How to MEASURE PERFORMANCE in Quantitative terms to IMPROVE MEMORY PALACE SKILL

General Memory Chat
1

In all the memory competitions (e.g. counting cards etc), we can measure our performance against time and figure out how well we are developing our memory palace skill and keep track of our improvement.

BUT how to track our skill when we are studying for subjects like history, geography etc ?

One way to do it would be test scores but how do we pin point if we are IMPROVING our ‘Memory Palace Skills’ or not?

Or should we use some measure like time taken to read a page of a ‘standard’ book/text etc?

What is the most optimal way to track our performance in order to improve our skill i.e. ‘organize data into mind palace in the fastest way possible’?

2 Likes
3

Hi Josar,

I would say that there is no “objective” method to quantitatively measure performance other than your own speed getting faster and more accurate in the fabrication of Mind Palaces or memorizing what have you. You can always use ANKI or another Smartphone App to build up speed and accuracy and perhaps apps can measure your percentile there anyway.

Start a Memory Journal my man

Stefos

1 Like
4

I track every step of progress towards memorizing a full chess opening repertoire and having memory palace coverage against every possible move that is good.

image
image1799×799 29.2 KB

1 Like
5

Hi Zweb,

I hope you’re having fun…To me Chess openings are irrelevant…It’s the subsequent moves that count.

Stefos

2 Likes
6

True to an extent but it is a lot easier to find the best moves in positions you are very familiar with.

One other utility is if your opponent is the one to vary from opening prep the game becomes more valuable to analyze. Even if they win you will learn the punishment for next time wheras if you forget your prep and play a bad move often the insight to be gained from analyzing that game is “play your prep next time”.

Openings only really give a significant advantage when players have equal tactical vision.

That said 95% of top ranked players have vast amounts of opening theory memorized so it begins to matter a lot above 2200 rating.

7

Do you use alphabet for indexing your memory palaces? If so, do you find it useful? (for example I saw you wrote A - Airport, B - Baseball, etc)

1 Like
8

Yes I put them all in loose networks. I plan to do a separate network for each “act” of my black openings Nimzo, Anti-Nimzo, Sicilian Sveshnikov, Anti-sicilian, Other which each will have ~10 chapters, 10 palaces.

I am almost done with Nimzo and doing Sveshnikov next so I will select ~10 palaces to place in a loose network for that next and it will probably be A-M with some skipped.

I also have one A-Z network I use for competitive memory and plan to use for chess calculation in the future. This is the main utility of the named network both to have a strict schedule to cycle through and to chain them together for long-form disciplines. I just copied the named network thing for chess because it was already habit but it isn’t as important in the chess project.

1 Like
9

Hi,

Understood…It seems very tedious unless you’re a competitive player for actual ranking

Stefos

1 Like
10

Yeah and most competitive players don’t really do much until a very high level. Its true its not valuable to study openings until you are losing to people who get winning positions out of openings.

If I was obsessed with only chess strength I should be spending this time studying openings practicing calculation and tactics but playing a strategic game is more fun to me and requires good openings and also study of the associated best strategies.

Its often though that I get completely winning positions with good openings and strategy and miss tactics and lose so you do make a good point.

11

I admit I know nothing of chess, but I’m interested in how people organise their memory palaces. I’m sure what you are doing can prove useful at a competitive level.

1 Like
12

Hi Josar,

Just a quick suggestion…
Why don’t you use the standard way of tracking skill or performance improvement?
There are several (very very very good) models of skill improvement, none of which will give you any trouble to understand if you already know how to read basic mathematical symbols or you are prepared to learn them. There’s no need to actually know (dah dah DAH!) CALCULUS in depth just what it means… But to be honest the only way to track, model any change (in anything) is to use the mathematics of change - calculus.

The standard basic models are the learning and performance curves and how they can be quantified from data (the times or quality of performance of the effort put in).

They are equations, differential equations.
And the ESSENTIAL thing that you are talking about above is THE RATE OF CHANGE - NOT THE ACTUAL CHANGE…

Give you a couple of examples (of the top of my head).

Rate of change of (say) memorising 100 digits in terms of time taken over attempts over say twenty eight days.
That can be MODELLED EXACTLY using the following:
K (Min possible time - time taken)
K is a constant.
It’s a curve that can never ever reach the minimum possible and looks roughly like half a parabola or stretched out quarter circle… This is not a bad model to use and track your progress. And will roughly give you a model to track improvement from a point where you are pretty comfortable with what you are trying to improve which has a definate ceiling. This type of curve is actually best used to track a PLANNED improvement from one time to a better target time, it’s used for example for swimmers trying to take four seconds off of a personal best over the course of three months … Or for a formula one motor racing team trying to get an extra one or two seconds off of a three minute lap… Or a student trying to improve his test averages.

That’s dy/dt = k(M-P) where P is current performance.

Better is what is called a sigmoid curve (so called because it looks like an S turned on its side). You can track actual progress and set up future target performance using this FABULOUS model.

The rate of change in this case is modelled by the following or similar: dP/dt = (that says the rate of change of performance with respect to time is equal to: k(P)*(1-P/M) some constant times your present level (at any time) multiplied by the percentage your performance is of the maximum you can actually expect or are targeting.

This is EXTREMELY accurate once you have some data.

Key points off this model are you improve very quickly at first and keep getting better at a faster and faster rate until you reach about half of your target improvement, then assuming the same effort and commitment you’ll find your improvement slows gradually until its non existent (because of you reach P=M, then (1-P/P)=zero.

These two improvement models (and similar ones) model, record and allow you to see at a glance any improvement (or backward step) the rate of change of performance. At a glance.

It’s not easy to give a metaphor to make this improvement model understandable to people who aren’t fluent in maths - but one is. You are on a journey which will take you twenty eight days, your start speed is best near zero and your end speed is most definitely zero, and the fastest speed you ever reached is at half way (fastest rate of improvement) and the shape of the curve of the speed is a parabola or (semi circle stretched out look) and the distance is an S curve.

I’m not sure anyone who isn’t interested in improvement would have read this far, but if they have - learning to model your changing scores in any discipline - any - needs the maths of change and that’s (only) calculus. There’s no other way. So learn it.

Hole that helps.

K

That

2 Likes