I've done some some experiments on Haste and shot rotations using Cheeky's spread sheet as a bases.
For my calculations I assumed the following to be true of hunter shot rotations:
3:2 (steadys only)
1.5:1 (Multi/Arcane/steady MM rotation)
1:1 (steady only)
1:2 (steady shot only)
I picked two gear sets to look at: A "best in slot" haste set with 297 haste rating with 4 piece T6, and a "best in slot" non haste slot with 54 haste rating (Blackened Naaru Sliver is just insanely good, i have it as roughly 118ap averaged out over the internal cool down).
In Cheeky's spread sheet i used the following conditions and settings: Full raid buffs (sadly, I wasn't thinking and just used my typicall raid buffs I get instead of setting the raid buffs to either none or full), Shot Rotation and Quick Shot Rotation were set to be the same per which ever shot rotation I was running varying haste buffs for. Latency in Cheeky's sheet was set to 0.05 seconds (my typical ping).
For each set i used the following haste increasing buffs and all the various combinations of such assuming a BM (20% haste) and a top level quiver (15% haste). To apply the haste buffs I calculated the total haste effect and figured out what the equivalent haste rating minus any haste from gear that would need to be added in the "hand adjustment" section on Cheeky's spread sheet that would yeild the same total haste effect.
Drums of Battle (80 haste rating)
Haste Potion (400 haste rating)
Heroism/Bloodlust (30% haste)
Rapid Fire (45% haste)
My results for the non haste set of gear:

|
Originally Posted by Praxx
Table 1
| buffs | | talents | | quiver | | haste % | | speed of Auto | | 3v2 | | 1.5v1 | | 1v1 | | 1v2 | | 0 | | 1.2 | | 1.15 | | 1.427464968 | | 2.10 | | 2471.12 | | 2488.81 | | 2332.87 | | 1843.5 | | 80.0 | | 1.2 | | 1.15 | | 1.497783439 | | 2.00 | | 2499.61 | | 2502.87 | | 2420.94 | | 1912.84 | | 400.0 | | 1.2 | | 1.15 | | 1.779057325 | | 1.69 | | 2517.78 | | 2534.71 | | 2757.57 | | 2216.47 | | 480.0 | | 1.2 | | 1.15 | | 1.849375796 | | 1.62 | | 2520.62 | | 2540.34 | | 2797.89 | | 2284.98 | | 487.2 | | 1.2 | | 1.15 | | 1.855704459 | | 1.62 | | 2520.86 | | 2540.82 | | 2801.42 | | 2290.78 | | 591.2 | | 1.2 | | 1.15 | | 1.947118471 | | 1.54 | | 2524.23 | | 2547.07 | | 2853.72 | | 2376.88 | | 730.8 | | 1.2 | | 1.15 | | 2.069824204 | | 1.45 | | 2527 | | 2550.47 | | 2872.06 | | 2504.59 | | 846.8 | | 1.2 | | 1.15 | | 2.171785987 | | 1.38 | | 2528.28 | | 2553.02 | | 2873.89 | | 2601.09 | | 1007.2 | | 1.2 | | 1.15 | | 2.312774522 | | 1.30 | | 2529.87 | | 2556.2 | | 2876.16 | | 2747.64 | | 1111.2 | | 1.2 | | 1.15 | | 2.404188535 | | 1.25 | | 2530.8 | | 2558.07 | | 2877.49 | | 2842.09 | | 1310.8 | | 1.2 | | 1.15 | | 2.579633121 | | 1.16 | | 2532.4 | | 2561.27 | | 2879.78 | | 3013.55 | | 1426.8 | | 1.2 | | 1.15 | | 2.681594904 | | 1.12 | | 2533.23 | | 2562.95 | | 2880.97 | | 3105.72 | | 1437.2 | | 1.2 | | 1.15 | | 2.690771465 | | 1.11 | | 2533.31 | | 2563.1 | | 2881.08 | | 3113.18 | | 1588.0 | | 1.2 | | 1.15 | | 2.823321783 | | 1.06 | | 2534.3 | | 2565.1 | | 2882.5 | | 3218.05 | | 2191.2 | | 1.2 | | 1.15 | | 3.353523057 | | 0.89 | | 2537.48 | | 2571.52 | | 2887.06 | | 3597.87 | | 2342.0 | | 1.2 | | 1.15 | | 3.486073376 | | 0.86 | | 2538.12 | | 2572.83 | | 2887.98 | | 3681.17 |
|
due to the large gaps in the Auto Shot times I graphed the Speed of Auto Shots versus the expected dps I obtained from Cheeky's spread sheet.
By using Line fitting it is possible to determine where dropping down a rotation will cause an increase in dps. For this gear set I found the following relation:
rotation vs Auto Shot times
3:2 > 1.9s using a 3:2 rotation or a 1.5:1 rotation is optimal for auto shot times greater then 1.9s
1.9s >= 1:1 > 1.2s dropping down to a 1:1 rotation is optimal for auto shot times of 1.9s or less but greater then 1.2s
1.2s <= 1:2 dropping further down to a 1:2 rotation is optimal for auto shot times of 1.2s or less
Results for the Haste gear set:

|
Originally Posted by Praxx
Table 2
| buffs | | talents | | quiver | | haste % | | speed of Auto | | 3v2 | | 1.5v1 | | 1v1 | | 1v2 | | 0 | | 1.2 | | 1.15 | | 1.641057325 | | 1.83 | | 2235.36 | | 2318.55 | | 2334.1 | | 1842.15 | | 80.0 | | 1.2 | | 1.15 | | 1.711375796 | | 1.75 | | 2239.56 | | 2325.68 | | 2411.5 | | 1902.56 | | 400.0 | | 1.2 | | 1.15 | | 1.992649682 | | 1.51 | | 2250.38 | | 2350.48 | | 2553.92 | | 2161.66 | | 480.0 | | 1.2 | | 1.15 | | 2.062968153 | | 1.45 | | 2251.48 | | 2353.57 | | 2555.18 | | 2227.35 | | 560.1 | | 1.2 | | 1.15 | | 2.133374522 | | 1.41 | | 2252.38 | | 2356.46 | | 2556.37 | | 2287.93 | | 664.1 | | 1.2 | | 1.15 | | 2.224788535 | | 1.35 | | 2253.48 | | 2359.95 | | 2557.8 | | 2369.78 | | 840.2 | | 1.2 | | 1.15 | | 2.379533121 | | 1.26 | | 2255.14 | | 2365.28 | | 2559.97 | | 2501.1 | | 956.2 | | 1.2 | | 1.15 | | 2.481494904 | | 1.21 | | 2256.13 | | 2368.44 | | 2561.25 | | 2591.41 | | 1080.1 | | 1.2 | | 1.15 | | 2.590444586 | | 1.16 | | 2257.09 | | 2371.56 | | 2562.52 | | 2684.97 | | 1184.1 | | 1.2 | | 1.15 | | 2.681858599 | | 1.12 | | 2257.84 | | 2373.99 | | 2563.5 | | 2755.51 | | 1420.2 | | 1.2 | | 1.15 | | 2.889342038 | | 1.04 | | 2259.38 | | 2378.94 | | 2565.5 | | 2894.3 | | 1536.2 | | 1.2 | | 1.15 | | 2.991303822 | | 1.00 | | 2260.05 | | 2381.13 | | 2566.38 | | 2964.79 | | 1652.3 | | 1.2 | | 1.15 | | 3.093393057 | | 0.97 | | 2260.68 | | 2383.19 | | 2567.21 | | 3026.12 | | 1803.1 | | 1.2 | | 1.15 | | 3.225943376 | | 0.93 | | 2261.44 | | 2385.67 | | 2568.21 | | 3100.98 | | 2406.3 | | 1.2 | | 1.15 | | 3.75614465 | | 0.80 | | 2263.96 | | 2393.88 | | 2571.49 | | 3395.98 | | 2557.1 | | 1.2 | | 1.15 | | 3.888694968 | | 0.77 | | 2264.48 | | 2395.6 | | 2572.18 | | 3444.1 |
|
Graphing the results yeilds:
using line fitting and extending the lines some to find the 3:2 to 1:1 switching point yeilds vertually identical results for when to drop down a rotation.
Conculsion: The ideal times to drop down a rotation for a
low ping connection player are 1.9 and 1.2 seconds. Ping can and does play a HUGE factor in which rotation will generate the best dps and can significantly change the "ideal" times to drop a rotation down (Ping and its effect on haste and rotation selection is going to have to wait for another day).
Regards,
Ravenfire
P.S. (I never seem to get tables to paste right....)
EDIT: Thanks Praxx for showing me how to post tables correctly