The following are both wrong
1) cap when you have the SP to do so
2) train until 30
In fact, this is (usually) the best way
* only cap the level before the decimal crosses the cap (xx.07)
* BUT, do not cap if...
A) compute how long until you'll cross the decimal AGAIN (in real time days, not levels)
B) calculate how much you could train now without capping
C) calculate how much you could train at 49 with capping
If B-C > 1, keep training.
Whoa! Complicated? That's math.
Example
---
EX1) You are at 24.6, and get 0.4 per level in gains. You have 24 SP.
strategy one: cap now, 48.6, level up = 49.0 + 5 SP
strategy two: keep training now, get an extra +1, so 25.6, level to 26, spend 23 SP = 49 + 6 SP
You gained an SP, but at the expense of training your NEXT thing. But what value did you lose there? Basically none, unless this lost time will prevent you from capping THAT on target. In my experience, the added value of training an attribute at 26 is something you'll really, really, REALLY wish you had at level 50. So you didn't really gain a SP, but you did gain TRAINING TIME by waiting until the decimal reached
EX2) You are at 20.0 and get 0.2 per level in gains. You have 29 SP
strategy one: cap now, train at 49.
- you get 15% per train. At 0.2 per level, it will take 5 levels to go over. Lets say 30 days. That is 12 trains, or + 180%
- so after 30 days, you'll be at 49 + 1 in level gains + 180% = 51, 80%, with 25 SP from those level-ups
strategy two: don't cap now, keep training at 20.
- you get 45% per train. Again, 12 trains = 540%. Training value will fall fast going up by 5, so lets say 520%.
- so after 30 days, you'll be at 20 + 1 in level gains + 540% = 26, 40%, with your original 29 SP + 25 from level ups
- lets assign the SP to get to 49, 20, then another 4 to 51, so
strategy one: 51 + 25 SP
strategy two: 51 + 30 SP
WHOA NELLY training longer helped. Sure, you missed 12 trains on something ELSE, you could have gotten that exact 520% on some other random attribute you wanted to do next, so you didn't actually gain SP, but what did you gain? You just got "free training time" Now you have even longer to train your other stuff, and once again, at level 50 the real value of your build is in STILL HAVING STUFF LEFT TO TRAIN!!!
EX3)
This is your problem right here. Lets pretend you don't want to train at 49, and in the punter situation decimals don't matter because it just +1. You have what, 27 punting? At level 4? So you have 30 SP. If you added those all now, you'd get to 53 (22 to 49, 8 to 53). Or, at level 3, you were at 26. You had 25 SP then. Cap to 49 costs 23 SP, so you could have gotten to 50. Then level and gain to 51. Then 5 SP, you get right back to 53 like before, but now with a left-over skill point.
So you trained at 26, but lost a skill point. I'm positive the value of training at 49 including that skill point is better than you training at 26.
I hope I didn't mess up any elementary calculations, I'm using google calculator and whipping this out really fast, but I hope my point is clear....
AKA, not clear... slowbuilding is no simple formula about when to add your skill points. You actually have to analyze the best way for your build.
Stop listening to people laugh at you for having 85 SP, or for capping at a base value of 23. Neither are wrong, the people laughing are probably the ones with sub-par slowbuilds.