stars or points ?

[kasperpeulen]

I'm thinking about changing the score system. In some way this game teaches now people to do a construction as fast as possible. That shouldn't really be the point of the game. What do you guys think ?

[Hexstream]

First-time play

The first time a level is played, the player is expected to have to come up with a resolution strategy "from first principles", which is expected to take some amount of time, perhaps a very large one in some cases, depending on the prior knowledge of the player in particular.

If the player manages to complete the level in a competitive amount of time on the first try, it probably means either that the level is really easy or that the player had prior knowledge of the class of problems the level is a member of. There is no point in rewarding prior knowledge by accounting for time taken to complete a level.

I don't see how inducing stress in the player by accounting for time in the score calculation could possibly help the learning process at all, rather than hindering it (or at best, having no beneficial effect). I also take it as an axiom that the focus should be on elegant solutions requiring a minimal amount of moves, rather than speed.

For the harder levels, the player might not even be sure he'll be able to complete the level on his own in any amount of time, so formalizing the notion of "time taken" by accounting for it in the score is unnecessary at best, especially as the player is probably already informally keeping track of it. ("Come on, I've been on this simple-looking level for an hour already...")

Revisiting levels for further edification

There is value in revisiting already-completed levels until the notions are well and truly acquired and intuitive. I was rather shocked on my second play-through at how many of the levels I still had a hard time with, not remembering how the heck I had completed them the first time around. It's only after a few more playthroughs that I could confidently complete most levels really quickly.

There's also the aspect of trying to find more elegant solutions, requiring less moves than before.

The speed gained from practicing and using more elegant solutions is a reward on its own, and I don't think accounting for time in score calculations helps any of this.

Revisiting levels purely for speed and score

The last use-case is where the player already knows all the levels' solutions by heart and is completing them again for the sole purpose of twitch speed and increased score.

This is the only scenario where accounting for time is a net plus instead of being a nuisance.

I think given my arguments above, it's obvious that accounting for time in the score, if it takes place at all, should only happen in a completely separate "time attack" mode built for recurring (addicted?) players.

Conclusion

I think the time component should be dropped from score calculations entirely.

[kasperpeulen]

Good points. I think I largely agree, I introduced the time bonus as that was the only way I thought it was possible to make a highscore. Otherwise the first 100 ? maybe 1000? players would have all the same score. But even if I play the game myself now, I'm getting stressed about the time ticking. I think mathematics is about slow and precise thinking, and the game should reward slow and precise thinking more than it rewards fast memorization, and not the other way around.

I'm thinking now about making different modes, or maybe different "campaigns" is a better word:

1. The mode that the original game had, that gives you a star or something, and message if you do the task in minimum amount of moves, but nothing more. If you complete all levels in this mode, you will unlock a new campaign.
2. The second campaign is very similar but you can only proceed to the next level if you do the level in minimum number of moves.
3. The third campaign is a little bit different, you cant use any other tools then the original 5. The goal is the same as the first campaign.
4. The last campaign, the only way to succeed is to complete the level in a minimum number of moves with only the 5 original tools.

[Hexstream]

I introduced the time bonus as that was the only way I thought it was possible to make a highscore. Otherwise the first 100 ? maybe 1000? players would have all the same score.

I think the emphasis should be on "challenge yourself" rather than "challenge others". It's nice if you can also make a nice system for challenging others, but in my view it's at least secondary and perhaps even optional.

---

The second campaign is very similar but you can only proceed to the next level if you do the level in minimum number of moves.

I would be content with just seeing the following additional information next to each completed level:

1. the number of moves I used;
2. the minimum known-possible number of moves required.

In fact, this could pretty much replace (and become) the score entirely, which is currently opaque and requires "calibration" (sometimes even giving "negative" scores as a result).

[kasperpeulen]

Well negative scores where only possible a couple of days ago. That is already solved. The minimum score is now 100 and maximum score is 900. And there is a time bonus if you are very quick.

After thinking about it a little more, I think that the first round people play the game, I shouldn't give to much emphasize on the number of moves. I think the first round should be more about one of the most important aspect of mathematics, which is how to derive a propositions from a number of axioms and how to use such a proposition to derive new propositions.

I'm thinking that if I would put to much emphasis on the number moves right from the beginning, then people may think that it is all about doing something as short or efficient as possible, which I think is not the case.

Therefore I think it may be better to only add the "minimum number of moves challange" after completing all 25 levels.

[reverofevil]

I think the time component should be dropped from score calculations entirely. I second this.

Minimal moves challenge is nice. The problem is that composite actions are actually not a "one move". One could do it in less primitive actions with compass and straightedge.