First time poster here. I have conducted a quick experiment to test the new training system.
Here is the test subject (the screenshot was from after today's training):
I picked him up as a 41% 18 y.o. fast trainer (FT) from auction. He gained about 6.5 skill% per 6x Lv4 Press, so in the words of the legendary Al Svanberg, his TEF (skill% gained/condition loss) was 0.36. After 3 seasons of mostly Gym and Press training with the occasional Fast Counter-Attacks (FCS) thrown in for Crossing, this is what he looks like now.
After the second match today, I gave him 6x Lv4 Press and nothing happened. Nada. 0 skill% increase. As a 7-starred FT under 21, his TEF should be 0.14. In other words, he should have gained between 2 and 3 skill%. I was confused.
So I decided to check the forum and of course I found out that the training system had been changed. To mix things up a bit, I gave him 6x Lv4 FCS just to see what would happen:
Incredibly, he gained 10 skill% from this set of training. His TEF should be 0.14 as mentioned earlier, so the expected gain from this set should be 22.5 x 0.14 = 3.15 skill%. Just to make sure this was no fluke, I did the same training set twice more:
From 3 sets of 6x FCS, he has gained 25 skill%, or 8.3% per set. This equals a TEF of 0.37 ((25/(22.5*3)), which is more or less what a 3 to 4-starred FT would gain. This is insane! Here is a 7-starred player who trains his lower skills as if he were still a 4-starred player!
As a side note, he gained 10% in Crossing (white skill) and 5% each from the 3 grey skills. So I think it is safe to say that white skills grow twice as fast as grey skills just like before.
Conclusions
1) Players gain skills just as fast (if not faster) as before if you train his lower skills.
2) You do not need to mix and match each individual set of training. This was my main concern when I found out the training system had been changed, since tailoring each individual set of training for each player to get the optimal result would be far too much work. As you can see from above, I have given my test player 18 FCS training in a row and the results are still good. You can continue giving your players 6x training with a particular drill, as long as you mix it up once certain skills are high enough.
Conjectures
1) In this quick experiment, the 4 skills being trained are all below 80% (i.e., 4 stars or below) and they were growing at a rate as if the player himself were 4-starred. I have a suspicion that the growth rate of a skill depends on the current level (star rating) of that particular skill, instead of the overall quality of the player like before. To put this another way, say you have a 7-starred player with certain skills that are under 80% (4 stars or below). If you pick a drill that only trains those low skills, the player will grow as if his overall quality were 4 stars or below. This is a pure conjecture, I could be completely wrong about it.
2) Someone asked about how training NP/SA works with the new system. I suppose there are only two logically explanations. Either i), it doesn't matter which drill you choose, just like it was before, or, ii) using the drills that train the lowest skills would provide the best results.
3) If I understand correctly, it seems like the long term effect of the new training system is that we will stop having players with 300% in some skills and 1% in others. On average, the white skills will be higher than the grey ones because they are still easier to train, but the gulf between white and grey skills will definitely decrease as the seasons go by. This is not a bad thing. Managers have been complaining about mutants and upselling for some time now, and I believe this is Nordeus' attempt to the mitigate the problems. If the new training system works the way I think it does, it will be extremely difficult to create 3-starred players with nearly maxed out white skills and extremely negative grey skills like the ones you see in the world's top associations.
-George