- 1,233
- VS & 74 (France)
- GT-Alex74
Update, not with cars, but with some Excel file tweaks.
First, the ease rating system. Giving such precise rates was not ideal and the exact figure for a specific car could vary according to my mood or various non-objective things. First, I wanted to rate the ease, but couldn't give a bad rate for fast cars, and then I think I shouldn't rate this car 10 because it is clearly not as easy as this one, but I rated this one 9 already and it's not as good... but it's more fun and blablabla... And for the moment, most cars are above 8,5 and I feel I have less margin to differentiate good cars.
So screw rates, there's the comment section anyway ! I swapped for a 1 to 5 stars system. In my mind, the progression through stars is not linear, as 3 stars don't represent a 5/10 rate, but more the average rate from the pack :
* : really difficult to handle. Doesn't mean it won't be fun, far from it, some of the most enjoyable cars in my opinion will fit through there, but you'll probably have to take a shower after trying to keep that facing the right direction.
** : the handling sports more cons than pros, that's generally the kind of car that you can keep on the road but doesn't want to go fast around corners.
*** : that's average, you won't find that particularly good nor particularly bad. Or it can be something potentially very fast, but difficult enough to handle so that beginners are more likely to stay at average cornering speeds.
**** : now we have something actually nice for racing, that allows you to keep a good rythm without worrying too much. It may be more difficult on its limits or have still a bit of over/under steer though.
***** : top class handling, goes very fast around corners and very easily, it will take a lot of abuse without betraying you. If the stars are golden, it's icing on the cake.
Now, second point about the Excel changes : I have developped a PP recalculation / cheated car indicator function. Got some thinking into that, some math, and some trial and errors, but I think I've put the finger onto something pretty nice now. I'm still experimenting though, so for the moment, I threw that in a third sheet : I'll try to integrate that to the full table when I'll be sure to be fully satisfied with it.
Here's how it works : I set up a time interval (I included that as a variable so I can change it easily, but +/- 1 second seems to work nice), and for each car, the function will look for all the entries that fits in that range around the currently studied car's time. Then, the average time and average PP are calculated and compared to the studied car's values : specifically, I calculate what percentage of the studied car's time the average represents, and take that as a multiplying factor to recalculate the new PPs. I added a slot to showcase the difference between new and old PPs, which can also be defined as a cheat indicator.
I only spotted one flaw with this method : the extreme values will be compared nearly exclusively to slower or faster cars. For example, the ZZ-II being at the moment at the very top of the board, there's no entries in the [0 / -1] seconds range. Though, it doesn't seem so bad at the moment, and I'll be able to correct that problem if necessary by tuning a few cars (power and weight only) or doing a few race cars and putting their results outside the board (the functions look for the whole columns regardless of tables). The bottom of the board can appear to be a bit more difficult to deal with, as time gaps are wider, and I can't really find slower things than the VW military vehicles. But I think it's much less problematic if the accuracy is a bit off for those cars, and as their times rely nearly only on top speed, having them not being compared to something else may not be that bad.
Also, keep in mind that the adjusted figures will update themselves each time a new entry will fit the range around a car, so they can be inaccurate until the testing is completed.
So, I'm updating first post accordingly.
First, the ease rating system. Giving such precise rates was not ideal and the exact figure for a specific car could vary according to my mood or various non-objective things. First, I wanted to rate the ease, but couldn't give a bad rate for fast cars, and then I think I shouldn't rate this car 10 because it is clearly not as easy as this one, but I rated this one 9 already and it's not as good... but it's more fun and blablabla... And for the moment, most cars are above 8,5 and I feel I have less margin to differentiate good cars.
So screw rates, there's the comment section anyway ! I swapped for a 1 to 5 stars system. In my mind, the progression through stars is not linear, as 3 stars don't represent a 5/10 rate, but more the average rate from the pack :
* : really difficult to handle. Doesn't mean it won't be fun, far from it, some of the most enjoyable cars in my opinion will fit through there, but you'll probably have to take a shower after trying to keep that facing the right direction.
** : the handling sports more cons than pros, that's generally the kind of car that you can keep on the road but doesn't want to go fast around corners.
*** : that's average, you won't find that particularly good nor particularly bad. Or it can be something potentially very fast, but difficult enough to handle so that beginners are more likely to stay at average cornering speeds.
**** : now we have something actually nice for racing, that allows you to keep a good rythm without worrying too much. It may be more difficult on its limits or have still a bit of over/under steer though.
***** : top class handling, goes very fast around corners and very easily, it will take a lot of abuse without betraying you. If the stars are golden, it's icing on the cake.
Now, second point about the Excel changes : I have developped a PP recalculation / cheated car indicator function. Got some thinking into that, some math, and some trial and errors, but I think I've put the finger onto something pretty nice now. I'm still experimenting though, so for the moment, I threw that in a third sheet : I'll try to integrate that to the full table when I'll be sure to be fully satisfied with it.
Here's how it works : I set up a time interval (I included that as a variable so I can change it easily, but +/- 1 second seems to work nice), and for each car, the function will look for all the entries that fits in that range around the currently studied car's time. Then, the average time and average PP are calculated and compared to the studied car's values : specifically, I calculate what percentage of the studied car's time the average represents, and take that as a multiplying factor to recalculate the new PPs. I added a slot to showcase the difference between new and old PPs, which can also be defined as a cheat indicator.
I only spotted one flaw with this method : the extreme values will be compared nearly exclusively to slower or faster cars. For example, the ZZ-II being at the moment at the very top of the board, there's no entries in the [0 / -1] seconds range. Though, it doesn't seem so bad at the moment, and I'll be able to correct that problem if necessary by tuning a few cars (power and weight only) or doing a few race cars and putting their results outside the board (the functions look for the whole columns regardless of tables). The bottom of the board can appear to be a bit more difficult to deal with, as time gaps are wider, and I can't really find slower things than the VW military vehicles. But I think it's much less problematic if the accuracy is a bit off for those cars, and as their times rely nearly only on top speed, having them not being compared to something else may not be that bad.
Also, keep in mind that the adjusted figures will update themselves each time a new entry will fit the range around a car, so they can be inaccurate until the testing is completed.
So, I'm updating first post accordingly.
But I still play online as well, and I intend to use that work to improve our online events 
For the gap with previous car, you don't have to use $ for the row, as you want it to update for both parameters, so it ends up to be just =C2-C1. You can also write $C1 to fix the column, but it's not necessary here.
But a friend asked me to try his next real life car, so Type-R's coming just after^^
machine.
It is the proper combination between Lotus's handling, V8's prowless and Lightweight's, well, lightness.
; fluid handling coupled with light weight. 👍 The 350 is all about that and a bit more; it's like a racier Honda NSX-R.
) for a while. Good car, but we bought it from the wrong guy. Anyways, most Peugeots are fun to drive, their hot hatches have an acute feel to them, and the RCZ is pretty stable although it is underpowered against the car people often mention (the Audi TT). 
