F1 2017 resource points3/10/2023 Though it is still hard to believe that the ratio between probability of failure and QC upgrade cost already levels out at 2. This would bring total for 100% upgraded car to 76.5K-81.5K resource points or 6.5-6.8 seasons. When counting the durability upgrades as well, you need an additional 11.5K resource points, or 1 more season. This seems resonable.įor the most cost effective development path, you will need around 65K-70K resource points or 5.5-5.8 seasons (assuming 600 resource points per race weekend (this might be a bit high assumption)) in order to fully develop the performance departments engine, aero and chassis. Also in the lower right plot we see, through the rightwards shift that the importance of quality upgrades increases as the failure probability increases. And as one might expect, a lower value results in lower overall R&D costs since the probability that each upgrade will fail is less. The effect of the parameter p_fail can be seen in the two lower plots. This provides the best balance between spedning resource points on the QC upgrades and avoiding having to respend on upgrades due to failed upgrades. Here it can be seen that the optimal solution for each department (although slightly different) are 2 QC upgrades (1 also being a good option). Look at the two right plots in the figure. Keeping the above point in mind, we move on to look at the optimal number of quality control (QC) upgrades. It gives a slow start to development, but as can be seen in the two left plots of the figure, gives great financial benefit in the long run. Running the simulation many times made it clear that, regardless of number of quality control upgrades applied, it's always beneficial to buy all 5 cost reduction upgrades at the start of development. Aero being the most expensive (it has most upgrades and most expensive cost reduction upgrades), engine the cheapest department and chassis somewhere in between. There are significant differences in R&D costs depending on department regardless of cost reduction and quality control upgrades. Some key features to notice about the results: The left y-axis in each plot shows amount of resource points required while the right y-axis show the number of seasons required assuming 600 resource points earned per race weekend. (The top two plots use probability p_fail= 50% that an upgrade will fail on first attempt). The two lower plots show the sum of all three departments and for different choices of the slightly uncertain parameter p_fail. The top two plots visualizes the R&D costs for each perfromance department (not counting the durability upgrades for engine and gearbox) as a function of cost reduction upgrades and quality control upgrades, respectively. Feel free to ask any questions if something is unclear! Here comes a discussion describing the results. The program simulates the statistics in the R&D process and code together with more results can be found on GitHub. For the ones who prefer the short version, scroll to the bottom of this comment for the conclusion. My aim of the simulation was to find the optimal strategy on how to spend resource points in order to get my Mclaren-Honda back to the top as fast as possible. Hey guys! These are results of a numerical simulation I wrote for the R&D process in F1 2017.
0 Comments
Leave a Reply.AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |