It doesn't? I'm pretty sure that you said it wasn't very meaningfull.. :S
my laptimes suck coz i did uber long stints so ner ner :P
It was too short race for me to get a clean lap! And also i drove too slow to have faster times!
oo, Sherlock.
All the spdo GT2 team kinda bottled it i think all of us were 0.5s slower than in practice stints.
i think all of us were 0.5s slower than in practice stints.Yeah same here miles off what I was doing in prac in fastest lap and consistency. :/
I said it's not meaningful as a measure of performance. Then you pointed out that if you compare two finishing times you can see who finished in front of who. That's pretty obvious and not something I'm going to disagree with.
Even if you say the sample isn't representative because I've excluded crash laps/pitlaps - it's still the same standard for all teams. I didn't include it for one and exclude for another. I just don't see a point in having an average lap 2 sec slower and not anywhere near race-pace just for the sake of leaving in pitlaps, everyone should have roughly the same pitstops anyway. As for loss in traffic, this is all included (if your fastest lap was a 2:22.00 then 105 % of that is 2:29.10) so only laptimes above that are ignored, most of the traffic laps in gt2 were around 1-2 sec off pace.
As far as damage is concerned, let's say for example in real life your mechanic doesnt change a tire properly and u have to pit again so you lose another 30 sec, as soon as you get out on track you're faster than anyone else.. is it really fair to look at average lap including those 30 seconds? It's not an accurate representation of what the driver did.
I've taken statistics and probability as a part of Mathematics 3B in college, I understand all of your points and arguments - but this data selection shows what I think is the most realistic representation of what a driver does on the track during a race.
As far as damage is concerned, let's say for example in real life your mechanic doesnt change a tire properly and u have to pit again so you lose another 30 sec, as soon as you get out on track you're faster than anyone else.. is it really fair to look at average lap including those 30 seconds? It's not an accurate representation of what the driver did.
I've taken statistics and probability as a part of Mathematics 3B in college, I understand all of your points and arguments - but this data selection shows what I think is the most realistic representation of what a driver does on the track during a race.
By this funny 105% magic the data comes complitely inaccurate when, for example, driver drives lets say 15 laps. First ten laps are full of driver mistakes and laptimes around 2.30. Then the five last laps he decides to do 2.21.5 five times in a row resulting 2.21.5avg because the first 10 laps are excluded by 105%. When in reality his "true avg" for that stint would be somewhere around 2.27.2(without the 105%). Of course this is exaggerated example but just to show that those avg laptimes tell something but definately dont tell the whole truth in some cases.
Oh Petripoodle, he's talking about decent drivers, your example is only relevent if im the case subject, where the average lap time would be the 1 clean lap i manage to do 
Sasaquatch, if you've taken both those courses then surely you'd have known the probability in people agreeing with you would be preeeeetty slim
2 pops if im right, big guy :bullwhip:
I'd say 105% is probably a little too wide, 102% might actually be more on the money. On a 2m15sec lap, to have a 2.7sec (+2%) spread from fueled up to running on fumes is about right, perhaps a tiny bit more for heavy traffic on full fuel. That covers say 2:14.3 through to 2:17.0 which covers 3ID's race with some wiggle room.
It should cover what the driver is capable of achieving over a stint, while dropping the poor laps which occured due to issues. It cant be 100% accurate, but i'd say its a fair representation and at the standard were talking about in MoE, it shouldnt be an issue where lap times are wildly different unless something has caused it, those being freak occurances and not the norm.
Ignoring driver errors obviously means it doesnt make these results the ultimate deciding factor of who's faster/better, but the race results do a fairly good job of factoring in that.
But I wouldnt argue these results arent accurate, simply because you only have to look at the official results and consider how anyone who suffers a disconnection or gets taken out, they have to have that included on their official result, so is the official results an accurate measurement of events? Hardly.
Sasa, stick up one for nfinity GT1, i know they spent 80% of the race holding hands with F1rst (though a lap down after their disco just after the 1hr mark) im guessing the results would look similar. We'll perhaps leave the GT2 car out of this
Sasaquatch, if you've taken both those courses then surely you'd have known the probability in people agreeing with you would be preeeeetty slim
2 pops if im right, big guy :bullwhip:
I'd say 105% is probably a little too wide, 102% might actually be more on the money. On a 2m15sec lap, to have a 2.7sec (+2%) spread from fueled up to running on fumes is about right, perhaps a tiny bit more for heavy traffic on full fuel. That covers say 2:14.3 through to 2:17.0 which covers 3ID's race with some wiggle room.
It should cover what the driver is capable of achieving over a stint, while dropping the poor laps which occured due to issues. It cant be 100% accurate, but i'd say its a fair representation and at the standard were talking about in MoE, it shouldnt be an issue where lap times are wildly different unless something has caused it, those being freak occurances and not the norm.
Ignoring driver errors obviously means it doesnt make these results the ultimate deciding factor of who's faster/better, but the race results do a fairly good job of factoring in that.
But I wouldnt argue these results arent accurate, simply because you only have to look at the official results and consider how anyone who suffers a disconnection or gets taken out, they have to have that included on their official result, so is the official results an accurate measurement of events? Hardly.
Sasa, stick up one for nfinity GT1, i know they spent 80% of the race holding hands with F1rst (though a lap down after their disco just after the 1hr mark) im guessing the results would look similar. We'll perhaps leave the GT2 car out of this
Yes of course Paulie, but what i tried to say with my exaggerated example is that more bad/exluded laps u have at the beginnin, the better avg laptime you get over the whole stint in the end. It works also both ways, if you have more excluded laps in the end then avg is obviously worse as the first part of the stint has more weight in this 105% calculation.
Sasas method is close to what it should be but its not the 100% truth either. It favors the guys doing mistakes in the start when the laps with high fuel dont have any value.
Sasas method is close to what it should be but its not the 100% truth either. It favors the guys doing mistakes in the start when the laps with high fuel dont have any value.
(v1rg0)
DELETED
by v1rg0
(v1rg0)
DELETED
by v1rg0
Give me back my jacket!!!
Where is the problem here?
If you don't like the stats scipy made on his own, then do your own ones rather than bashing him for doing it not your prefered way (whatever that is) in the first place...
Don't trust any statistic you didn't fake yourself
If you don't like the stats scipy made on his own, then do your own ones rather than bashing him for doing it not your prefered way (whatever that is) in the first place...
Don't trust any statistic you didn't fake yourself
some of them really look nice!
Lynce's reflections and colours? Looks pretty nice, although the AA could've been better. (2nd pic)
moe5.jpg would be the start of the race where I back out thinking everyone was going to bin it entering T1.
1
2
12 Hours of Kyoto GP Long: Post-Race Discussion
(42 posts, started )
FGED GREDG RDFGDR GSFDG