I don't fully agree that this statement:

12 hours ago, LorenzoGatti said:

Penalty kicks are not a different game, they are only a specialized form in which the ball is kicked into the goal unusually fast and intercepted by the goalkeeper with unusual difficulty: not only the same skills, but the same rules as normal play.

...can agree with this statement:

1 hour ago, LorenzoGatti said:

Is every sport in which a ball is used to score points and the team without the ball defends against the scoring attempts of the team with the ball the same? No

I think it's clear that there is a whole continuum of similarity here and that what one person might call "a specialised form" someone else might call "a different game". (Besides, penalty kicks don't really have the same rules as normal play, since other players are not allowed in the penalty box, the goalkeeper has to remain on their line, the attacker can only touch the ball once, etc.)

As such it's not about trying to arbitrarily find the line where the game or the 'form' changes into a different one, but about understanding how far a game might change during play, and how comfortable a player is with that.

It's interesting to note that many soccer players and fans are not happy about penalty shootouts. They feel it is a less fair way to settle a match, and is not entirely in keeping with how the game should be played. So obviously changing the form of the game has issues outside of just videogames.

A game should stick to what it's good at. Say I'm making a platforming shooter (which I am :p). There's basically two elements that will need attention at the game's core: platforming and shooting. So I should spend plenty of time tweaking my mechanics so that the platforming and shooting and everything else in the game complimenting those two things are as good as I can make it.

So if I suddenly turn around and suddenly make it a puzzle platformer halfway through the game, two things happen. The first thing is that the player will probably be frustrated with my game. He/She spends so much time mastering the mechanics in the action platforming and then suddenly they have to throw all that skill away because the game is a puzzle platformer now and I'm demanding they play what is essentially a completely different. They might just want to shut the game off and uninstall it! They wanted an action platformer, and they got it for some time but then I yanked it away from them. I guess you could say I only gave that player half a game. Now let's look at some examples that are actually out there in the world today. A lot of people already mentioned the games from the 80s and 90s. For other examples, I'd even remind us that some action games in the early 2000s (I think, it may have even gone onto to early 2010s if I recall) would have these great, good feeling action mechanics in play, you're feeling like a champ, downing enemies...and then suddenly you have to solve some random puzzle in the middle of the game to go back to bashing enemies. I don't know about anyone else, but that annoyed me and I'm glad we don't really see that anymore.

Then there's the development side of things. It's just too difficult on me as well as most if not all developers to make what may as well be two games in one. Having to shift mindsets throughout development...go into action building mode then back to puzzle building mode over and over is taxing and honestly, not fun at all. And most of the time, splitting your time between two very different mechanical styles just makes the whole game bad.

Get good at one thing. Try to do a whole bunch of things in a game, and most likely all those things will be mediocre and won't leave a lasting impression on a player. Whereas you do one thing REALLY well, and it'll stick with a player in a good way,