ABCDEFGHIJKLM
1
TextšŸ‡ŖšŸ‡ŗ PALšŸ‡ÆšŸ‡µ JPšŸ‡ÆšŸ‡µ ā€“Ā šŸ‡®šŸ‡¹
2
segtext# boxes1st boxtotal# boxes1st boxtotalāˆ† totalgrand total:+163f+5.44s
3
aestart1242412828+4made by shoutplenty
4
d-bussm2144112222-19Info
5
g7sm1222211414-8This is a fully experimental model, made from TASed text. The method is documented on the source videos. Note: text means set of textboxes (as replaceable with fast text); textbox means a set of up to 3 lines cleared with one a-press.

This has been checked robustly for transcription errors and optimal inputs to clear textboxes. Other errors may be encountered in that optimal text lengths depend on the quarterframe cycle on which the text is started. This model uses lengths as determined by the single sample of each text recorded, with a goal of emulating RTA-ish text triggering.
6
q2start-11212112020-1
7
q2start-23227231764-8
8
q2end3228131461-20
9
q3end1161611616+0
10
q6start724199824203+4
11
q6end22960423100+40
12
q7sm11414114140
13
p1ystart720165825183+18
14
p7sm1181811414-4
15
b7sm11414114140
16
d-susm1212112424+3
17
s1start11212701327362+92
18
she2end419103517110+7
19
she3end2183622145+9Vs Previous Model
20
she4end1323222853+21Compared to the previous model, the estimate for šŸ‡ÆšŸ‡µ text length relative to šŸ‡®šŸ‡¹ has decreased from 7.13 to 5.44 (-1.69s). The previous model was only theoretical and had never been experimentally verified so this large inaccuracy isn't surprising. I don't understand it well enough to assert what the differences are but I tried to guess them based on a comparative analysis (on the vs previous) tab. These are my best guesses:

ā€¢ gaps: +42f
the previous model didn't count the gaps between textboxes within the same text set at all, and šŸ‡ÆšŸ‡µ has an additional 12 of these, each incurring ~3.5f

ā€¢ off-by-1: -90f
after adjusting for gaps, the previous model counts šŸ‡®šŸ‡¹ less accurately, with a linear error appearing to be around 1f per textbox, of which there are 90, which indicates an error in the algorithm

ā€¢ noki 3 pause: -14f
the (second) noki 3 text's last textbox has a long pause in it before it closes in šŸ‡®šŸ‡¹, about 14f longer than the corresponding pause in šŸ‡ÆšŸ‡µ
21
sce4end1151511818+3
22
she5end42193524114+21
23
sce5end1151511717+2
24
s6start22045516113+68
25
s6end1303022643+13
26
she7end2264422235-9
27
s7hsm11414114140
28
n3estart-1625142816193+51
29
n3estart-2527160424102-58
30
n4estart-13317532263-12
31
n4estart-21252511818-7
32
n5start-12265111919-32
33
n5start-23216031764+4
34
n5end3257631460-16
35
n7sm1191911414-5
36
r2istart3166431466+2
37
r7sm11414114140