Some one averaging 70 in first 10 test has a very slim chance of keeping that average if he plays for 60 tests.
Some one averaging 15 has a very slim chance of keeping that average if he plays 60 tests.
Notice the word, slim chance. It's not impossible, but chance is extremely low. That's what some posters are saying.
In both cases, newness combined with unique circumstances may result in anomaly for a short period. It has happened before as well, not necessarily in first 10 tests, but 10 successive tests.
This is based on long history of cricket. You are convinced that this time it's different, but others are not. We just have to agree to disagree.
No I'm not convinced, at all, but I'm interested in making the argument. I'll be the first to acknowledge the unlikelyhood of someone like Abbas appearing, but once he appears, I'd like to sit and think with it a bit. The law of averages is not a law, it is statistically speaking nothing at all. That is precisely why someone like Bradman can exist. But we can also look to less spectacular anomalies, like Adam Voges, who retired with a batting average of 61 after 20 Tests, or Mickey Hussey, who if he had retired earlier could have done something similar.
Now, as to the points you make. Someone who averages 70 after 10 Tests probably won't play any more tests. But if he is really that awful for that many Tests, I don't think it is unlikely that he will continue to be that awful. We can probably find hundreds of examples of batsmen who bowled occasionally with little result and thus carried averages of around 70 across their careers.
There is a difference between predicting the outcome of a dice roll, in which any side has equal chance to come up, and you would expect all outcomes to equalize over time, and predicting the outcome of something like a bowler's performance, where every incremental performance actually tells you something more about the bowler's abilities, and thus shifts the entire distribution. By the time Bradman was half way into his career you would not have bet
against him continuing to average 100.
As for someone averaging 15 in Tests over a long period of time, it is of course as per the available evidence extremely unlikely, and has not been achieved even by all time great bowlers. But since it has actually happened that someone has managed to get to 10 Tests with this average, one would have to ask, how often has
that happened? If we are already inside the anomaly, do the usual likelihoods apply, or do we have to recalibrate? Can we find parallels to Abbas' case?
I can only think of one pacer myself, Philander, who sat at 15 and thereabouts after 10 Tests, and he managed to sustain an average of 18 until 20 Tests. So it doesn't seem completely implausible that Abbas could do the same. Particularly since Philander's initial burst was done on extremely helpful South African and New Zealand pitches.
That is why I think it is a fair gamble that Abbas sustains his average or even pushes it lower over the SAF series, before it starts to rise again. If he only plays 30 something Tests, which is also not unlikely, given his age and Pakistan's schedule, I don't see why he couldn't stay below 20.