Quote Originally Posted by seb_sh View Post
I was bored last evening and played around in a spreadsheet a bit and calculated the difference per kilometer compared to the stage winner for each stage separately. This removes stage length from the equation but to me it shows the "raw pace" on each stage. I also made an average of these differences and a ranking based on that. The average is less relevant to the "real time" classification because the pace on all stages has equal importance to the average regardless of length but I think it can show some things about the raw speed potential.

I removed the last stage from the average per rally because that was much affected by weather and I removed data that was affected by mechanical problems but not the driver mistakes. So here are some interesting things:

1. Evans had the best pace since SS9 on average. In real-time he was 0.2s slower than Ogier on all of the stages on Saturday but that's because Ogier had a smaller advantage over longer stages while Evans had a bigger advantage on shorter stages.
2. On SS12 the stage before he crashed Neuville had his slowest pace from all of the stages so far, by some margin. Maybe he started to backed off, relaxed and lost focus? In the post stage interview he said "I have the snow tyres this afternoon. This morning we were on slicks. It was the safe choice. We have a good lead, but we have to be clever." Maybe i'm over analyzing :P
3. Latvala was on average the 4th fastest, slightly faster than Sordo. He had 6 stages where he was under 0.5s/km slower than the winner and 4 more where he was under 1s/km slower. However, Meeke and Breen were not far behind, I'm pretty sure Meeke could have outpaced him and Breen was just 0.260s/km on average slower in a 2016 car.

I dunno, maybe it's a silly thing to look at but the results fit some of my intuition/feeling.
How can you use statistics to take into account the context behind the stage times though? Ie that everyone who was on Rally2 would be under instruction to finish the rest of the rally at all costs, so naturally their pace is going to be off when they Rally2 (see Juho and Kris as good examples)... And what about the difference in tyre choice vs the correct tyre choice? I don't think you can use data unless you add in the variables that impact the validity of the data...