Data Analysis discussion

**aeronaut** · 06-19-2021, 08:42 AM

Is this worth it's own thread? Would we reach critical mass? Lot's of good info and support in other forums, especially Rennlist.
But, I have questions.

**olemiss540** · 06-19-2021, 12:29 PM

Sadly Rennlist gets much more traffic. Happy to see this develop if the BMWCCA racers get more active around here.

Personally, I have barely broken my AIM dash open enough to get the lap times to show. Car still getting the kinks worked out to be able to get it steady enough to make use of any data. Look forward to if this gets some action or even somehow a database of PB laps get data housed somewhere for comparison purposes.

**matthugie** · 06-19-2021, 09:25 PM

I started using an AIM Solo2 last year and while I'm just scratching the surface of what it can do it has proved valuable so far. I think I need to find a better way to mount it though, I use the suction cup mount on the windshield and have it wedged against the dash but still get some vibration noise in the G readings.

**mslevin** · 06-22-2021, 02:49 PM

I am definitely new to data analysis, but have been doing some reading as I've just installed an AiM MXL2 in my new car. I found this book to be very helpful, and generally just a very interesting read:

https://www.amazon.com/Making-Sense-.../dp/0983259313

**aeronaut** · 06-22-2021, 07:03 PM

I've browsed that book, there are some nuggets in there.

It'd be cool to have a good repository of lap data to look at and compare! We'd need a good following for that to reach critical mass.

**matthugie** · 06-23-2021, 12:04 AM

Originally Posted by aeronaut

I've browsed that book, there are some nuggets in there.

It'd be cool to have a good repository of lap data to look at and compare! We'd need a good following for that to reach critical mass.

One thing that could be a hiccup is data file formatting. I've exchanged data with some competitors before and if they aren't using the same system it can be a bit of a challenge.

**aeronaut** · 06-23-2021, 08:20 AM

There are a few analysis tools that can read a lot of the different data formats, including AiM. But yea, hoping to have multiple data sets of one track, by several drivers, using the same equipment, and somewhat similar cars..whew! We can dream.

**BlackHawkRacing** · 06-23-2021, 03:37 PM

Originally Posted by aeronaut

I've browsed that book, there are some nuggets in there.

It'd be cool to have a good repository of lap data to look at and compare! We'd need a good following for that to reach critical mass.

A friend and I did this when M3Fourm was still a thing but we lost a lot of it when that died. We ran into two major problem, there is a ton of shit data people think is great and racers (with good data) will NOT share it outside of close friends.

I'm pretty deep into the analysis stuff compared to my friends. I think the best resource for people just getting into it are the videos on AIM's YouTube page. The webinars they have been putting on since COVID are invaluable and included a lot of great extras like math channels and user profiles. James Colborn also has a number of great videos geared towards beginners.

**ScotcH** · 06-23-2021, 04:14 PM

Literally, the best way to use data to your immediate benefit is to do some hot laps, then have someone else who is faster than you do a few laps. Compare. Comparing between different cars is next to useless unfortunately, as lines and braking points etc. will be totally different.

There is of course a lot you can learn from just looking at your own data and doing analysis (ie, how much time you spend off throttle? Braking force? How early are you on throttle at exist? how does your line look on the track ... using all the surface? etc). If you can compare with another driver in your car, you'll have a gold mine.

**aeronaut** · 06-23-2021, 05:30 PM

Good stuff.
I'd be happy if this thread were just a resource to ask a few questions about tools, or products. Or to find out BMW specific stuff, like how to splice into a TPS sensor, etc. For example, I'm happy with my SmartyCam, but I was actually surprised to find out what it actually it can, and can't, do after buying it.

Driving analysis by sharing my data, or comparing to others would be a worthy high bar. And/or, if someone wants some inside scoop on a particular track, we might share that, with data.

**BlackHawkRacing** · 06-23-2021, 06:38 PM

Originally Posted by ScotcH

... There is of course a lot you can learn from just looking at your own data and doing analysis (ie, how much time you spend off throttle? Braking force? How early are you on throttle at exist? how does your line look on the track ... using all the surface? etc). If you can compare with another driver in your car, you'll have a gold mine.

I'll second this, there are a lot of great math channels that can be helpful here. Things like GSum help show if you are maximizing available grip and then you can get into channels like 'Coasting' or "On Throttle"/"On Brake" etc. These are all things that can be done with just a basic Solo. Obviously, it's more accurate if you have sensors but far from required.

**ScotcH** · 06-23-2021, 11:54 PM

Originally Posted by BlackHawkRacing

I'll second this, there are a lot of great math channels that can be helpful here. Things like GSum help show if you are maximizing available grip and then you can get into channels like 'Coasting' or "On Throttle"/"On Brake" etc. These are all things that can be done with just a basic Solo. Obviously, it's more accurate if you have sensors but far from required.

Ya, I was blown away at how much time is gained around a track from just comparing the on throttle and on brake times between drivers. Since we've had the fortune to have many drivers in our car (we offered arrive and drive rental for a few years), we have a lot of good data to show what works and why some are faster than others with apparently the exact same lines and braking points.

Find the really good laps, and figure out WHY they were good. You'll probably find you carried a couple kph more through a corner, or got on throttle a bit sooner, which translates to +5 kph at the end of the straight.

**golgo13** · 06-24-2021, 06:53 PM

Buddies in similar cars all share our AiM data with each other. We also do zoom calls to discuss what we're doing and study each sector while screen sharing to discuss ways to improve.

It's clunky with just GPS speed and the accelerometer, but you can still get a good enough sense of where you are on the gas/brake based on speeds and G loading.

**aeronaut** · 06-24-2021, 07:04 PM

That's pretty good. The most me and my buddies do is share a few vids, a minor amount of data (Vmins, etc), and then a bunch of "you're cheating" texts. lol.

- - - Updated - - -

Originally Posted by ScotcH

Ya, I was blown away at how much time is gained around a track from just comparing the on throttle and on brake times between drivers. Since we've had the fortune to have many drivers in our car (we offered arrive and drive rental for a few years), we have a lot of good data to show what works and why some are faster than others with apparently the exact same lines and braking points.

Find the really good laps, and figure out WHY they were good. You'll probably find you carried a couple kph more through a corner, or got on throttle a bit sooner, which translates to +5 kph at the end of the straight.

I finally added both a brake pressure and TPS channel. Enlightening, for sure.

**aeronaut** · 07-11-2021, 06:47 AM

AiM RS3-beta has been updated. It's a pretty major update and is getting pretty "usable". Still a few annoying bugs, but much better than previous versions.
(eg, Splits is showing my 'best rolling" as slower than my "best".)
FYI, the web page link is incorrect, and downloads the previous version, but you can update using the tool, and it's correct.

**ScotcH** · 07-11-2021, 10:39 PM

Jeebus, it's about damn time ... RS2 analysis has been around for what like 10 years now?!? I miss Motec I2 Pro sometimes

**circuit.heart** · 07-13-2021, 03:15 AM

Originally Posted by ScotcH

Find the really good laps, and figure out WHY they were good. You'll probably find you carried a couple kph more through a corner, or got on throttle a bit sooner, which translates to +5 kph at the end of the straight.

This, but it's just a mindset, you don't need any datalogging till you're at the last few percent. Constantly timing yourself in your head going in and out of a corner and experimenting with inputs will yield results for $0; also a little W2W or lead-follow action does wonders as you now have a ghost car to chase and a LOT of motivation to fill out his/her mirrors.

**aeronaut** · 07-13-2021, 07:43 AM

Originally Posted by circuit.heart

This, but it's just a mindset, you don't need any datalogging till you're at the last few percent. Constantly timing yourself in your head going in and out of a corner and experimenting with inputs will yield results for $0; also a little W2W or lead-follow action does wonders as you now have a ghost car to chase and a LOT of motivation to fill out his/her mirrors.

My limited experience....
I agree that W2W or L/F with another driver/car of equal limits yields fantastic results. I'm also a believer in data analysis for the intermediate and higher driver. And if HPDE ever goes instructor-less, it could be useful for the Novice driver. I also agree that only the driver can determine if a $1k-$2k investment is worth the results, but the results are available. A solo2 and a smartycam, with throttle and brake pressures (not a cheap investment) can be humbling for even advanced drivers and racers.

**BlackHawkRacing** · 07-13-2021, 08:05 AM

Originally Posted by circuit.heart

This, but it's just a mindset, you don't need any datalogging till you're at the last few percent.

I disagree here. A lot of lower level drivers get stuck because they don't know how far below the actual limit they are. Data does an excellent job of showing them this and also helps build a plan on how they can best improve. It is a worthy investment for anyone intermediate and above who wants to actually improve.

**circuit.heart** · 07-14-2021, 09:05 PM

Originally Posted by BlackHawkRacing

I disagree here. A lot of lower level drivers get stuck because they don't know how far below the actual limit they are. Data does an excellent job of showing them this and also helps build a plan on how they can best improve.

Novices usually don't even know what to look and feel for, which is why I'm not a fan of recommending a data rig so early. Telling them that they can brake 2s later for turn 1 isn't going to change squat if they don't buckle up and actually try it. A lot of building speed from the novice ---> intermediate stages IMO is getting comfortable with the track and the car to the point that you feel safe to experiment for more speed. That means getting enough raw seat time to understand Input A begets Output B from the car given Conditions C, under many, many different variations of C.

Barring that, sim racing >>> data collection for novices. Because there's no fear factor in sim racing, you can freely send it and realize "oh, wow I can make that" or you can find videos/data from other drivers with the same "car". Apply it back to the fearless sim, get used to the timing, and all of a sudden when you get back to the real thing it seems a lot less daunting.

Originally Posted by BlackHawkRacing

It is a worthy investment for anyone intermediate and above who wants to actually improve.

I agree here. Once you know what your target is, data viz is a great how to help you get there.

**aeronaut** · 09-22-2021, 05:21 PM

AiM Race Studio 3 Beta update release today.
It's getting pretty mature and usable.

https://www.aim-sportline.com/docs/r.../software.html

**golgo13** · 09-29-2021, 02:25 PM

My biggest complaint is that AiM makes the software with a really shitty UX that you almost have to take their course just to learn how to do basic shit with it.

Honestly, if it wasn't so widely adopted early I think it would have crashed and burned. When you're the only show in town, you kinda get to iterate on your shitty user experience whenever you want lol

**aeronaut** · 09-29-2021, 05:49 PM

Originally Posted by golgo13

My biggest complaint is that AiM makes the software with a really shitty UX that you almost have to take their course just to learn how to do basic shit with it.

Honestly, if it wasn't so widely adopted early I think it would have crashed and burned. When you're the only show in town, you kinda get to iterate on your shitty user experience whenever you want lol

Agreed. The latest (Race Studio and Analysis 3, Beta) is a lot better in some regards, and no better in others.

There's a lot of, why in the world would you do it that way? But there's also a lot of improvements over Race Analysis 2.

**BlackHawkRacing** · 09-30-2021, 07:38 AM

I think people overblow the interface. It's a specialist piece of software, not something that needs to work for everyone from your 15 year old cousin to your 85 year old grandmother. A lot of the other data software I've used is very similar, motec for example. That said, I do think there is a place in the market for a more user friendly, and slimed down version aimed (no pun intended) at your more casual weekend track person.

**aeronaut** · 09-30-2021, 07:56 AM

I agree with your second point. Their software is very powerful (take math channels, for example), but all those power knobs are available and even sometimes needed by the user that just wants to look at their data for 10 minutes after every event. And, the inconsistencies in their interface are many and confusing. Scatter charts; oh look, a list of channels to click on for the Y axis, cool. X-axis..how do I...where is...oh...in here?..yea....context menu, settings, set-x-axis, search for name, ahhh....there it is.

For their beta, here's a biggie I want, get rid of the f'ing vertical white space:

layout-RSA3-view-option.gif