---
Awareness plays a large roll in track now. The difference between 0 awareness at 60+ is very large. Even players with 70-80% Track can be useful with high awareness %
//__Kaspar__: The difference is not in **speed** but only **movecost** while tracking(from Elestir and 'help awareness'). Lets test how much awareness affects movecost while tracking!)
//
Heigo provided me with enough data to make a reliable prediction how per and int affect the track skill. We collected samples from 36 builds with different int and per and checked the track percentages.
The earlier track formula was obviously wrong when you went to the extremes. I fed the data into mathcad and the results are below. The percentages can be off by 1% here and there but the overall table has proven to be correct. Note that this is not an exact formula but a very close approximation.
PER
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
INT ---------------------------------------------------------------------------------------------------
1 18 27 34 40 46 52 56 61 64 68 71 74 77 79 82 84 86 89 92 95 98 101 105 109 114
2 24 32 38 44 50 55 59 63 67 70 73 76 78 81 83 86 88 91 94 97 100 104 108 113 118
3 30 36 43 48 53 58 62 66 69 72 75 77 80 82 85 87 90 93 96 99 103 107 112 117 123
4 35 41 47 52 56 61 64 68 71 74 76 79 81 84 86 89 92 95 98 102 106 110 115 121 127
5 39 45 50 55 59 63 67 70 73 75 78 81 83 86 88 91 94 97 101 105 109 114 119 126 132
6 43 49 54 58 62 66 69 72 75 77 80 82 85 87 90 93 96 100 103 108 113 118 124 130 138
7 47 52 57 61 65 68 71 74 76 79 81 84 86 89 92 95 98 102 106 111 116 122 128 136 143
8 51 56 60 64 67 70 73 75 78 80 83 85 88 91 94 97 101 105 110 115 120 127 133 141 150
9 55 59 63 66 69 72 75 77 80 82 85 87 90 93 96 100 104 108 113 118 125 131 139 147 156
10 58 62 65 68 71 74 76 79 81 84 86 89 92 95 98 102 107 111 117 123 129 137 145 154 163
11 61 64 67 70 73 76 78 80 83 85 88 91 94 97 101 105 110 115 121 127 134 142 151 160 171
12 63 67 70 72 75 77 80 82 85 87 90 93 96 100 104 108 113 119 125 132 140 148 158 168 179
13 66 69 72 74 77 79 81 84 86 89 92 95 98 102 107 111 117 123 130 137 145 155 165 176 188
14 68 71 74 76 78 81 83 85 88 91 94 97 101 105 110 115 121 127 135 143 152 161 172 184 197
15 71 73 76 78 80 82 85 87 90 93 96 100 104 108 113 119 125 132 140 149 158 169 180 193 206
16 73 75 77 80 82 84 86 89 92 95 98 102 107 112 117 123 130 137 146 155 165 177 189 202 217
17 75 77 79 81 83 86 88 91 94 97 101 105 110 115 121 128 135 143 152 162 173 185 198 212 228
18 77 79 81 83 85 87 90 93 96 100 104 108 113 119 125 132 140 149 159 169 181 194 208 223 239
19 78 80 82 85 87 89 92 95 98 102 107 112 117 123 130 138 146 156 166 177 190 203 218 234 251
20 80 82 84 86 89 91 94 97 101 105 110 115 121 128 135 143 152 162 173 186 199 213 229 246 264
21 82 84 86 88 90 93 96 100 104 108 113 119 125 132 140 149 159 170 181 194 208 224 240 258 277
22 84 85 88 90 92 95 99 102 107 112 117 123 130 138 146 156 166 178 190 204 219 235 252 271 291
23 85 87 89 92 95 98 101 105 110 115 121 128 135 143 152 163 174 186 199 214 229 246 265 285 306
24 87 89 91 94 97 100 104 108 113 119 125 133 140 149 159 170 182 195 209 224 241 259 278 299 322
25 89 91 93 96 99 103 107 112 117 123 130 138 146 156 166 178 190 204 219 235 253 272 292 314 338
NOTE: Can be wrong by a point here and there but should mostly hold water.
Linear model Poly33:
track = f(per,int) = p00 + p10*per + p01*int + p20*per^2 + p11*per*int + p02*int^2 + p30*per^3 + p21*per^2*int
+ p12*per*int^2 + p03*int^3
Coefficients (with 95% confidence bounds):
p00 = 2.803 (-12.67, 18.28)
p10 = 10.01 (7.536, 12.49)
p01 = 7.021 (4.817, 9.224)
p20 = -0.4692 (-0.6464, -0.2921)
p11 = -0.6966 (-0.8237, -0.5694)
p02 = -0.2277 (-0.3582, -0.09712)
p30 = 0.009648 (0.005319, 0.01398)
p21 = 0.02234 (0.0198, 0.02488)
p12 = 0.01469 (0.01087, 0.01852)
p03 = 0.003257 (0.0005643, 0.005949)
Goodness of fit:
SSE: 4.876
R-square: 0.9995
Adjusted R-square: 0.9994
RMSE: 0.3292