Jump to content

krz9

Members
  • Content Count

    6
  • Joined

  • Last visited

Community Reputation

1 New Car Smell

Gaming Setup

  • Platforms
    Steam
  • Peripherals
    Controller
  1. krz9

    AI Difficulty Calculator

    Thanks for sharing the link! 🙏 I appreciate it. I'll check it out and see if we can add some of this data on our site to improve the results. Awesome, glad to hear it! Please let me know if there's anything we can do to make it more useful for you! (reach out here or email krz@f1laps.com)
  2. krz9

    AI Difficulty Calculator

    Ha, sorry to hear that it gave bad results! As Ollof mentioned, Singapore bight be one of those tracks where player skills are all over the place. But at the same time, I do think that the difficulty calculator data we published still has a lot of room for improvement. Which reddit calculator are you using? Curious to check it out. Maybe I can improve ours accordingly too.
  3. krz9

    AI Difficulty Calculator

    The tricky thing with AI difficulty calculators is that lap times depend on car performance. Which means that, in order to compare lap times, you need to use a Time Trial car (because, as you mentioned, e.g. in My Team season 1 the car is much worse). The difficulty calculator should be based on Mercedes Time Trial data that maps to the corresponding AI difficulty (which in turn is assumed to be uniform across game modes). Not sure if my explanation makes sense 😄... but the takeaway is that - if the difficulty calculator has good enough data - than the resulting difficulty level should work well for My Team. At the same time though: if this calculator is more confusing than helpful, than I recommend you just don't use it. At the end of the day, the best rule of thumb imo is that your quali result should be roughly equal to your teammate. The easiest way to find a good AI difficulty is to just run some qualis and adjust the difficulty based on the result! Awesome - that's great to hear! And yes, data nerdiness is exactly how we got to building this site. Please ping me anytime if you have questions or feedback - would love to improve the site to make it more useful for you. You can also DM me here (not sure if this forum supports DMs?) or email at krz@f1laps.com.
  4. krz9

    AI Difficulty Calculator

    Following up on my last comment. I made good use of the days off and published some of our time trial laps data. Check out this new page for Austria (or the list of all tracks here). You can enter your own laptime and benchmark it against average AI laptimes per difficulty level on that track. A laptime of 1:09.600, for instance, would result in a suggested difficulty level of 38. As a reminder, this is all based on Mercedes Time Trial data -- that's the only way to make data comparable. As I mentioned in my last post, you could alternatively (once you know your average "baseline" difficulty level) adjust your AI difficulty based on the average per-track difficulty adjustments. And/or track your own AI difficulty levels to see your own progress over time.
  5. krz9

    AI Difficulty Calculator

    Hi Lurtz - good question! There are different approaches to AI difficulty. None of which are perfect imo. Two that I've considered: Baseline difficulty + adjustments per track: once you have a sense for your average ("baseline") difficulty, you can then adjust it for each track based on how well you know/think you'll do. As an example, my own difficulty baseline is around 90 (meaning over the course of a season my average is about 90). For tracks I do well, such as Austria, I increase my difficulty by 5. For tracks I suck at, such as Vietnam, I decrease by 7. The page I've linked to is helpful with this approach, as it shows the average difficulty adjustment for each track. Laptime comparison: this approach is similar to what you've asked about. You first set your own laptime on a track, then compare it with some online benchmark which spits out the recommended difficulty. This works well if you're willing to always hop into Time Trial mode to set a few laptimes. It's important that you use Time Trial (and always the Mercedes), because laptimes depend on factors such as car performance, tire wear, engine mode, etc. Time Trial is the only mode that keeps those constant. If you'd set a lap in My Team Practice, for instance, you'd get a much slower laptime (e.g. because engine mode is lower, or because your car is weaker); and the slower laptime would then result in a lower suggested difficulty by any "difficulty calculator". Personally, I don't want to go to Time Trial each time I'm about to race a Grand Prix... hence I stick with approach #1. I'd love to hear if anyone has different approaches to this! With regards to the second approach, my friends and I are working on publishing our average laptime by difficulty data as well -- I'll post it here shortly, stay tuned. Please let me know if there's anything else I can help with - would love to hear more opinions on how to improve our data and make everyone's F1 experience better 🙂
  6. krz9

    AI Difficulty Calculator

    Jumping in here with my first comment. I loved reading this thread as I found that a good AI difficulty is a huge factor in having a good, immersive F1 game experience. In the past, I tracked my difficulty adjustments myself. Since I started discussing this with some friends we just ended up taking a few days to build a basic tool (F1Laps). It allows you to track your myteam/career progress & AI difficulty. The tool also calculates our average difficulty & difficulty adjustments per track, which we've published here: https://www.f1laps.com/pages/resources/f1-2020-game-difficulty-by-track (updates daily). Our adjustments are not super consistent with the calculator data -- I guess it's another case of each one of us having different track strengths & weaknesses, and not enough aggregate data yet to smooth out individual inconsistencies. Personally, my own approach mostly relies on my previous seasons. I prefer setting the difficulty based on how I perform "realistically" - even if it's not consistent with the "average" driver.
×