Differing Analysis

When I play a game on the tablet using the chess.com app, I look at an analysis of the game afterwards. As the app does not offer as full an analysis as I would like, I copy the PGN into the chess.com web page analysis. It gives me different results, like a different accuracy percentage and different numbers for mistakes, inaccuracies, etc.

I'm just wondering why this might be. Are the app and web page using the same engine?

Maybe different depth.

Because I do not have access to my home computer.

It is common for it to alter slightly due to factors like different engine depth, or letting the engine run longer (to reassess quality of the moves even if same depth). The "accuracy" (CAPS) is not an exact science - rather it was designed to give a decent "report card" on how each side did. It really isn't good to get too focused on exacts though, because MANY factors impact accuracy. 

For example, accuracy is tougher to score well with in tactical games. Also accuracy will likely be inflated for you if the opponent blunders early (if they blunder a Queen out of the opening, then basically ANY move you play is winning, so you are more likely to find good moves easier). There are many games I've scored high 90s in accuracy and other games by Grandmasters (which I view as much higher quality than mine) which score 50s or 60s in accuracy. There are just too many variables which impact it to take accuracy that seriously; it is just a fun little statistic to have though

Thanks for the answer and video. Yes, I've noticed that when my opponent blunders early, my accuracy is pretty good. A few days ago I achieved a 99.1% accuracy, probably because my opponents blunder on move 6 made things rather easy for me.