Help! Antiskate with only a weight...no dial, and she's skating away!


I have a project rpm 10 carbon with 10cc evolution tonearm that has a weight on a string for antiskate. There are three notches on which to attach the string based upon the tracking force range of the cartridge. I currently have an ortofon cadenza bronze tracking at 2.5g and have the antiskate weight in the appropriate notch (according to the Pro-ject manual) from which it hangs. The table is level--checked and adjusted to ensure. The tracking force is at 2.514g (the range for the cadenza is 2.2-2.7 with 2.5 suggested by ortofon) checked with a digital scale (Riverstone Audio digital scale). The soundstage sounds great, vocals are centered, other instruments are placed in space according to the recording... Also the alignment was carefully set up using the WallyTractor and is spot on. 

But sometimes when I lower the stylus to the lead in groove, it will slide very quickly towards the spindle as though no antiskate were present (it doesn’t skip over the record, it falls into the first song groove--and yes I have confirmed that the stylus is present). But it’s a big jump vs just sliding into the groove.

So I found a blank side of an album and lowered the stylus onto the surface and it immediately slid all the way across the surface towards the spindle as though no antiskate were in play. I then disengaged the antiskate weight and experienced the same (expectedly so). But there seemed to be little or no difference between antiskate being engaged/disengaged.

So I engaged the weight again and lowered the stylus, but this time I placed a little extra force on the weight with my finger and was able to get the tonearm to stay in position--applicable antiskate force in play with this extra force. Of course, I have no way of measuring how much extra weight I applied.

The help I need:
Why is the recommended antiskate parameters set by pro-ject seemingly having no effect?
Is something else wrong?
The table and tonearm are obviously manufactured to handle this level of VTF, no?
The tonearm wires don’t appear to be impeding the arm movement.
What can I do to remedy this?
Do I need to do something to remedy this?
I wonder if I’m causing harm to the cantilever with what appears to be no antiskate, yet the music sounds great and the Analogue productions test LP record antiskate tracks "sound" equal to my ears. (But my ears aren’t young anymore, so I don’t think I can place full confidence in that audible test).

Any thoughts, suggestions would be greatly appreciated.
cabalaska
stringreen:
I'm sure it's not broken, bent etc. It's brand new and only has about 15 hours on it so far. I have inspected it under higher magnification and everything appears intact.

lewm:
but it is skating across the whole record--that is a blank record.  Which tells me (with what little knowledge I've gleaned thus far) that there is no antiskating force in play.  You said the grooveless record experiment is meaningless, and I agree it certainly isn't the same characteristic as if it were in the groove, but shouldn't there be some antiskate character being observed.  I did this with my rega rp6 with 2m black and antiskate set to about 1.2 or so and it didn't slide all the way across the blank record--which led me to conclude the antiskate was in effect? no?

How would I go about checking the internal wiring? The wiring attaching to the cartridge appears loose and without a drag effect, but of course I can't see the inside. 
Here are two videos of the issue: The first is with antiskate engaged which includes the additional weight (total of around 5.5g) added to the string.  
The second is without antiskate.

https://youtu.be/cGG9-H_ZD6w 

https://youtu.be/uRm_xAZIdUM
millercarbon.

Thanks. your post gives me a lot to consider and makes sense. Agreed that the VTA on this arm is a PITA, but at least it's available whereas Rega's are not. I'm trying not to obsess over this as each day passes, but the nag is there in the background for now.  I try to remind myself, that many of you who've been at this for a long time don't give much credence to AS and to the contrary find it sounds better w/o in some systems.  I'm still weighing the pros cons and whether this arm reveals or hides with/ without.  Either way, thanks again for all the feedback. It is much appreciated. 
With only a few notches just try them all and listen for the one that sounds best, this worked for me with a Naim Aro anyway which has a similar arrangement.
With a replicant stylus the VTA is pretty important to get right but you should haer when you land on the right setting, it might take some finding.
Cabalska, a likely problem is that the arm was installed incorrectly or the hole for the arm is out of spec. You have to measure the overhang. Check Pro-Jects spec for that arm. If the arm is located too close to the spindle it will skate incessantly. It the arm tube is longer than it should be the same thing will happen. Otherwise you have been doing all the right things. On a groove-less record the arm should drift slowly towards the spindle at the end of the record. 
The anti skating force should be 10% of the VTF. The skating force will decrease slightly as you go towards the center of the record. It does not decrease at the null points. That is just lay instinct. 
If the arm is in spec then somebodies math is wrong and you just add weight until you get the right effect. Or you could tell Pro-Ject that the table is defective and you would like another one. Actually measuring the anti skating force is possible but the devices to do it are either expensive (the WallySkater) or unavailable (my Gizmo). Trying to gauge your particular problem with a test record would be difficult but you could try it.
Lewm is correct in that without anti skating the right channel (outside) groove will miss-track and distort first, too much and the left or inside groove will distort first.
Although anti skating is a ball park proposition it is extremely important for stylus and record wear not to mention sound quality. As the force drifts to far one way or the other it pushes the cantilever into a non-linear zone where moving the stylus one way takes more force than moving it the other. It also displaces the coils, magnet or iron out of the center of the magnetic gap. Any asymmetry is unfavorable for stereo reproduction. If there is a reason straight tangential arms sound better the lack of any skating forces is a more likely candidate than a reduction in tracking error.
God luck in figuring it out. Most people would never have noticed there was a problem.