Michigan International Speedway

Brooklyn, IL


Type Speedway
Configuration D-Shaped Oval
Length 2 Miles
Characteristics Turns Banking 13 Width 73'
Frontstretch Banking 12
Backstretch Banking 5
Track Record Ryan Newman 194.232 mph (June 18, 2005)
Capacity Grandstand--136,373

MIS

Click on image for a large view

Fast Tracks

After Talledega and Daytona, Atlanta is generally considered the 3rd fastest Cup track with a challenge from Texas. Don't forget Michigan though--it is still officially faster than Texas--by .008 seconds. They are both 3 mph slower than Atlanta. Lowe's has also been done at 190+. I need to do a study on which teams consistantly perform well at these four venues and how it reflects on or translates to their performance at all of the other 1 1/2 to 2-mile speedways.

NR2K3 Stuff

The Records Problem

I've been working hard for you. Really I have. Running laps, tuning cars, checking facts... That's where the problem start. When I first obtained NR2K3 Dale Earnhardt Jr. held the Michigan track record for qualifying at 191.149 mph. After much work I finally drove it at 186.196 so my relative top speed is at 97.4%. Then Ryan Newman comes along and does a 194.232 mph lap. I'm back at 95.8%. So I have to really work the setups, playing with all the little things to wring every spare tenth of a mph out of the track that I can get with my limited skills. I hit 188.487 and I'm back at 97%. Then something clicks with the expert setup: I run 191.699 and my average speeds are better than my previous record lap. So I actually top the old Earnhardt record but the only way to catch Newman is to shave a full half-second off my time. Not gonna happen methinks.

So here's the problem: too many variables. The tracks change, the cars change, even the drivers' abilities change. An AI that was originally set at 100% should now probably be at 102%. I should move into a 100% race according to the numbers. But I'm just overmatched there--I'm still overmatched at 98%. I like having at least a chance to pick up a win. I saved the replay where I fell to 23rd and started picking cars off with 21 laps to go, passing Tony Stewart on the last lap for the win. That was at 93% though and I can't race there in good conscience anymore. I think I may have taken all the fun out of it by over-analyzing.

I should be straight up with you. I have two goals when I visit a track. One is to see how fast I can go and how well I can do against the track record. Two is to find a level that I can "compete" against the Cup regulars. The two don't always exist easily together; a setup that works great for testing can be completely unsuitable on race day. Sometimes the problem is the air flow when in close proximity to other cars. Other times it's the AI itself. At certain tracks I don't like the way it behaves. To this point I have avoided tweaking the .ini files but there will come a day...

Radt says:

Since we run the Cup schedule we are here several times a year. LK has yet to hit on the perfect setup here. Usually it will work fine until I go into the frontstretch oval and then it will buggy-whip me if I'm off my mark just a tad. Ya can't always run the same line in a race as ya do in practice. Does he not see that? Hello? There are other cars who want the same spot. Y'know, every day is a new day for this guy. I'm trying to get him to race online so he can see how the real drivers do it. Those AI folks he's always talking 'bout? They're dumb as stumps and ya can quote me.

See Also:

MIS Website

My Michigan AI & Qualifying Spreadsheet Analysis