This is a little script that tries to simulate different outcomes for F1 Fantasy and help you pick the best team.

It generates a bunch of possible different qualifying grids as well as the final race position. This is based on a pace map, whereby the more pace difference, the less chance a driver has of finishing ahead of a faster driver. It generates all possible team compositions given the available budget. For each quali grid and race finishing position combo, it assigns the respective points to each team. Across all the top 100 scoring teams in each quali-race combo, it finds out the most frequent teams and the ones that have the best average score, provided they show up at least 5 times

Clone this repo and then run

npm install

This was done in node v14.9.0. Make sure you have a compatible version. Recommend using nvm for node version managing

Go to the root directory and run

node --stack-size=24000 main.js

The 24000 should be high enough to calculate everything, but if you get a stack call error, just increase it.

There are three command line options you can pass:

node --stack-size=20000 main.js --iterations=100 --TD=PER --budget=100

• Iterations is how many quali-race grid combos you want. Default is 100. Keep in mind increasing this can dramatically slow down calculations. This code is not optimized
• Budget is the available budget to build teams. Default is 100
• TD is your turbo driver. Bear in mind they will heavy skew your results, so try to pick a decent choice. I suggest PER. If not specified it will just ignore turbo points.

In the data/data.js file there's a pace object specifying the relative pace for drivers and constructors. The fastest drivers get 0, the slower a driver/car the bigger that value is.

For example:

• HAM is the fastest driver so we give him a 0.0

• BOT is a bit slower, so we give him 0.2 (I think of it as 2 tenths slower, but that's not necessarily the interpretation here. It's just a value to indicate how much slower)

• MER is fastest constructor so they get a 0.0 too

• VER is as fast as HAM so we give him 0.0 as well

• RBR however, is a bit slower, so they get 0.1

That makes means the following:

HAM + MER = 0.0
VER + RBR = 0.1
BOT + MER = 0.2

So there's a higher chance VER will beat BOT in most samples.
Try to adjust the values to however you see fit.

This does not take MD, streaks or fastest lap into considerations. Nor does it generate grid with DSQs or DNFs.
I might include some of those features at some point but this is a good first set so far and I think useful enough.

My next thought is to actually create a UI before tacking those issues