It depends on the market. A honest comparison with other games of similar type and quality will provide reference prices.

We'd need a heck of a lot more information than just "a Side-Scroller game". Do what LorenzoGatti said.

Never do 1$ ! Do 0.99$

but seriously, alsways use those .99's because people will usually calculate the cost rounding down, and will with higher chance buy your game.

$1.99 is better then :)

The frank answer is that on iOS, essentially no software is "sold" for a cover price these days. Any price at all is too much friction for people to overcome when deciding to try out your game -- If you can't get over the hump of people wanting to try it out, then you have no downloads, if you have no downloads, you'll never make it to a top-apps list, if you don't make it to a top-apps list, then no one will ever hear about your game. That's why essentially all iOS software today, games especially, are given away for free and the developers make money by selling extra content, hocking various micro-transaction schemes, or advertising, to people who have already downloaded the game. Basically, the model is to give away the razor in order to sell the blades.

Unfortunately, as you can see in the games that are popular on the marketplace today, it can be very hard to monetize just any game with post-purchase transactions. Something like a run-and-jump scroller would be quite difficult to monetize effectively through micro transactions (which are generally the most lucrative of the reasonable monetization strategies). You could give away the first "episode" of your stroller for free, and then charge $0.99 for the rest of the game's episodes (This is the same shareware strategy that buit iD software, with Wolfenstien 3D and Doom), its hard to pull off though, because the part you give away has to be good enough to make your players be willing to pay for more, and then the stuff you sell them has to be good enough for them to be happy about their purchase. You can generally shoe-horn advertising into any kind of game, but its also generally the least lucrative of all options, unless your game is a viral hit like Flappy Bird was. Sometimes, people give away the whole game for free with advertising, and then charge $0.99 to remove advertising from the game, but I don't know how effective this is.

If your goal is to make real money (as opposed to making pizza-money from your hobby), then you need to design your game to monetize in some way. If it were me trying to monetize an iOS side-scroller, then I would do a combination of downloadable content (The Doom shareware model), and some kind of micro-transactions to enhance the game -- perhaps you could sell different characters that play differently, or selling extra inventory slots (something like how Super Mario World let you keep one power-up in reserve), or maybe enabling extra secret areas or more-frequent checkpoints. If there's no competitive aspect to your game, you're pretty free in the kinds of 'convenience' items you can sell, because you don't have to worry about breaking multi-player balance by letting people buy their way to winning.

You could make the app free and put ads in it.

Just to maybe bring this back around, its absolutely true basically everyone is forced into a freemium, micro-transaction, or advertising-based revenue model. In large part for the reasons I mentioned, but also in large part because of Android (which suffers the same visibility/over-saturation problems as iOS), where side-loading cracked applications (piracy) is too easy and common for devs to make money on an initial sale of an app on that platform -- To make money on android, you can't expect to make money on the initial sale, to make money post-sale, you need to design your game around that, and if your game is cross-platform to iOS and elsewhere, then your app monetizes the same way everywhere. Unfortunately, its simply the way that the herd is moving, and when the herd really gets moving you've got no real choice, in this instance, but to go along or just get trampled underfoot.

With that being said, the vast majority of the herd is the absolute last place I'd look for inspiration on how to execute freemium models or micro-transactions well. Sadly, there are far fewer companies and games that do freemium well, than there are who make enough money at it to keep the lights on. Many games are designed around exploitative mechanics which 'force' users to make frequent purchases of consumable goods to continue competing or progressing, others exploit the kind of flaws you find in those who have gambling addictions. You cannot absolve the player from their part in playing the willing victim, but still those kinds of monetizations aren't especially good for players or the industry. There are those few who do it well, who offer value for money and where purchases and traded virtual goods enhance the play experience for everyone, rather than simply buying a shortcut or unfair advantage (Some of the popular MOBAs are good at this).

The point I'm pushing at is that just because we are unfortunately pushed into a model that's not based on the initial sale, doesn't' necessarily mean that we have to stoop to exploitative mechanics or pervert our core gameplay in support of that. You can, in fact, design a freemium game in which your players are perfectly happy to participate in your ecosystem, and to receive good value for doing so.

Valve gave a really great talk at Steam Dev Days last month, entitled In-Game Economies in Team Fortress 2 and Dota 2 which I would highly recommend you to watch if you're interested in how to design freemium games that players love to participate in economically.

Speaking of Steam, if you really can't imagine your side-scroller as a freemium title of any kind (or DLC or Ad-supported), then perhaps the answer is that its not a good fit for the iOS or Android ecosystems. Steam, and to a lesser extent the Windows Store or other online storefronts for PC games, is one of the last bastions of the sticker-price model that is ubiquitous and accessible.