Here's yet another episode of SCIENCE and myth-busting.
The first series of test is touching on the so called "speed cap" I've seen being mentioned here and on some other places. Basically, it's said that once the speed of a suit hits past 24.9, any further increment to the speed stat will not make your MS go faster.
Except it does not really make sense. What's the point of Netmarble giving all those S ranks and etc raw speed that goes way past 24.9? Just so that they have some buffer speed to counter slow effects? Kind of stupid if you ask me.
Some on Bahamut says that there is no hard cap on speed, but there's diminishing returns once you hit past somewhere around the 24.0 mark. Make sense.
And so, here's me doing all the testing. It's done in TW, so if you think that it's not accurate enough since not done on KR or whatever, feel free to do that. Since I am playing on TW, this test is good enough for me.
OC6 Ball K. Why Ball K? It hits the all important 24.9 speed with 4 custom investments into the speed stat.
Gundam Fight Arena
Edge to edge distance: 62 squares
Traversable distance: 61 - 1 + 0.5 = 61.5 squares (roughly)
1. Start from edge of map
2. Boost straight till no boost left
3. Calculate distance travelled
And here's the test.
And 2 charts.
There's no hard cap on speed when it comes the raw stats on the graph. However, improvements are much less significant once you hit past speed of somewhere around 23~24.
Therefore, for Ball K, if you want to dump speed, just do C3 speed, and use that extra custom for something else. Like agi or atk. Unless you really need that 0.3 square boost.
Here's a neat little video on agi. It just proves that the benchmarks for agi in olgame is not really accurate. But then again, olgame did say that it's just something for reference.
So yeah. Better to just invest agi bit by bit and see which works for you the best. There's not too much difference between 16.5 and 18.3 in the video (for that MS anyway), so that's something worth noting.