Rating the BTP NET 2021 Team Performance Projections

How the PCS team standings can help us predict future team performance

The new BTP podcast is live. I talk about Wout van Aert’s potential UCI suspension, EF’s strange moves, and appear on PEZ Cycling’s new YouTube series to talk about BTP and my approach to covering professional cycling.

Rating the 2021 BTP NET Ratings

Over the past off-season, I created the BTP NET team rating via the simple process of taking each team’s total Pro Cycling Stats (PCS) points total from the 2020 season and added/subtracted the number of aggregate points they gained/lost in the transfer market, which shows up how the teams would rank if every rider currently on the roster was on the team in the year prior. While clearly not a perfect system, it was an attempt to bring some sort of objectivity to judge teams’ off-season activities and give us a baseline from which we could estimate how they might fare in the coming season.

2021 BTP NET Projections & Explainer
Tier 4: 19-15
Tier 3: 14-10
Tier 2: 9-5
Tier 1: 4-1

To give us a benchmark against which to measure the BTP NET ratings, here are the final 2021 Pro Cycling Stats points standings for all WorldTour teams.

2021 Team PCS Points Standings

1) Deceuninck - Quick Step-9580
2) INEOS Grenadiers-8614
3) Team Jumbo-Visma-8141
4) Bahrain - Victorious-7301
5) UAE-Team Emirates-7227
6) BORA - hansgrohe-5751
7)Movistar Team-4936
8) AG2R Citroën Team-4816
9) Trek - Segafredo-4686
10) Astana - Premier Tech-4665
11 Israel Start-Up Nation-4595
12) Groupama - FDJ-4448
13) Cofidis, Solutions Crédits-4218
14) Team BikeExchange-3917
15) Intermarché - Wanty - Gobert Matériaux-3865
16) EF Education - Nippo-3793
17) Lotto Soudal-3656
18) Team DSM-3579
19) Team Qhubeka NextHash-2858

And here are my PCS points standings projections from last off-season:

2021 Team BTP NET Rankings

1)INEOS Grenadiers - 7742
2)Deceuninck - Quick Step - 7483
3))UAE-Team Emirates - 7457
4)Team Jumbo-Visma - 6170
5)BORA - hansgrohe - 4802
6)Astana - Premier Tech - 4480
7)Groupama - FDJ - 4299
8)Movistar Team - 4053
9)Bahrain - Victorious - 4008
10)Team BikeExchange - 3944
11)AG2R Citroën Team - 3822
12)Team DSM (aka new Sunweb) - 3459
13)Trek - Segafredo - 3483
14)Israel Start-Up Nation - 3281
15)Cofidis, Solutions Crédits - 3247
16)Lotto Soudal - 2918
17)EF Education - Nippo - 2811
18)Intermarché Wanty Gobert - 2248
19)Team Qhubeka ASSOS - 2035

My initial takeaway from comparing the NET projections to the PCS ranking is that outside of Qhubeka NextHash, no team performed exactly as projected relative to the rest of the WorldTour. This isn’t particularly surprising, since a raw tally of the previous year’s PCS points for the incoming riders can’t predict the nuance of a full season of racing. Also, the total number of points was lower almost across the board for every team, since the COVID-shortened 2020 season included fewer races, and thus, meant riders generated fewer points.

Below, we can see each teams’ PCS ranking next to their NET projection, and the delta between their expected and real ranking.

2021 NET Rating vs 2021 PCS Ranking

Performed Exactly As Predicted (1):
Qhubeka-ASSOS

Outperformed Model (10):
Bahrain-Victorious +5
AG2R Cirtroën +3
Trek-Segafredo +3
Israel Start-Up Nation +3
Intermarché +3
Cofidis +2
Deceuninck - Quick Step +1
Jumbo-Visma +1
EF Racing +1
Movistar +1

Underperformed Model (8):
DSM -5
Groupama -5
Astana -4
BikeExchange -4
UAE -2
Ineos -1
Bora -1
Lotto Soudal -1

This delta between expected and real PCS finish shows us how the team performed relative to the quality of its roster. For example, Ineos underperformed their projection despite finishing 2nd overall in the PCS rankings and third in total wins, while Cofidis overperformed while finishing 13th in the PCS rankings and 15th in wins. This is a rough reflection of the quality of each squad (based on past PCS point accumulation) versus their performance on the road compared to the year prior.

On its own, this might be of interest to some superfans, but I believe where the NET projections really drive value is when we compare them to the raw win, podium, and top ten totals from each WorldTour team.

Team Rankings by Wins, Podiums & Top Tens

We see that DQS once again finished first in wins, podiums, and top-tens, but what is most interesting is that via the pre-season BTP NET projections, we know that according to their rosters, at least measured by individual talents, DQS is not as formidable as Ineos, yet their performance as a whole blows them out of the water.

This shows us that once again, DQS’s ‘secret sauce’ isn’t bringing in major talents, but their ability to generate on-road success from almost every rider they bring into the program. Meanwhile, Ineos is on the opposite end of the spectrum, bringing in endless top talents and big names, but struggling to convert that into the same on-road success.

A way to mine even deeper and measure the ‘quality of its team’s points is to measure this ‘quality to results’ ratio by looking at each team’s podium to top-10 conversion rate, or how often they finish inside the top three relative to how often they finish inside the top ten.

These numbers show us there is a massive gulf between Deceuninck - Quick Step (75%) and Jumbo-Visma (74%) and the rest of the field, and why Ineos (56%), finished below DQS in the PCS points standings despite being projected to finish top of the table. It also shows us that while AG2R and Cofidis might have out-performed their projections, it was done so with somewhat ‘empty calorie’ 4th-10th place finishes.

However, the real utility of combining these two data points is that while looking at their total wins, or podium to top ten rate, one would assume AG2R, Intermarché, Qhubeka-ASSOS and Cofidis all simply underperformed along with DSM and BikeExchange. But by comparing the BTP NET projections to PCS final standings, we can see that these two sets of teams actually had wildly different seasons. DSM and BikeExchange underperformed relative to the talent on their rosters, while the rest simply didn’t have the roster makeup to convert top tens to podium finishes and at least relative to the talent on their rosters, had successful seasons.

Major Takeaways

1) Deceuninck - QuickStep proved once again that in terms of racking up points (and in turns, wins) they are inevitable even though they might not have the ‘strongest’ roster on paper.

• The model didn’t project them to have the firepower to land the top spot, but they still ran away with the top PCS stop, which shows that the team’s hyper-focus on winning stages and one-day races, while eschewing grand tour overall placings is paying massive dividends.

• Their 75% podium/top-10 conversion rate means that 3 out of every 4 times they placed a rider in top ten, they were either winning the race or on the podium. This rate is almost inconceivably high.

2) When the data is laid out, Bahrain’s great, and DSM’s dismal, seasons are really put into context.

• Despite being separated by only three places in the projected standings, they finished a whopping 14-places apart in the final PCS point standings.

• This is a great example of how two teams with similar talent can have wildly divergent seasons and underlines the importance of quality rider management and building a positive professional environment.

3) We can also see that a few teams had somewhat secretly good seasons.

• Despite failing to shine in the raw number of wins, Israel Start-up (9th in wins), AG2R Cirtroën (13th in wins), Trek-Segafredo (8th in wins), Intermarché (16th in wins), and Cofidis (15th in wins) all healthy outperformed their BTP projection.

• Granted, this is my own made-up projection model, but it does show some sort of success and overachievement inside the teams. It essentially shows that their woes aren’t related to on-road performance as much as their roster building.

4) On the flip side, teams like Ineos (3rd in wins), UAE (4th in wins), and Groupama (7th in wins) all underperformed their BTP NET projection while having ‘good’ seasons according to the raw win numbers.

• Groupama’s relatively high standing in win total but poor finish relative to their BTP projection exposes the team’s underperformance at major races (which yield a high amount of PCS points) and strategy of racking up wins at smaller, regional French races.

• While UAE won the Tour de France, which automatically means you had a great season, and Ineos won the Giro, a solid consolation prize for a Tour-focused team, the data shows that outside of their top-performing stars (Tadej Pogačar and Egan Bernal), the rest of team slightly underperformed.

• For example, Pogačar alone accounted for 40% of UAE’s total wins, and Ineos only won three one-day races while running up the winning score by peaking for one-week stage races that lacked serious competition.

• This isn’t a clear sign of problems per se, but it is a good indicator of what is going on with the team as a whole. It is also worth mentioning that UAE took steps to remedy any potential performance dips by bolstering their roster with some blockbuster off-season signings, headlined by João Almeida, while Ineos has mainly stayed the course.

5) At the opposite end of the spectrum, Jumbo-Visma suffered a major blow when Primož Roglič crashed out of the Tour de France, but unlike UAE, their extreme depth still allowed them to outperform their projected finish.

• Having a generational talent like Wout van Aert as a ‘backup plan’ doesn’t hurt, but even setting Van Aert aside, the team so full of talent that they were still able to finish ahead of Ineos’ A option Richard Carapaz with their C (or frankly maybe even lower) option Jonas Vingegaard at the Tour de France.

• And most worrying for rival teams is that these numbers don’t even factor in Tom Dumoulin, the runner-up in the time trial at the recent Tokyo Olympics, who essentially sat out the trade team racing season.

• If Roglič can avoid crashing, Van Aert can convert some of his near-misses in major one-day races, and Dumoulin makes a full return, the team could easily improve on this performance in 2022.

• Their 74% podium/Top-10 ratio allowed them to pick up a significant amount of points every time they got a result and is a testament to the amount of talent on the roster.

Once the team 2022 team rosters are set in January, premium subscribers will get updated BTP NET projections for each team for the coming season, and potentially even add regression/progression factors such as riders’ age and injury history to increase the accuracy of the exercise.