Philosophical setup instructions for TagTime

[harding]

@zedmango

Not just how many minutes you spend, but also the length of time over which that task is performed. For instance, if you work for a large number of clients but only work on each individual client for a few days or weeks, TagTime is likely to miss some of those.

Ultimately it’s just total time spent doing a particular task. You’re correct that if you work only, say, 5 hours on each project, there will be some projects that never receive even a single ping (about 0.1% in this case[1]), but it doesn’t actually matter whether you do those 5 hours all in one day or spread the work out over an entire year—the probabilities of being pinged (or not) are purely dependent on the amount of time spent on the project.

[1] exp(-5/0.75) = 0.001272

[zedmango]

Oh, very cool! That makes sense since each minute is equally likely to be picked.

So a 15 minute task would get pinged 1 - exp(-0.25/0.75) = 28.3% of the time, and a 45 minute task would get pinged 1 - 1/e = 63.2%. Those are much higher probabilities than I would have guessed!

I made a table:

hours	prob
0.1	0.124826681
0.2	0.234071662
0.3	0.329679954
0.4	0.41335378
0.5	0.486582881
0.6	0.550671036
0.7	0.606759279
0.8	0.655846213
0.9	0.698805788
1	0.736402862
1.1	0.769306818
1.2	0.798103482
1.3	0.823305554
1.4	0.845361735
1.5	0.864664717
1.6	0.881558171
1.7	0.896342871
1.8	0.909282047
1.9	0.920606068
2	0.930516549
2.1	0.939189937
2.2	0.946780656
2.3	0.95342385
2.4	0.959237796
2.5	0.964326007
2.6	0.968779073
2.7	0.972676278
2.8	0.976087007
2.9	0.979071987
3	0.981684361
3.1	0.983970642
3.2	0.985971533
3.3	0.98772266
3.4	0.9892552
3.5	0.990596437
3.6	0.991770253
3.7	0.992797545
3.8	0.993696604
3.9	0.994483436
4	0.99517205

Now I’m curious if there’s a formula for the expected error for what TagTime says.

[dreev]

I think your table is what @harding made a graph of above, and expected error is what I was getting at here:

Philosophical setup instructions for TagTime

Just did some math on the question of how much time you have to measure with TagTime for the estimate to be probably approximately correct: If TagTime says you spent 45 minutes there’s an 86% chance it was really between 0 & 2 hours If TagTime says you spent 3 hours there’s a >90% chance it was really between 1 & 6 hours If TagTime says you spent 6 hours there’s a 71% chance that’s within 37% of the truth If TagTime says you spent 8 hours there’s a 67% chance that’s within 33% of the truth If …

[eugeniobruno]

I have tried to do monte carlo simulations of how wrong in relative terms tagtime can get, depending on days of use and real time per activity.

I have no idea if I simulated this right, though. Could any tagtime/math nerd sanity check this?

this is an activity done about 15 minutes per day: https://i.imgur.com/f4ajdGK.png
this is an activity done about 45 minutes per day: https://i.imgur.com/M5uTDj2.png
this is an activity done 8 hours per day:
https://i.imgur.com/t3wirmW.png