classification after stage 4:

Stage 5

Stage 5 Results:

General Classification after stage 5:

Stage 6

Stage 6 Results:

General Classification after stage 6:

Stage 7

Stage 7 Results:

General Classification after stage 7:

Stage 8

Stage 8 Results:

General Classification after stage 8:

Stage 9

Stage 9 Results:

General Classification after stage 9:

Stage 10

Stage 10 Results:

General Classification after stage 10:

Stage 11

Stage 11 Results:

General Classification after stage 11:

Tracking the Specialists: Stage-by-Stage Performance

Analysis: What I learned

What a blast! At the heart of the Grand Tour I played was a phenomenal duel between two giants from my dominant Visma | Lease a Bike team. We witnessed an epic battle between the climbing specialist (Grimpeur) Jonas Vingegaard and the incredibly versatile powerhouse (Polyvalent) Wout van Aert. The Yellow Jersey changed hands multiple times throughout the race. From van Aert’s early dominance to John Degenkolb’s surprise two-stage stint in the overall lead as a classic Flandrien, the tension never dropped.

But it all came down to a nail-biting finale on the final mountain stage! In a spectacular display of endurance, Vingegaard snatched the ultimate victory, securing the Yellow Jersey by a mind-blowing margin of just 2 seconds!

The fierce competition extended to the other classifications as well. Van Aert rewarded his consistent brilliance by taking home the Green Jersey with 10 sprint points, while the explosive Puncheur Julian Alaphilippe conquered the grueling climbs to claim the Polka Dot Jersey with 13 mountain points.

Whether you’re a die-hard cycling fan or a newcomer to the peloton, this Grand Tour was a perfect showcase of strategy, grit, and the incredible dynamic between Sprinteurs and Rouleurs.

Bots tend to dominate the stage beginning

Let’s take a look of how the bot teams performed!

If you have raced against bots over a multi-stage tour, you might have noticed a specific pattern: they tend to absolutely dominate the first half of a stage, only to completely bonk and fall apart in the finale.

After running 11 stages of a Grand Tour with an early version of the app, I dug into the math behind the decks to figure out why this happens. The core issue lies in how the red an yellow movement points—the triggers that let a bot draw a second card—are distributed. In the original design, these colors are often attached to the highest value cards. Early in the race, when the deck is perfectly clean, the bots will constantly hit these triggers. They draw two cards, filter out the weaker ones, and burn through their top speeds immediately. But once the mid-game hits and their draw piles get clogged up with Exhaustion cards, all those valuable suffixes are sitting in the discard pile. The bots have no lifeline left to skip a bad draw, leading to a massive drop in performance just when the sprint to the finish line starts.

To fix this for the new Medium and Hard difficulty levels of my app, I didn’t just blindly inflate the numbers. Instead, I re-engineered the trigger system. I stripped the red and yellow values from the absolute peak cards and strategically moved them to the mid-tier values (like the 5s and 6s).

I think this changes the pacing. In the first half of the race, the bots now play their top speeds straight up without filtering their deck too fast. The start is slightly more controlled. But in the crucial final stretches, when their decks are heavily diluted with Exhaustion, drawing a mid-tier card now triggers a second draw. This gives the bots a much higher chance to bypass an Exhaustion card right when it matters most, keeping them consistently competitive until the very end.

More realistic bot movements

I also realized that simply adding a flat +1 movement to every card to create a „Hard“ mode breaks the game. If a bot gets more than a dozen extra movement points per deck cycle, no human player managing standard cards and slipstreams can catch them over a 70-square stage.

Because of this, the Easy level remains Fredrik Stahre’s balanced original baseline. For Medium and Hard, the base speeds of the bots remain grounded so the peloton doesn’t instantly shatter. The difficulty scales entirely through targeted tactical spikes. The bots get significantly higher peaks on their specific terrain—like a Super Sprinteur hitting a massive 13 on the final straight, or a Grimpeur surviving a steeper Uphill limit—making them incredibly dangerous exactly where their real-world counterparts would attack.

Check out my app