There was a huge thread on supposed problems with "the Math" in D&D Next ([5E] The Maths is Wrong!) on RPG.net recently. One of the arguments that the math was wrong struck me as actually showing a completely different set of problem with WOTC editions of D&D in general: the way WOTC editions handle skills and set difficulty numbers.
The argument in the thread went something like this. The 5e rules set the DC for walking a tightrope at 25, which means that even an expert thief, who has dedicated everything to tightrope walking, couldn't amass enough bonuses to guarantee walking a tight rope more than 60% of the time. Instead of assuming that perhaps the DC should be lower for walking a tightrope, most posters complained that this showed how "the Math" in 5e just did not work.
Someone then pointed out that performers in the Cirque du Soleil, who do things like riding a unicycle on a tightrope while juggling, manage to do what they do in performances all the time without falling and killing themselves. In fact, there has only been one death in something like 25 years worth of Cirque du Soleil performances. This was used as further evidence that "the Math" for 5e was completely off as there was no way D&D Next characters could ever be skilled enough to do this.
However, it sounds to me like there are three problems here. None of which has much to do with "the Math" but with the way Next (and other WOTC editions of D&D) handle DCs and skills.
First, as I mentioned above, the DC for tightrope walking is probably too high. More correctly, it doesn't take into account the exact situation. A true tightrope that is carefully stretched to the correct tension and is level is going to be much easier to walk than a rope stretched across a crevice in a cave that probably isn't stretched to the correct tension and is unlikely to be completely level. And this is before you take into account environmental factors in the cave (Is it damp? Is there a downdraft or an updraft? Etc.). So while a DC of 25 is probably way to high for a royal command performance in the castle courtyard, it might be just right in a damp cave where the far side of that crevice is a couple of feet higher the the near side. DCs really need to be set by the GM based on the specific situation, not set by some standard DC list for various activities in the rulebook. The GM should just decide if the specific instance of the task is easy, average difficulty, hard, very hard, extremely hard, etc. and assign the appropriate DC. Some examples of things that are easy, hard, etc. could be in the rule book, but they should just be examples.
The second problem is WOTC D&D's handling of skills. The rules seem to make it sound like you should roll for every use of the skill, no matter how normal and mundane -- or at least they seem to read that way for many players. IMHO, if you have a skill, normal usage of that skill under normal circumstances should not require a skill roll at all, the character should just succeed at the task. (And no, "take 10" and "take 20" do not actually provide this automatic success). Otherwise, even a low 0.01% chance of failure would mean that failure is very likely to happen more often than it does in real life. Someone with the acrobatics skill, for example, should not even have to roll to successfully cross that tightrope strung up level and at the correct tension in the castle courtyard unless there is some reason that it would be harder than normal (it's wet, greased shoes, etc.).
The third problem is that skill failure and skill fumble should be two different things. For a skilled person, failure at walking a tightrope should just mean you have problems and it takes you longer; or you have to make a DEX check to avoid dropping something as you flail around trying to regain your balance; or you can't make it across for some reason and have to go back; or the like. You shouldn't have to worry about falling unless you fumble the skill check and therefore have to save to grab the rope before you fall off or something. If you don't have the skill at all, failure probably means a disaster of some type, but if you have the skill, it shouldn't.
The Cirque du Soleil example isn't so much a rules issue as it is a common sense issue. For a person with the appropriate training (skill), practice often makes even the extremely difficult routine. A Cirque du Soleil performer doesn't just get up one morning and decide to add a new "riding a unicycle on a tightrope while juggling" routine to his performance. He develops the routine and practices it until he's as close to perfect as he can get (which may require weeks or months of practice) before he debuts the routine before an audience with the tightrope high in the air and no safety nets in sight. By the time, the performer is doing the act in front of an audience, he has practiced his specific routine so much that it has become a normal task for him. Even though his routine may be extremely difficult with a DC of 40, it has become a normal task for him because of his many hours practicing the exact steps of the routine. And as a normal task, no roll is needed for a basic success.
If I were running a skill-based game and had a Cirque du Soleil group of PCs I'd probably still require a skill roll when the character performs his routine but it would be to determine how well the performance went that day. A success means you did well enough that any minor missteps aren't noticed by the audience (a critical success would mean even your fellow performers/trainers don't notice any). A failed roll would mean that you had a problem that the audience could not help but notice. A fumble means you had real problems and have to make a second skill roll to avoid injury. Success on the second means just a really bad personal performance -- the sort that might ruin the entire show for the audience. While a failure on the second roll means not only did your poor performance ruin the show but you also suffered a minor injury (a fumble means you also suffered a major injury). If the character has an appropriate Performance skill, a high roll there might allow you to make even a bad mistake in your routine look like a part of the show.
In summary, I think "the math" is getting the blame for a problem that can't really be fixed by changing the math. What is needed is flexible DCs, recognition that skill rolls really should only be required when the circumstances aren't normal, and an understanding that most skill failures just mean you failed and not that a "near worst case" disaster has struck.
The third Sponsor Microlite81 (and Lords & Wizards) Mini-Drive for RetroRoleplaying Cancer Fund is running now. The goal is $500 with the following giveaway items: First Fantasy Campaign (1st Printing), Judge Guild Character Chronicle Cards, Basic Boxed Set (complete including original dice), Expert Boxed Set (complete including original dice), Journey to the Center of the Circle (from Wilmark Dynasty), and The Vampyre's Mirror (also from Wilmark Dynasty). Each $10 donated gives you one chance at one of these items. The top three donors will have a separate drawing for a set of 1st printings of the original three TSR adventure modules (The Giants series as three separate modules: G1, G2, and G3) This is in addition to the usual PDF downloads and other benefits of a donation to the RetroRoleplaying Cancer Fund and sponsorship listings in the upcoming Microlite81 and Lords & Wizards games. To get help us pay our cancer treatment related bills (and to get access to some special downloads and a chance at the above-mentioned items), send a donation in any amount -- small or large -- to me via Paypal. Thank you!
As of 1 August 2013, $115 has been donated in this mini-drive, an additional $385 in donations is needed to trigger the drawings for the items mentioned above.