Advertisement
Guest User

SpeedDemon

a guest
Nov 1st, 2014
201
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 1.80 KB | None | 0 0
  1. CLOCK TICKS, TURN COUNTS & SMOOTH TIME: timing in FFT |
  2. '''''''''''''''''''''''''''''''''''''''''''''''''''''''
  3. The way time actually flows in FFT is based on 'ticks'.
  4. Every tick of the clock adds a players speed to his charge
  5. meter. If the number passes 100, you take an 'observed turn'
  6. and reset your meter. If two players fill on the same tick,
  7. they both play. A likely sorting method is highest overshoot
  8. Goes first. Ex:
  9.  
  10. Sp Charge Over
  11. 11 99->110 10
  12. 12 96->108 8
  13.  
  14. If Sp_11 and Sp_12 fill on the same tick, 11 plays before 12.
  15. The problem is 12 will now fill before 11 because they both
  16. start at ZERO. Rule:
  17.  
  18. For two characters, the higher speed will eventually play
  19. twice in a row.
  20.  
  21. In the previous post we used the term '100/Sp' as our 'turn time'.
  22. This looks problematic as FFT doesnt actually use fractions. Lets
  23. Reformulate using turn counts. The math that tells you turns for a
  24. given player based on tick counts is as follows:
  25.  
  26. T = TICKS, total clock ticks
  27. C = CYCLE COUNT = CEILING(100/Sp), this gives the ticks per turn
  28. N = TURN COUNT = FLOOR(T/C), this is played turns given TICKS
  29.  
  30. Now we can calculate something interesting. Assume we have
  31. multiple players. At a large TICKS value, compute two new values:
  32.  
  33. N_total = N_1 + N_2 + N_3 + ... , sum of all turn counts
  34. Sp_total = Sp_1 + Sp_2 + ... , sum of all speeds
  35.  
  36. This gives us 'statistical speed' or 'Ss':
  37.  
  38. Ss_j = N_j/N_total*Sp_total
  39.  
  40. Close with a 3 player example:
  41.  
  42. TICKS = 257
  43. Player 1 2 3
  44. Sp 6 9 13
  45. C 17 12 8
  46. N 15 21 32
  47. Ss 6.2 8.6 13.2
  48.  
  49. This gives credence to the Sp/100 value used for computing power as it
  50. nearly equals a scaled probability derived from turn counts. They are
  51. Identical for 1 player. They likely converge at infinite player count.
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement