While collecting Scarlet and Violet Data I was informed by NW that per-box data was very consistent, as in every box hat hit counts within an unusually small range that appeared much more consistent than expected randomness. On investigation not only was this obvious and easily outside of margin of error, but it applied to Lost Origin too (if to a lesser extent). Any per box data I find I'll note down here. Note that this is raw data, my code includes a correction to set it to the correct pack pull rate (as it will be more accurate due to higher sample size)

Lost Origin (n=415 boxes)

RR(52.9%), Alt Art(15.2%) or Gold(31.9%)

0 in a box - 30.6%

1 in a box - 45.3%

2 in a box - 19.518%

3 in a box - 3.8554%

otherwise 4 in a box

Full Arts per box

0 in a box - 9.879%

1 in a box - 59.03%

2 in a box - 30.6%

otherwise 3 in a box

Vstar(86.3%)/Vmax(13.7%) per box

0 in a box - 3.125%

1 in a box - 38.31%

2 in a box - 52.53%

3 in a box - 5.3%

otherwise 4 in a box

V per box

3 in a box - 7.95%

4 in a box - 66.988%

5 in a box - 22.89%

otherwise 6 in a box

TG gold per box

0 in a box - 65.3%

otherwise 1 in a box

TG V(43.36%), Vmax, Supporter(56.64%) per box

0 in a box - 6.265%

1 in a box - 75.66%

2 in a box - 17.35%

otherwise 3 in a box

TG holo per box

2 in a box - 12.289%

3 in a box - 73.01%

4 in a box - 13.25%

otherwise 5 in a box

Radiant per box

1 in a box - 24.33%

2 in a box - 71.8%

otherwise 3 in a box

Scarlet and Violet (n=107 boxes)

UR

2 in a box - 61.68%

3 in a box - 35.514%

otherwise 4 in a box

DR

3 in a box - 0.934%

4 in a box - 17.757%

5 in a box - 72.8%

otherwise 6 in a box

HR and SIR combined (36.85% chance for a HR, otherwise SIR)

1 in a box - 15.88%

2 in a box - 83.17%

otherwise 3 in a box

IR

2 in a box - 27.1%

3 in a box - 71.96%

otherwise 4 in a box