I am updating the forecast based on the fact that there will definitely be more AVs on the road in the first quarter of 2023 than in the last quarter of 2022, although haven't made a quantitative estimate yet. Here is a graph I am using as a proxy for the growth of all AVs vs time.
from: https://seekingalpha.com/article/4499111-notable-progress-at-cruise-increases-gms-value
No Scores Yet
Relative Brier Score
0
Forecasts
0
Upvotes
Forecasting Activity
Forecasting Calendar
No forecasts in the past 3 months
| Past Week | Past Month | Past Year | This Season | All Time | |
|---|---|---|---|---|---|
| Forecasts | 0 | 0 | 0 | 0 | 2 |
| Comments | 0 | 0 | 0 | 0 | 10 |
| Questions Forecasted | 0 | 0 | 0 | 0 | 1 |
| Upvotes on Comments By This User | 0 | 0 | 0 | 0 | 6 |
| Definitions | |||||
New Prediction
Probability
Answer
10%
(0%)
Less than 20
65%
(-15%)
Between 20 and 29, inclusive
23%
(+13%)
Between 30 and 39, inclusive
2%
(+2%)
Between 40 and 49, inclusive
0%
(0%)
More than or equal to 50
Files
New Prediction
Probability
Answer
10%
Less than 20
80%
Between 20 and 29, inclusive
10%
Between 30 and 39, inclusive
0%
Between 40 and 49, inclusive
0%
More than or equal to 50
Why do you think you're right?
according to https://www.dmv.ca.gov/portal/vehicle-industry-services/autonomous-vehicles/autonomous-vehicle-collision-reports/, the number of collisions looks like it is on a downward trend and may even get below 20. In 2021 in the same quarter, it was 23. So perhaps the car manufacturers figuring out how to fix these issues. What do you think @WilliamKiely ?
Files
Why might you be wrong?
New vehicle releases with new technology could perhaps create an uptick in incidents.
Files
Yes agree, it may be too tight. I wonder if 2022 was an unusual year too, since it is the highest year in all of the recorded record. Thinking about all of the factors that may affect make incendents: number of AVs on the road, introduction of new AV systems, introduction of new traffic conditions, maybe others. I expect that the # of AVs on the road keeps increasing, and software fixes for the other two reasons probably affect all of the cars about the same time, that maybe increases and decreases in incidents may look like step functions. I am going on an assumption that the number of incidents per number of AVs goes down over time as the software becomes better and better. Let me think about this more... Will probably revise soon.
Files
New Badge
We probably want to know what the total number of vehicle collisions (including AVs) in California looks like over time to make an accurate forecast. I created this graph about LA's total vehicle collisions over time from their website https://data.lacity.org/Public-Safety/Traffic-Collision-Data-from-2010-to-Present/d5tf-ez2w. You can clearly see the effect the pandemic had on the number of collisions (assuming the pandemic was the cause). Now if I could get a similar graph for all of CA.... hmm... actually I think San Francisico and bay area collision data would be more applicable to this forecast, if the bay area is the epicenter of AV rollouts ( i don't know how much it is ).