November 6, 2011

Inside GP: Reh at Melbourne

Every racing weekend one driver might explain how he prepared himself for his race. I (Reh) am gonna make the start this week.
For me everything started two weeks ago when the pack and later when the drivers were released. First I drove some laps on every track to learn about the car, the tracks were released a few weeks earlier.
After a busy week with some messed up races in the other compos in the GRIF I even forgot to race the Qualifying. I just trained a little in Melbourne, my personal best was a high 17 seconds I think. Most of the other guys were better, so I wouldn't have got a better starting position probably anyway.

Though, all through the last weeks I figured out how I could discover the best pit stop strategy for the race - it turned out I put much too much work on that...
Anyway: The race time is the sum of all laps and the time you lose in the pit. Sounds logical, doesn't it?

On Sunday after I did all the other races I started to race a few single stints, saved the rehplays and put the times via Majortoms GameAnalyser into an Excel sheet.
I determined the average laps (every first lap of the stint, every second lap, and so on) by ignoring the two worst laps to get less mistake of it.
Also I split the first lap into a imaginary laptime and an absolute term, the time you lose at the start, which is the same every time.
Then I checked all possibilities of strategies (the numbers on the right) and respectively put the mathematical terms together.
The best strategy has the lowest value. :)
Though, these numbers are from runs without AIs, so I tried the 4 stop strategy with high AI and with low AI, for me the faster AI worked much better. I could get them almost as fast as the low AIs at the start but had less traffic though the race:
With slow AI i had 145-146 km/h, 147 with fast ones. So I had my strategy :)
Well, when you convert this into seconds (the numbers in the table), you get 494 seconds at 147 km/h - which is a lot different to 473. So it is much more important to get in the box at the right time in the race than to work hours on a stupid formula...

Earlier for other series' I tried it another way assuming that the lap times become slower in a linear way - so I figured out the increase and calculated the strategy in another way, with bigger error, but a bit faster to work out.
Anyway, all this is kind of unnecessary when you have a good gut feeling - mostly it is obvious which strategy is the best.

I have to say my race was lucky in the end, the strategy worked perfectly, which wasn't self-evident at all.

And thats how i prepared for the race :)