A remarkable account of Tadej Pogačar’s history-making Milan-San Remo victory suggests the world champion did not realise his Colnago Y1Rs cracked in the crash at the foot of the Cipressa climb, a UAE Team Emirates mechanic reporting they only found the damage after the finish.
It would mean Pogačar rejoined the front of the race, attacked multiple times and bombed down the descents of the Cipressa and Poggio, putting Tom Pidcock under pressure on the latter, before outsprinting the Brit to claim a maiden title at the opening Monument of the season, all aboard a Colnago Y1Rs with a cracked frame.

While the pictures and TV images from the day are inconclusive, Pogačar’s mechanic Boštjan Kavčnik told Slovenia’s national daily newspaper Delo the crash caused a crack to his “rear stays” and the world champion undoubtedly “wouldn’t have descended so aggressively” if he had “known what condition the bike was in”.
“He won, even though his frame cracked in the crash and the disc was rubbing against the braking surface,” Kavčnik explained.
“The rear stays were damaged, but fortunately they didn’t fail. If Tadej had known what condition the bike was in, he wouldn’t have descended so aggressively that even Tom Pidcock could barely follow him. This bike will now go into Tadej’s special collection, even though it’s damaged.”
The UAE Team Emirates mechanic, who accompanies Pogačar at all his races, also suggested his rear derailleur went into crash mode in the fall, even with the impact on the non-drive side.

“Tadej himself got it back into operation; he didn’t feel anything else was wrong, so we didn’t change the bike,” Kavčnik continued, before explaining a couple of technical tweaks they made ahead of the race, namely switching from 30mm tyres to 28mm and opting for a 1x set-up.
It is believed to be the first time Pogačar has raced a single chainring set-up for a road race, the team apparently tempted by the weight and aerodynamic savings on offer.

“He also used deeper rims with narrower tyres. Last year they were 30mm, this year 28mm. Our performance department calculated that this would be best for such a fast race.”
The preparation for Pogačar’s second tilt at Paris-Roubaix has long since been underway. After Saturday’s success, it is the final Monument left for the greatest cyclist of the modern era (some would argue any era) to win.
Kavčnik confirmed Pogačar will race this year’s edition on the Y1Rs, rather than the V5Rs which he rode to second place last time. The team has been testing numerous tyre and rim combinations throughout the winter and another recon is scheduled on the Roubaix cobbles in the week before the Tour of Flanders.

5 thoughts on “Did Tadej Pogačar win Milan-San Remo on a cracked Colnago?”
I can’t believe the author never mentioned Lance Armstrong’s stage win in the 2003 TDF at Luz Ardiden on a bike with a (maybe worse) cracked chainstay. Near the bottom of the climb, Lance fell due to hooking a musette bag that a fan was holding and Iban Mayo rode straight into his chainstay, cracking it (you can probably find many YouTube videos). Lance remounted, almost immediately crashed again as the bike lurched around when he put the power down, but then managed to nurse it home and win at the top. I’m not a Lance fanboy, but it’s an interesting bit of nerdy history for us bike mechanics.
No actual picture of the crack?
FWIW, I cracked the rear triangle of a dual suspension mtb and didn’t know until I took the bike in to get a creaking pedal serviced and the mechanic thought the crack might have been the noise.
My point is that carbon frames are pretty robust and a minor crack not be such a big issue.
Did he actually run with a 1x chainring? on a dura-ace? i didn’t know that was feasible as a shimano backed brand (sure you could technically do that, but i am surprised they did it on an offical bike).
Many riders have been doing the same on time trials for a few years. I’m not aware of Shimano telling them not to. It makes sense when you can average about 38 km/h on a climb with a 4% average gradient – no small ring needed.