Methodological approach to improvement

[zjn]

I watched couple of youtube tutorial about how to train for improvement. I also read here and there in the forum.

Trying to summarize at very high level I'd say it's a typical process feedback loop. Execute -> Check -> Identify Actions -> Restart.

Based on this I developed this spreadsheet to write down things instead to rely on my (poor) memory.

The plan would be to pick the same track, pick the same car, and run 10 lap session.
Review 3 laps: the worst lap (escluding the messed up ones), an average lap, the best lap

The spreadsheet is articulated into different areas:
the top one, where to put information using the conventions identified in the bottom area
the second top one, where to put information about the choosen laps
the middle area, where to put the references used.

Looking at the different data I should be able to spot differences with more precision that just remembering impressions by memory. I'm also a lazy guy, so I prefer to have a standard format instead of writing non structured sentences

I'd like to have your feedback about this approach and on the spreadsheet itself.

Thanks

 

[Tommy_D]

Most of the good drivers who I see on YT load up ghosts that are slightly faster than them and they race the ghost trying to find little bits of time here and there. I guess it all depends on how you absorb data. If a spreadsheet helps you understand the things you are doing right and the things you are doing wrong, then it is a useful tool. But if it's just a means of tracking yourself and you aren't seeing much improvement, sometimes going up against a better driver will show you faster lines around the track and help you improve.

 

[SrGS]

I think if we had telemetry data, it would make this process much easier and precise. Unfortunately we don't have that, so a spreadsheet could be useful in a way. Personally, I would use the sector times that the game provides after we finish a time trial, since it's the most precise information available. Take the best sectors times, and work from there. A purely time based analysis.

I don't know if you will be watching the replay and taking the information from it. If so, there is no precise indication on the track from where one turn starts or ends. So every lap you do, for every turn, you will be taking data from a slighly different section of the track. Other than that, please keep us updated on your results, I'm curious to see it!

 

[i386]

Is there tire wear and fuel consumption involved?

 

[zjn]

Most of the good drivers who I see on YT load up ghosts that are slightly faster than them and they race the ghost trying to find little bits of time here and there. I guess it all depends on how you absorb data. If a spreadsheet helps you understand the things you are doing right and the things you are doing wrong, then it is a useful tool. But if it's just a means of tracking yourself and you aren't seeing much improvement, sometimes going up against a better driver will show you faster lines around the track and help you improve.

Thanks for your replay.
I also do it, when replays are available (es: TT), but also I'd like to develop my own skill about understanding what I did better by retroengineering my laps.
Sometimes in the heat of the lap I cannot figure out what I did better.

 

[zjn]

I think if we had telemetry data, it would make this process much easier and precise. Unfortunately we don't have that, so a spreadsheet could be useful in a way. Personally, I would use the sector times that the game provides after we finish a time trial, since it's the most precise information available. Take the best sectors times, and work from there. A purely time based analysis.

I don't know if you will be watching the replay and taking the information from it. If so, there is no precise indication on the track from where one turn starts or ends. So every lap you do, for every turn, you will be taking data from a slighly different section of the track. Other than that, please keep us updated on your results, I'm curious to see it!

Yes, telemetry would be very nice!

I defined for this exercise these 2 values:
Brake speed: the highest speed I can read approaching a turn
Apex speed: the lowest speed I can read driving through the turn ( I do understand that this may not be the speed at the apex )

I didn't define an exit speed because it's too difficult to judge. I think the best reference for this is the brake speed on the next turn

When I lap alone I know which reference points I have. I consider them a constant (in therory at least). So looking at the replay of my lap I can evaluate when I nail them.
By now my first goal is reproduce laps at the most constant pace. My second goal is improve times.
This may be achieved by 1) understanding/feel I can do better in a specific turn 2) lapping faster then expected and figuring out what went "by mistake" better 3) compare what I do with what others do

 

[zjn]

Is there tire wear and fuel consumption involved?

Not by now, also the setup should be always the same.
To be onest with you I feel that, the level I'm now, 10 laps with fresh tyres and a definite amount of fuel, the most variability on the lap times is me, more the tyre wearing / weight decrease

 

[dabz343]

By now my first goal is reproduce laps at the most constant pace. My second goal is improve times.
This may be achieved by 1) understanding/feel I can do better in a specific turn 2) lapping faster then expected and figuring out what went "by mistake" better 3) compare what I do with what others do

I'm impressed with your spreadsheet, but here are some things to consider so you can be accelerate your learning pace.

Consider first identifying what goal you have before approaching a session. Is it to improve consistency or pace? If it's consistency, then I would suggest an open-ended session where the goal is to string 3 laps together, all within 0.2 sec. These could come laps 2-4 or 13-15, doesn't matter. When you find that consistent rhythm, stop, review the replay and take notes on what you are doing right and wrong to produce consistency. Deconstruct all the phases of the corner (initial braking, trailing, mid/apex, initial throttle, exit) along with steering and racing line. Once you have identified how YOU produce 3 clean and consistent laps, then note this in your spreadsheet as a baseline. Caution, do not do endless laps. If you struggle to string 3 laps together within 15 laps, attempting 30 laps isn't going to necessarily help. After you get good at 3 laps, move onto 5, 7, 10, 20.

If you are looking for pace, then my suggestion is to do a session where you vary your inputs and tactics to find what works for you. A 3 part session of 12 laps. First 4 laps are all about establishing your current pace. Consistency matters, but what is important here is to explicitly recognize your inputs/line/aggression. Laps 5-8 are about pushing the boundaries a little. Where to push takes knowledge on how insightful you are about laps 1-4. For example, if you recognize you are a little abrupt in trailing off the brakes during corner entry, then focus on holding the brakes slightly longer. But as you trail deeper into the corner, your steering mid corner and line may have to adjust, not to mention your initial braking point too.

After laps 5-8, pause and think about what is working, not working, difficult to notice, difficult to execute. Once you have a firm understanding, then apply new tactics that go even further in laps 9-12. This is the challenging part and you will likely spin. This is fine, you are not looking for consistency here. You want to poke around to see how you (and car) behave. Explore with an incremental mindset (not out of control), taking slight variations that you can notice -- noticing (not luck or happy accidents) is a bit part of gaining pace. Once you have found more pace by identifying a new technique/strategy, make a detailed note on what you found, how you found it and what you see as the next step. Then run your next session to find consistency with you newfound pace.

Lastly on finding pace. I recommend starting out just focusing on one thing and save yourself the frustration of trying to juggle multiple variables at once. Braking is usually where most races find their initial chunk of time so start here. Once corner entry feels comfortable, run a session with tire wear so you can step up to another level where you have to start varying your braking. This will challenge you to really know how to find the sensitivity in braking and provide you with the last few tenths of time. Same goes for all the techniques involved in all phases of the corner.

 

[zjn]

Consider first identifying ...

thanks for your very detailed comments. I'm reading through it couple of times to capture the implications of your writings against my current approach.

I'm currently looking for consistency. When I started this trip, couple of months ago, I was really inconsistent I really not able to figure out differences between laps.

Within this consistency search I sometimes identify spots for pace improvement and apply them.

You're helping me to kind of disconnet the two moments, and have separate mindsets for both of them.

 

[dabz343]

Yes, in practice mode my experience has been that it is better to focus either on consistency or finding pace, not both at the same time. Think of going for consistency as trying to be a repetitive machine -- this will build your muscle memory the fastest. However, when going for pace, you want to be in a creative and curious mode, because you are searching for something new. Knowing how to vary your inputs will raise your overall racecraft -- physically and mentally.

But, you have to do this smartly. If you just throw the car around hoping for something to work, then only frustration will rise. Take on a challenge that you can handle. I recommend starting with braking and vision.

Once you do a few rounds of practice, you will get more comfortable and may want to adjust the format. I encourage this, because we don't all learn the same way...as long as you are being intentional about how you practicing, good results will come.

 

[zjn]

Thanks dabz343. What do you mean when you say starting with vision?

 

[dabz343]

Good vision allows you to have more time to make more precise inputs. Key is to find a rhythm throughout the entire corner. It's gonna feel a bit strange at first to look 100 - 200 meters ahead, especially as you try to clip the apex, but once your eyeballs build muscle memory it will be obvious how seeing more improves everything.

 

[zjn]

Good vision allows you to have more time to make more precise inputs. Key is to find a rhythm throughout the entire corner. It's gonna feel a bit strange at first to look 100 - 200 meters ahead, especially as you try to clip the apex, but once your eyeballs build muscle memory it will be obvious how seeing more improves everything.

Oh, yes, got it

That was my biggest sin. I'm better now, but still long way to go, expecially in tracks newer to me

 

[SURV1V3]

Any updates on your progress with this method?

 

[Magog]

My best advice is to look further up the track than seems natural. For me it's natural to look at the entrance of the next corner but I am faster if I force myself to look to the exit of the next corner instead.

 

[zjn]

Any updates on your progress with this method?

is obviously a matter of preference, however, assuming that learning inclinations are similar to mine I can list a few positive and negative aspects of the method.

First of all: the way I am made I can watch replays for hours, but until I write down some very practical notes (e.g., he brakes at 100mt, I at 120, or he goes through the turn a certain way, I another) things don't stick in my head unless they are either macroscopic or very few in quantity.
If your type of learning propensity is different, this method or some derivative of it is not for you.

Negative aspects:
time-consuming, in the sense that writing down all that information and keeping it "alive" for it to be of practical use takes several minutes. If you have little time to drive (as I sadly found I do) then it is not worth spending it on paper
it does not prioritize: the turns from this approach all have the same weight, in reality some turns are more important than others
you may lose focus toward game developments: time trials are biweekly, daily races weekly. If you have many hours to devote to driving then you can apply the method to these challenges, otherwise either do it offline on your own, or apply it "in a hurry" to time events

Positive aspects:
The very fact that it does not prioritize and is time-consuming has led me to sharpen my thinking and figure out where an surrogate version of this method is worth tackling. I already see this as a big improvement for me (maybe instead it is a triviality for you because you are starting from a higher level than me). For example, in the current TT, realizing that I lose a lot particularly on the two low-speed hairpins makes me apply the method to these two turns only, while on the other turns I am closer and content to follow the ghost to improve. The last corner leading onto the main straight, although it is the most important corner on the track, I already do it well enough on my own, so there is no point in me spending time to gain 2 tenths at the end of the straight when I lose 6 tenths in another corner.
In short, working out this method and using it a few times has improved me in terms of understanding where I can improve, and marking things down has helped me focus on how I need to improve.

One of the things I realized is that I don't have enough CPU to work on more than 2-3 aspects at the same time. I am talking about very practical aspects such as making a certain trajectory in a certain corner or taking apex in a certain place.
Once these are in place, I move on to the next 2-3 aspects by prioritizing.
The way I am made I get more results this way than working on 6+ aspects all at once, because in this case I realize I can't take care of them all. Maybe on one round I focus on 1-3-4-6, on the other on 1-2-3-4 and this inconstancy doesn't settle into the muscle murmur what I want to do.

Sorry for the long-winded answer

 

[SURV1V3]

is obviously a matter of preference, however, assuming that learning inclinations are similar to mine I can list a few positive and negative aspects of the method.

First of all: the way I am made I can watch replays for hours, but until I write down some very practical notes (e.g., he brakes at 100mt, I at 120, or he goes through the turn a certain way, I another) things don't stick in my head unless they are either macroscopic or very few in quantity.
If your type of learning propensity is different, this method or some derivative of it is not for you.

Negative aspects:
time-consuming, in the sense that writing down all that information and keeping it "alive" for it to be of practical use takes several minutes. If you have little time to drive (as I sadly found I do) then it is not worth spending it on paper
it does not prioritize: the turns from this approach all have the same weight, in reality some turns are more important than others
you may lose focus toward game developments: time trials are biweekly, daily races weekly. If you have many hours to devote to driving then you can apply the method to these challenges, otherwise either do it offline on your own, or apply it "in a hurry" to time events

Positive aspects:
The very fact that it does not prioritize and is time-consuming has led me to sharpen my thinking and figure out where an surrogate version of this method is worth tackling. I already see this as a big improvement for me (maybe instead it is a triviality for you because you are starting from a higher level than me). For example, in the current TT, realizing that I lose a lot particularly on the two low-speed hairpins makes me apply the method to these two turns only, while on the other turns I am closer and content to follow the ghost to improve. The last corner leading onto the main straight, although it is the most important corner on the track, I already do it well enough on my own, so there is no point in me spending time to gain 2 tenths at the end of the straight when I lose 6 tenths in another corner.
In short, working out this method and using it a few times has improved me in terms of understanding where I can improve, and marking things down has helped me focus on how I need to improve.

One of the things I realized is that I don't have enough CPU to work on more than 2-3 aspects at the same time. I am talking about very practical aspects such as making a certain trajectory in a certain corner or taking apex in a certain place.
Once these are in place, I move on to the next 2-3 aspects by prioritizing.
The way I am made I get more results this way than working on 6+ aspects all at once, because in this case I realize I can't take care of them all. Maybe on one round I focus on 1-3-4-6, on the other on 1-2-3-4 and this inconstancy doesn't settle into the muscle murmur what I want to do.

Sorry for the long-winded answer

No worries, good answer. Thanks^