Comprehensive Climbers Ranking

[Peyresourde]

Despite reading this forum for several years, I have never posted anything until now. Judging from the discussions, many posters are interested in comparing climbers and climbing performances. How much better than the rest is Pogacar really? Who is the better climber between Evenepoel, Roglic and Almeida? (Spoiler: it is quite close according to my data). Are perfomances now better than in the 90s?

Over the last months, I have compiled over 2000 climbing performances from the 1950s until now. To make them easily comparable, I used 3 steps:
1) Calculating the Watt/kg . I use the formula provided by James Martin et al and 60kg standard weight like Lanternerouge. Of course I have tried to era-adjust parameters like CDA, CRR and equipment weight as best as possible, but there is always a margin of error.

2) Based on the W/kg and the time, I assign a raw Index for each performance. An Index of 100 equals an extremely strong performance (currently the Top 24 all time are over 100). E.g. 7 W/Kg for 25 minutes or 6.66 W/Kg for 40 minutes would equal a 100 Index.
1 Point difference equals 0.025 W/kg, so 4 Points = 0.1 W/kg and 10 Points = 0.25 W/kg.
If you want to think in trendlines, the purple trendline on Lanternerouge.com is around a 95 Index, while the red trendline would be 85 (0.25 W/kg less).

3)Then I adjust the raw Index for the following factors: Stage Hardness (170km and 3000m elevation gain before climb would be around average), Altitude, Approach to the climb (are you climbing already before the segment or is the approach a false flat downhill?), regularity, follow-up (is it a mountain-top finish or still 50 km to the finish) and weather (plus points for rain and temperatures over 30 °C in the valley). Overall, I try to adjust for most relevant factors, but rather too little than too much.



Example 1: Pogacar in Tirreno Adriatico 2021 pushed 6.45 W/kg for 36:06 on Prati di Tivo. (Wout van Aert also did his best ever W/kg there) An extremley impressive performance for 2021 standards, it equals a raw performance Index of 89. But the stage before was easy (-2), the approach is basically 25km without having to pedal (-3), the climb is very regular and has a very clear start and finish (-2), and also slighly low altitude (-1).
So the final Index would be 81 (-8). A good performance (equal to Chris Froome's PB), but nothing super special as the time gaps on the climb also indicate.

Example 2: Farrapona, Vuelta 2014 Stage 16. Probably the best performance of Froome and post-ban Contador, but in pure W/kg nothing special: Contador 6.59 W/kg for 17:43, Froome 6,58 W/kg for 17:58 (less draft) --> raw Index 74 for both. Adjustment : Altitude (+3), Stage Hardness (+2), Approach (+3), Regularity (-1). Final Index 81 (+7). The time gaps are big for a relativley short climb and this performance was also not beaten on the same climb recently in the 2025 Vuelta, showing that it was very good.

Fun fact: Biggest adjustments: (-14) for some pure time trials and (+28) for Claudio Chiappucci, 1992 Tour de France Stage 13 to Sestriere.

Disclaimer: I also include some very short efforts and shallower climbs like Cipressa. Of course, the uncertainty for these efforts is much higher, but they are too intersting not to include. Also, because the adjustments are linear and not percentage-based, they do not work 100% perfectly for very short and very long efforts. (Adjustments too low or too high respectively).
I only consider performances as they happened on the road, please no Clinic talk here.



I plan to publish 3 lists here, but I will not post everything at once to encourage a nice discussion. I welcome feedback, guesses, questions or tips .

1) Top 24 Performances of all time (Every performance of 100 or higher + some honourable mentions. You will see that the top 25 are heavily occupied by 2 famous riders)

18 | Tadej Pogacar | 102 (+1): 7.79 W/kg for 8:57 on Cipressa (MSR 2025)
19 | Marco Pantani | 101 (-6): 6.91 W/kg for 36:20 on Piancavallo (Giro 1998 Stage 14)
20 | Evgeni Berzin | 101 (-2): 8.45 W/kg for 4:46 on Santuario de Oro (Euskal Bizikleta 1995 Stage 2) (very uncertain)
21 | Santiago Perez | 101 (+5): 7.20 W/kg for 16:25 on Navacerrada (Vuelta 2004 Stage 20)
22 | Tadej Pogacar | 100 (-1): 8.62 W/kg for 3:59 on La Redoute (LBL 2025)
23 | Tadej Pogacar | 100 (0): 8.56 W/kg for 4:07 on La Redoute (LBL 2024) (very strong headwind)
24 | Tadej Pogacar | 100 (+4): 6.77 W/kg for 30:20 on Sormano (Il Lombardia 2024)

HM | Tadej Pogacar | 99 (+1): 6.71 W/kg for 35:09 on Hautacam (Tour 2025 Stage 12)
HM | Alberto Contador | 98 (+1): 7.07 W/kg for 20:40 on Verbier (Tour 2009 Stage 15)
HM | Lance Armstrong | 98 (+5): 6.53 W/kg for 38:03 on Alpe d'Huez (Tour 2001 Stage 10)
HM | Ivan Basso | 97 (-5): 6.62 W/kg for 46:16 on Monte Bondone (Giro 2006 Stage 15)



2) Top 50 climbers of all time based on a weighted average of their top 10 performances (The best performances is worth roughly 2.5 times more than the 10th)

3) Best performance per year since 1986 (and a few individual years before where I have good data).

Notes: I have been inspired mainly be lanternerouge.com, watts2win.eu, chronoswatts.com and climbing-records.com. I calculated all the Watts myself, but the segment data is often from these and other sources, so I would like to credit them. As I did a lot of this work manually, there will be some mistakes and of course no method can be perfect. I also have not video studied every second of every climb, so things like drafting, wind etc. are always estimates that lead to uncertainties. That said, I hope to give a basis for discussions and hopefully a new standard with which to compare climbing performances. Also, if anyone wants to know how good a specific performance (of their favourite rider) was, I will gladly answer these questions.

 

[hayneplane]

Given some of the shorter efforts listed I am intrigued to see whether Pogacar on Oude Kwaremont will be included and what the compensatory value might be for cobbles.

 

[Peyresourde]

It is a bit of a coincidence that there are a few short efforts in the first entries. I have calcaluated the Oude Kwaremont and have spent quite a bit of effort on it. The problem is that Pogacar never went all out for the whole climb in 2025. For cobbles, I simply increase the rolling resistance (CRR), but I am also unsure by how much would be appropriate. I have a feeling that riders cannot push 100 % of their power on cobbles because it is so shaky.

The best effort on Kwaremont (cobbles only) from 2025 is 82 (+7): 8.82 W/Kg for 2:35. But it is extremely uncertain and Pogacar also visually did not go 100% in the flatter middle part.

 

[houtdffan]

did you mean to post the entire top 24? Only 19 through 24 appeared

 

[Dazed and Confused]

did you mean to post the entire top 24? Only 19 through 24 appeared

I plan to publish 3 lists here, but I will not post everything at once to encourage a nice discussion. I welcome feedback, guesses, questions or tips

 

[Brullnux]

Despite reading this forum for several years, I have never posted anything until now. Judging from the discussions, many posters are interested in comparing climbers and climbing performances. How much better than the rest is Pogacar really? Who is the better climber between Evenepoel, Roglic and Almeida? (Spoiler: it is quite close according to my data). Are perfomances now better than in the 90s?

Over the last months, I have compiled over 2000 climbing performances from the 1950s until now. To make them easily comparable, I used 3 steps:
1) Calculating the Watt/kg . I use the formula provided by James Martin et al and 60kg standard weight like Lanternerouge. Of course I have tried to era-adjust parameters like CDA, CRR and equipment weight as best as possible, but there is always a margin of error.

2) Based on the W/kg and the time, I assign a raw Index for each performance. An Index of 100 equals an extremely strong performance (currently the Top 24 all time are over 100). E.g. 7 W/Kg for 25 minutes or 6.66 W/Kg for 40 minutes would equal a 100 Index.
1 Point difference equals 0.025 W/kg, so 4 Points = 0.1 W/kg and 10 Points = 0.25 W/kg.
If you want to think in trendlines, the purple trendline on Lanternerouge.com is around a 95 Index, while the red trendline would be 85 (0.25 W/kg less).

3)Then I adjust the raw Index for the following factors: Stage Hardness (170km and 3000m elevation gain before climb would be around average), Altitude, Approach to the climb (are you climbing already before the segment or is the approach a false flat downhill?), regularity, follow-up (is it a mountain-top finish or still 50 km to the finish) and weather (plus points for rain and temperatures over 30 °C in the valley). Overall, I try to adjust for most relevant factors, but rather too little than too much.



Example 1: Pogacar in Tirreno Adriatico 2021 pushed 6.45 W/kg for 36:06 on Prati di Tivo. (Wout van Aert also did his best ever W/kg there) An extremley impressive performance for 2021 standards, it equals a raw performance Index of 89. But the stage before was easy (-2), the approach is basically 25km without having to pedal (-3), the climb is very regular and has a very clear start and finish (-2), and also slighly low altitude (-1).
So the final Index would be 81 (-8). A good performance (equal to Chris Froome's PB), but nothing super special as the time gaps on the climb also indicate.

Example 2: Farrapona, Vuelta 2014 Stage 16. Probably the best performance of Froome and post-ban Contador, but in pure W/kg nothing special: Contador 6.59 W/kg for 17:43, Froome 6,58 W/kg for 17:58 (less draft) --> raw Index 74 for both. Adjustment : Altitude (+3), Stage Hardness (+2), Approach (+3), Regularity (-1). Final Index 81 (+7). The time gaps are big for a relativley short climb and this performance was also not beaten on the same climb recently in the 2025 Vuelta, showing that it was very good.

Fun fact: Biggest adjustments: (-14) for some pure time trials and (+28) for Claudio Chiappucci, 1992 Tour de France Stage 13 to Sestriere.

Disclaimer: I also include some very short efforts and shallower climbs like Cipressa. Of course, the uncertainty for these efforts is much higher, but they are too intersting not to include. Also, because the adjustments are linear and not percentage-based, they do not work 100% perfectly for very short and very long efforts. (Adjustments too low or too high respectively).
I only consider performances as they happened on the road, please no Clinic talk here.



I plan to publish 3 lists here, but I will not post everything at once to encourage a nice discussion. I welcome feedback, guesses, questions or tips .

1) Top 24 Performances of all time (Every performance of 100 or higher + some honourable mentions. You will see that the top 25 are heavily occupied by 2 famous riders)

19 | Marco Pantani | 101 (-6): 6.91 W/kg for 36:20 on Piancavallo (Giro 1998 Stage 14)
20 | Evgeni Berzin | 101 (-2): 8.45 W/kg for 4:46 on Santuario de Oro (Euskal Bizikleta 1995 Stage 2) (very uncertain)
21 | Santiago Perez | 101 (+5): 7.20 W/kg for 16:25 on Navacerrada (Vuelta 2004 Stage 20)
22 | Tadej Pogacar | 100 (-1): 8.62 W/kg for 3:59 on La Redoute (LBL 2025)
23 | Tadej Pogacar | 100 (0): 8.56 W/kg for 4:07 on La Redoute (LBL 2024) (very strong headwind)
24 | Tadej Pogacar | 100 (+4): 6.77 W/kg for 30:20 on Sormano (Il Lombardia 2024)

HM | Tadej Pogacar | 99 (+1): 6.71 W/kg for 35:09 on Hautacam (Tour 2025 Stage 12)
HM | Alberto Contador | 98 (+1): 7.07 W/kg for 20:40 on Verbier (Tour 2009 Stage 15)
HM | Lance Armstrong | 98 (+5): 6.53 W/kg for 38:03 on Alpe d'Huez (Tour 2001 Stage 10)
HM | Ivan Basso | 97 (-5): 6.62 W/kg for 46:16 on Monte Bondone (Giro 2006 Stage 15)



2) Top 50 climbers of all time based on a weighted average of their top 10 performances (The best performances is worth roughly 2.5 times more than the 10th)

3) Best performance per year since 1986 (and a few individual years before where I have good data).

Notes: I have been inspired mainly be lanternerouge.com, watts2win.eu, chronoswatts.com and climbing-records.com. I calculated all the Watts myself, but the segment data is often from these and other sources, so I would like to credit them. As I did a lot of this work manually, there will be some mistakes and of course no method can be perfect. I also have not video studied every second of every climb, so things like drafting, wind etc. are always estimates that lead to uncertainties. That said, I hope to give a basis for discussions and hopefully a new standard with which to compare climbing performances. Also, if anyone wants to know how good a specific performance (of their favourite rider) was, I will gladly answer these questions.

Where is christian scaroni ?

 

[YavorD]

Where is christian scaroni ?

I think he is still in Oman.
The best is yet to come though, he is preparing to stomp the competition (and this ranking) by some Pantani-esque watts on Piancavallo and bring joy to all the cycling fans globally.

 

[Red Rick]

Can't be *** about W/kg estimates - I'd much prefer an elo style system where you try to make a normalized average gap to the fastest climber on a given day. The question is how you pool climb data, how you deal with outliers, and how you account for data points where riders are not present.

If you only want to care about W/kg, you're just trying to select for outlier results where the whole top 10 did PB numbers, which makes it a completely redundant exercise.

 

[Peyresourde]

Where is christian scaroni ?

I don't think it is a big spoiler to tell you that he does not quite make the Top 50 yet. His performance from today is 73 (-5): 6.67 W/Kg for 18:20

Can't be *** about W/kg estimates - I'd much prefer an elo style system where you try to make a normalized average gap to the fastest climber on a given day. The question is how you pool climb data, how you deal with outliers, and how you account for data points where riders are not present.

If you only want to care about W/kg, you're just trying to select for outlier results where the whole top 10 did PB numbers, which makes it a completely redundant exercise.

I know you are not convinced by W/kg calculations. But with my adjustments the whole data set is very consistent and I would say there are barely any outliers and those that exist can be explained quite well.
And I think there is not a single climb where more than 5 people did their PB if we exclude very old performances from riders whome I barely have any data from or some pure mountain time trials.

 

[Red Rick]

I don't think it is a big spoiler to tell you that he does not quite make the Top 50 yet. His performance from today is 73 (-5): 6.67 W/Kg for 18:20

I know you are not convinced by W/kg calculations. But with my adjustments the whole data set is very consistent and I would say there are barely any outliers and those that exist can be explained quite well.
And I think there is not a single climb where more than 5 people did their PB if we exclude very old performances from riders whome I barely have any data from or some pure mountain time trials.

Most algorithms that standardize everything on profile will still have a problem with this because it will miss the actual racing strategy or hidden variables that make a climb more or less friendly to high performance.

Any algorithm will also have riders fluctuate enormously in index performances from day to day, which doesn't make that much sense.

Fatigue will also affect riders differently, so it will be overestimating fatigue for the very top riders on a climb while underestimating it for lesser riders.

I don't know your data, but I know for example W2W had Finestre in the Giro last season produce 9 PBs and an SB. Another mass notorious mass overperformance was Piancavallo 2020.

 

[Peyresourde]

Most algorithms that standardize everything on profile will still have a problem with this because it will miss the actual racing strategy or hidden variables that make a climb more or less friendly to high performance.

Any algorithm will also have riders fluctuate enormously in index performances from day to day, which doesn't make that much sense.

Fatigue will also affect riders differently, so it will be overestimating fatigue for the very top riders on a climb while underestimating it for lesser riders.

I don't know your data, but I know for example W2W had Finestre in the Giro last season produce 9 PBs and an SB. Another mass notorious mass overperformance was Piancavallo 2020.

Your points are all very valid. The adjustments are standardized, so of course they will be inaccurate for certain rider who can only perform after low kilojoules. Especially until the 80s basically no one could perform outside of time trials, while today almost anyone can perform decently after hard stages, so the historic comparison is also not clean. And strategy etc. is of course not included.

On Finestre, W2W has a completely unrealistic adjustment of +5. It is not high altitude (start at 400m) and after 160km of soft pedalling. My adjustment there is -2 and no one except Yates has a PB. W2W has a tendency to overestimate long climbs anyway (probably because the altitude effect is indirectly included twice in the formula).
As for Piancavallo, this climb seems to have a secret tailwind or something, even Mikel Landa did his best performance of the 2010s there in 2017.


This entry deserves a seperate post because it will be very relevant to anlayze for this years Milano Sanremo and how difficult it is to drop MvdP. Needless to say the watts here could be very inaccurate (but very plausible) with an average speed of 38 kph. But the W/kg itself are not important here, the differences between Pogacar and van der Poel are.

18 | Tadej Pogacar | 102 (+1): 7.79 W/kg for 8:57 on Cipressa (MSR 2025)
Matthieu van der Poel | 87 (+1): 7.40 W/kg for 8:58 on Cipressa (MSR 2025)

So the Pogacar pushed around 0.4 W/kg more than MvdP without dropping him. But remember that these are standardized for 60kg. This is a very shallow climb where absolute watts also matter (this is the reason why Ganna is good here). If we assume Pogacar (66kg) and MvdP (75kg), then in real W/kg:

Pogacar: ~7.52 W/kg Absolute Watts: 496 W
MvdP: ~ 6.95 W/kg Absolute Watts: 521 W

So MvdP is almost 10kg heavier, but only has to push 25 Watts more over the whole climb. Now from various performances, we can estimate that MvdP's FTP is only about 0-15 Watts higher than Pogacar, so it is not completely impossible for Pogacar.
But you also have to consider that this analysis is for the whole climb. If you only consider Pogacar's attacks, then van der Poel has to push less absolute Watts than Pogacar at that moment as the draft benefit at 45 kph+ is so huge. Considering he is more explosive, it should be close to impossible to drop MvdP if circumstances are similar in 2026.

 

[Pozzovivo]

Quality posts @Peyresourde! Credit to you for trying to bring it all together. I have been quite skeptical of comparing "performances" across different races, while I can aknowledge that the W/kg estimates can be fairly accurate.

For me, the problem is what Red Rick speaks about earlier, something akin to multicollinearity. When one rider "perform" very well, they all tend to do so. This indicates to me that something about the racing conditions contribute substantially to the range of performances that was possible on that day, and comparisons between the races are very nosiy.

For example, can you help me understand how come the W/KG is so much better on PDB (S15, 24'TDF), compared to the day before? Am I correct in stating that, and if so how does that make sense to you?

 

[Tim Cahill]

Berzin was Stage 5 no? Could be getting mixed up, but all the old Spanish races from the 90s are on Pablo MJs Youtube channel, what makes that performance incredible is how hard they were clearly going beforehand, in an absolutely crazy stage. That might be the best hilly team of all time with the obvious caveats.

So many moving parts just makes it almost impossible especially without direct access to the data, people embedded within teams probably have their own models but that will be limited to their own riders, has to start somewhere I suppose.

 

[CyclistAbi]

Based on my calculations Primož Roglič came on top, historically speaking. Then some of the other names mentioned followed so good job on that part.

 

[Peyresourde]

The other Berzin performance from stage 5 is still coming. This was indeed a crazy performance, he already went incredibly hard the second last climb. Also the winner of the stage Frattini (total one hit wonder) and Zülle did amazing performances that day (both 99 and 98).

For me, the problem is what Red Rick speaks about earlier, something akin to multicollinearity. When one rider "perform" very well, they all tend to do so. This indicates to me that something about the racing conditions contribute substantially to the range of performances that was possible on that day, and comparisons between the races are very nosiy.

For example, can you help me understand how come the W/KG is so much better on PDB (S15, 24'TDF), compared to the day before? Am I correct in stating that, and if so how does that make sense to you?

Yes, this phenomenon indeed exists, especially on longer climbs. Most of it is a pacing issue in my oppinion, but a lot of riders just do not/ are not psychologically capable of going all out on (shallower) climbs if the certain conditions are not met. And although everyone is different, humans also have many similarities and in races the riders do similar efforts day after day. So it is also not unlogical that on some days conditions are met where many perform, while on some days many perform badly.

Pla d'Adet and PdB are very interesting. Actually, only the top 4 (Pogacar, Vingegaard, Evenepoel, Landa) performed better on Pdb. The rest of the Top 10 was better on Pla d'Adet the day before.
Pogacar in that Tour could simply hold close to 7 W/kg for extended periods, but the first 60 % of Pla d'Adet were paced far too slow for him (and also Vingegaard) to make up for after his attack. Just because the first part is paced slower does not mean he can push 8 W/kg the rest of the climb. A hard attack also takes something out of you (he did like 9 W/kg for 2 minutes in the middle of the climb)

So on stage 14 the pacing was perfect for place 4-10. This climb also has 2 small descents which makes calculating the watts (and also the real performances) harder/lower.

On stage 15 the Visma pacing was hard enough to tire everyone but the top 4 before the final climb, so they performed worse. Then on the climb it was the most perfect pacing in the history of cycling and Pogacar even went 110% to the line instead of celebrating.
Additionally, it was not as big of an outlier as some may think. The Lanterne Rouge guys e.g. got the wind wrong (they mentioned a slight headwind when it was a tailwind in reality). It was also not one of Pantani's better records. His form in the 98' Tour was not that great even though he won.

And one performance simply has to be the best, and it is statistically normal that the best performance is 0.1 W/kg better than the second best.

 

[Red Rick]

The other Berzin performance from stage 5 is still coming. This was indeed a crazy performance, he already went incredibly hard the second last climb. Also the winner of the stage Frattini (total one hit wonder) and Zülle did amazing performances that day (both 99 and 98).


Yes, this phenomenon indeed exists, especially on longer climbs. Most of it is a pacing issue in my oppinion, but a lot of riders just do not/ are not psychologically capable of going all out on (shallower) climbs if the certain conditions are not met. And although everyone is different, humans also have many similarities and in races the riders do similar efforts day after day. So it is also not unlogical that on some days conditions are met where many perform, while on some days many perform badly.

Pla d'Adet and PdB are very interesting. Actually, only the top 4 (Pogacar, Vingegaard, Evenepoel, Landa) performed better on Pdb. The rest of the Top 10 was better on Pla d'Adet the day before.
Pogacar in that Tour could simply hold close to 7 W/kg for extended periods, but the first 60 % of Pla d'Adet were paced far too slow for him (and also Vingegaard) to make up for after his attack. Just because the first part is paced slower does not mean he can push 8 W/kg the rest of the climb. A hard attack also takes something out of you (he did like 9 W/kg for 2 minutes in the middle of the climb)

So on stage 14 the pacing was perfect for place 4-10. This climb also has 2 small descents which makes calculating the watts (and also the real performances) harder/lower.

On stage 15 the Visma pacing was hard enough to tire everyone but the top 4 before the final climb, so they performed worse. Then on the climb it was the most perfect pacing in the history of cycling and Pogacar even went 110% to the line instead of celebrating.
Additionally, it was not as big of an outlier as some may think. The Lanterne Rouge guys e.g. got the wind wrong (they mentioned a slight headwind when it was a tailwind in reality). It was also not one of Pantani's better records. His form in the 98' Tour was not that great even though he won.

And one performance simply has to be the best, and it is statistically normal that the best performance is 0.1 W/kg better than the second best.

Do you have the Plat d'Adet before/after attack split?

I think Plat d'Adet had more recent fatigue compared to Plateau de Beille, with Tourmalet an Ancizan before the MTF without much flat inbetween, and that's also how you get tired domestiques.

Plateau de Beille is the sort of stage where you clearly had enough flats inbetween the climbs to recover, and the top 3 clearly recovered very well on the final flat before PdB to do a near peak performance, combined with a perfect pacing and tailwind.

To be honest, LR rouge claiming a headwind on Plateau de Beille is laughable itself, considering there were flags everywhere.

I'm also pretty sure positive splits are even better than perfectly even splits because of the air density difference over a large amount of altitude distance. Perfect Finestre had 6.3 W/kg in the first 33 minutes.

Another often forgotten factor is tarmac quality, especially if a climb is resurfaced to host a GT. now they tend to only distinguish between tarmac and gravel or hormigon style stuff.

 

[topcat]

In on page 1. I have nothing to contribute. I don't understand why you don't understand.

 

[Pozzovivo]

Thank you for your detailed answer, @Peyresourde.

You raise several interesting points, and I follow your reasoning for the most part. A few aspects I would like to clarify.

First, I agree that pacing is an under-communicated factor when comparing performances, and that it likely explains some of the variation we observe both between races and from day to day. The question, then, is how we account for pacing when evaluating performance. If support riders set and maintain a very high tempo, and this in turn shapes the outcome, what exactly are we comparing when we speak about “performance” across riders and races? It becomes somewhat unclear whether we are isolating individual capacity or race dynamics.

Second, the timing and structure of attacks clearly influence the time gaps we see at the finish, as you point out. However, if Pogacar executes a large positive split mid-climb on Pla d’Adet or in the opening kilometres of Hautacam, should that not count positively in an evaluation of performance level?

Third, you suggest that some riders are psychologically unable to produce their best power on shallow gradients. I am inclined to think this is primarily physiological rather than psychological. Differences in body position between shallow and steep gradients alter biomechanics, muscle recruitment, and efficiency, which in turn affect power output. Riders are not identical in this regard. Not sure about that.

Finally, even if the Visma pace disproportionately harmed the weaker climbers, should we not expect all riders to be negatively affected to some degree? It was an exceptionally demanding stage by modern standards. Fatigue is not categorical but continuous right? it should affect everyone, albeit to varying extents.

I agree that the tailwind likely inflated the PDB performance estimate. How much is really the difference between the performances then?

Thank you.

 

[Red Rick]

Thank you for your detailed answer, @Peyresourde.

You raise several interesting points, and I follow your reasoning for the most part. A few aspects I would like to clarify.

First, I agree that pacing is an under-communicated factor when comparing performances, and that it likely explains some of the variation we observe both between races and from day to day. The question, then, is how we account for pacing when evaluating performance. If support riders set and maintain a very high tempo, and this in turn shapes the outcome, what exactly are we comparing when we speak about “performance” across riders and races? It becomes somewhat unclear whether we are isolating individual capacity or race dynamics.

Second, the timing and structure of attacks clearly influence the time gaps we see at the finish, as you point out. However, if Pogacar executes a large positive split mid-climb on Pla d’Adet or in the opening kilometres of Hautacam, should that not count positively in an evaluation of performance level?

Third, you suggest that some riders are psychologically unable to produce their best power on shallow gradients. I am inclined to think this is primarily physiological rather than psychological. Differences in body position between shallow and steep gradients alter biomechanics, muscle recruitment, and efficiency, which in turn affect power output. Riders are not identical in this regard. Not sure about that.

Finally, even if the Visma pace disproportionately harmed the weaker climbers, should we not expect all riders to be negatively affected to some degree? It was an exceptionally demanding stage by modern standards. Fatigue is not categorical but continuous right? it should affect everyone, albeit to varying extents.

I agree that the tailwind likely inflated the PDB performance estimate. How much is really the difference between the performances then?

Thank you.

Hautacam should be highly rated because of the gaps behind, not the pacing strat. If anything. Everyone was *** on that climb and the hard start was so short it shoudln't really be an excuse for the rest to drop times that were bad by 2014 standards.

 

[Breh]

Just dropping by to say I appreciate the effort. It drives quality posts and discussion.
I'm not too knowledgable on the topic to be offended so I'll just enjoy the read.

 

[Pozzovivo]

Hautacam should be highly rated because of the gaps behind, not the pacing strat. If anything. Everyone was *** on that climb and the hard start was so short it shoudln't really be an excuse for the rest to drop times that were bad by 2014 standards.

I agree that a combination of timegaps and strength of competition is the key overall.

But, if we are trying to comment on the level of a performance based on climbing times, I would certaintly say that it is more impressive as a performance to do X w/kg if you did a massive positive split compared to an almeida style effort.

 

[Peyresourde]

@Red Rick
Tarmac quality is a factor that is quite difficult to estimate for me, but I think with newer tires it is slightly less important than before. It also does not make a gigantic difference unless the quality is really bad or speeds are really high. In the 90s the tarmac was often amazing in the grand tours. I have also gradually reduced the general CRR I use from 0.004 in the 2000s and 2010s to 0.0037 since 2024. Though I have no idea if that is enough.

If you are talking about air densitiy only in the sense of air resistance (and not oxygen uptake) then the most important factor is the gradient which changes often on most climbs. Riders already (subconciously) push harder on steeper parts and less on flatter parts. Making this the basis of your pacing plan is somewhat convulted in my oppinon.

I rate Hautacam from this year high although I have "only" given it a 99. It was the hottest day of the Tour (36°C). I give +2 for that. Everone was suffering and 80% of the Top 10 pushed less Watts than on Mont Ventoux a few days later (which is 20 minutes longer). But I prefer to adjust less than reality. I also adjust less for altitude than some studies suggest. Sometimes on the first mountain stage everone just underperforms and we think the winner has done the best performance in history when in reality they were just the only guys at their normal level (La Plagne 1995, Hautacam 2000, Pierre Saint Martin 2015) . I do not have a scientific explanation for that, so I cannot adjust for it.

As for pacing, climbs like Hautacam 2025 and Alpe d'Huez 2001 (~7.4 W/kg for the first 10 minutes) had extreme splits. Many of the Pantani climbs also have very positive splits. I have thought about adjusting for that, but it is a real hassle to calculate splits for every climb and also very difficult to find the right adjustment.

In my opinion, not every performance should and can be adjusted until they are all the same. I just aim to find the general peak and consistency level of riders. I will also publish the TOP 50 climbers list the next few days where I use 10 performances each. That will give you some more insights. E.g. since 2024 the performance gap between Evenepoel and Pogacar has been very consistent in my data. When they were both in good shape and went all out, the gap was always around 0.4 W/kg (or 16 points in my Index)

 

[Red Rick]

@Red Rick
Tarmac quality is a factor that is quite difficult to estimate for me, but I think with newer tires it is slightly less important than before. It also does not make a gigantic difference unless the quality is really bad or speeds are really high. In the 90s the tarmac was often amazing in the grand tours. I have also gradually reduced the general CRR I use from 0.004 in the 2000s and 2010s to 0.0037 since 2024. Though I have no idea if that is enough.

If you are talking about air densitiy only in the sense of air resistance (and not oxygen uptake) then the most important factor is the gradient which changes often on most climbs. Riders already (subconciously) push harder on steeper parts and less on flatter parts. Making this the basis of your pacing plan is somewhat convulted in my oppinon.

I rate Hautacam from this year high although I have "only" given it a 99. It was the hottest day of the Tour (36°C). I give +2 for that. Everone was suffering and 80% of the Top 10 pushed less Watts than on Mont Ventoux a few days later (which is 20 minutes longer). But I prefer to adjust less than reality. I also adjust less for altitude than some studies suggest. Sometimes on the first mountain stage everone just underperforms and we think the winner has done the best performance in history when in reality they were just the only guys at their normal level (La Plagne 1995, Hautacam 2000, Pierre Saint Martin 2015) . I do not have a scientific explanation for that, so I cannot adjust for it.

As for pacing, climbs like Hautacam 2025 and Alpe d'Huez 2001 (~7.4 W/kg for the first 10 minutes) had extreme splits. Many of the Pantani climbs also have very positive splits. I have thought about adjusting for that, but it is a real hassle to calculate splits for every climb and also very difficult to find the right adjustment.

In my opinion, not every performance should and can be adjusted until they are all the same. I just aim to find the general peak and consistency level of riders. I will also publish the TOP 50 climbers list the next few days where I use 10 performances each. That will give you some more insights. E.g. since 2024 the performance gap between Evenepoel and Pogacar has been very consistent in my data. When they were both in good shape and went all out, the gap was always around 0.4 W/kg (or 16 points in my Index)

I don't think you need to adjust for those factors such as pacing plan and overall average level. I just think they're interesting to note and associate other performances with.

 

[Peyresourde]

As for the 24' Tour, I will just spoil some performances so you can compare:

Stage 11 Puy Mary + Pertus
Pogacar 94(+11)+80(+11)
Vingegaard 90(+11)+90(+9) (his second best day in the tour and where he beat Pogacar.

The perfomances on Puy Mary are easily underestimated because for a short climb you are basically at altitude here and you are also climbing for 30 minutes already.

Stage 14 Pogacar 98(+3) Vingegaard 92(+3)
Stage 15 Pogacar 111(+2) Vingegaard 105(+2)

Stage 19 Pogacar 105(0) Vingegaard 86(0)
Stage 20 Pogacar 94(-2) Vingegaard 94(-2)

Some notes:

1) My adjustment for Pla d'Adet is higher than PdB even though the stage was easier. That is because the approach and the nature of the climb (does it have a clear start and end point? is it regular or ar there small descents? is an important factor that many do not consider (enough)

2) Vingegaard clearly had his best day (ever) on PdB and a bad day on Isola. This can be seen in my numbers but also in the actual time gaps.

3)I have Isola adjusted with 0 as it has one of the easiest approaches in pro cycling (1 hour barely pedalling). But these adjustments are not necessarily 100 %accurate. If I adjust slightly more for altitude and slightly less for the approach, then this perfomance would basically be equal with PdB

4) On stage 20 Vingegaard was back to his standard level, while Pogacar did not go all out. Amazingly, Vingegaard had one of the most consistent Tours in history in 2023, then still had good consistency in 2024 after his crash only to have awful consistency in 2025 (basically 3-5 bad days depending on how you count)

 

[Pozzovivo]

So it is also not unlogical that on some days conditions are met where many perform, while on some days many perform badly.

Agree! Which is why I think charting w/kg like (LR and W2W do) and then speaking about "performance" can be so misleading. As in "the level at Paris-Nice was lower than the level at Tirreno" or "this Vuelta was a higher level than that Giro" - which always seem to be the case for some reason. Maybe you understand why @Peyresourde?

What we know is that they had different racing conditions and aprt from a few individuals the other best guys are very closely matched.