The rate of withdrawals really went up in the last week of the Tour.
Two Kaplan Meier diagrams, one this year's, one comparing 2022 to 2021 and 2020
You can see that by the end, 2022's withdrawals are almost the same as 2021's. The main difference is in the Did Not Start-type withdrawals, which were mostly due to the COVID protocols.
The percentage of Did Not Starts really was high this year - 69% vs. 30-33% in the previous two years, which again suggests COVID hitting hard.
When the withdrawals happened also suggests that is what's happening - normally when a team lose a rider they start losing other riders too because of the strains of maintaining performance with fewer riders but this time it's just been a steady drip, drip, drip of withdrawals.
That can also be seen when you look at the withdrawals by stage, either by order of stage or by how many riders dropped out on each stage.
Historically one or two stages have most of the withdrawals but this time the withdrawals really were quite evenly spread, with 1 stage losing 6 riders, then 1 stage with 5, 2 stages with 4, 3 stages with three, 4 stages with 2 and 2 stages where 1 rider dropped out.
Types of withdrawals by which week they happened in:
Did not starts increased every week.
The mid-stage abandonments had almost the opposite pattern.
There were only 2 Over the Time Limit withdrawals, 1 in week 3 and 1 in week 4.
So, what can I summarise - 2021 no longer looks like the outlier in terms of withdrawals, 2020 does, but that may be because of circumstance and COVID.
As we have a women's Tour de France this year, I will also look at the rate in that event, and I will also crunch the numbers for the Giro Donne to see which rate is more common.