It was reading blog of DevZing on development of games and it made me reflect.
One of the greater dangers than we have testers is “the contamination”, is an aspect that repels directly in the balance of the product. It is necessary to differentiate, tests to detect bugs (already we will speak seriously at another moment) to balance a game.

Normally a base is created where the game is theoretically to point, later, when already they have reviewed bugs, it balances on emulators and finally tested by people other people's to the project. These last ones, if that plays, the rest has been done test after test.
Another subject is the market towards which it is directed, mobile, consoles, Pc… the design and tests are different in each case. It is not the same to validate a console that a mobile.
For it one of the options is to have an equipment of testers and a beta-testers set, but this in most of the companies usually does not happen, more workers = too much money…

It is like everything, there is very good ideas and also good methodologies but in my opinion all are not applicable to the 100%, I am explained, is necessary to be realistic at the time of seeing the duration of the project, which usually happens habitually, since we are in a market that does not allow developments of more than 45 days if you really want to be inside.

In order to avoid it it is possible to have prepared patterns with the general batteries of tests (the programmers also must say this way), with the purpose of proving aspects that can be common in the games, sees menus, options, audio modules etc.
And a battery of tests makes specific, grouped by different sorts for example, tests for a Arcade that for a strategy game.
This avoids in a high percentage to test aspects of the game that do not have relevance, in addition, we proved those parts that more interest to him to watch the programmer.
Another one of the aspects that there are to consider is the type of game, if he is casual or hardcore, the content and the complexity that it has, according to this also we would test of IA: it is not possible to be validate in only a pair of games that each component has the behavior that is expected. There will be difference to each other is a Pac or a The fourth Seal the diversity of the behaviors and states causes that a more exhaustive analysis is made.