Yes, stats such as these can be attacked this way, and that's way making claims from them is problematic. Most of them don't really determine anything about the causality; they are just just-so stories, and it is very possible to make up other just-so stories that explain most of the stats.
This is an unfounded inference. One can just as well say that winning from one set up shows that you well when you can relax more and don't have to battle as hard mentally. And when including from one set down, why not mention from two sets down? As far as I know, Fed has the more than anyone of those. That could very well be interpreted as a sign that he is supreme at scrambling mentally when he is having a bad day. And afaik, Nadal has very few of those win, which could be a sign that he is not as good at scrambling mentally when his game is off. But – in line with my point here – such theories are just conjectures. It's okay to use stats to form some theories of course, but some humility as to what the stats really show is good.
1. First have in mind we are debating a subject that is not written anywhere. How good Nadal is Federer is etc are all subjective. I can have nothing to show for my claim yet believe Isner serves better than Roddick. So how am I going to tell others Isner serves better? Data points. Of one kind? Na. Of many kinds. How many of them? As many as possible. This is where I tried to provide as many stats that I could to get an indicator. I have relied on ATP's website as well as some other sources.
Does Federer really have a match-up problem against Nadal? Most of them happened on clay, no? Yes. The rest happened less no? Yes. So a logical conclusion cannot be formed?
You can argue that, but not for me. The point is 10-23 h2h tells you the story to X degree. Of course it is not perfect. There are A and B lower than X, and determinable or nondeterminable degrees Y and Z that are higher.
It is quite clear, or should be at least, for any sensible sport follower that stats present that X degree. The higher degrees are always welcome. I did what I could present with, rather than pulling out conclusion of thin air. I dont need the
humility to explain because the factor X should be apparent for anyone with commonsense. That the combined set of data points indicate something to X degree. You can form an opinion based on X or not depending on how sound you think X overall is. You can also rubbish it if you think some factor disturbs the overall balance of X. But you can not pinpoint individual weakness of various attributes of X. Because it's the totality of the entire set of stats that is being the indicator here. Any degree higher than X for most laymen like me would be unattainable.
For eg, your point that Federer's age/longer play should be considered is a valid one. Because it affects the totality of my inference. But things like Federer's peak play ended matches before final etc isn't valid one because that is nitpicking individual attributes. You can do that to anything.
I can provide ace rate, first serve %, double fault rate, mean serve speed etc and not prove but indicate one is a better server. On the other hand one can also criticize it saying it doesn't consider the tournament, the quality of opposition etc which while is fair, doesn't disqualify the findings of the original guy. In any stat analysis we give law of averages like principles a lot more importance. Stats, records etc are there for a reason. It's not complete, but it's the most objective way to approach.
2. It is not an unfounded inference. Let's agree to disagree, but at least few in ATP is with me. As I said before, it can be attacked, but it adds to my view of what mental strength is.
Indeed it shows scrambling mentality. It's otherwise called choking. To not choke you need a certain degree of mental strength. In other words another data point for me.