[Suggestion]: Option to randomize piece sequence and garbage holes

12/19/2013 7:43:03 AM
Gravatar
Total Posts 514

[Suggestion]: Option to randomize piece sequence and garbage holes

In my opinion the outcome of large rooms is not random enough. For example I played FFA in a 20 player room yesterday and got 4th place (+/-1) 10 matches in a row. My main problem with Cultris is that the time based combo system puts emphasis to speed rather than good piece placements, which I sometimes refer to as artificial skill gap. One has to do 8 combos in order to send huge amounts of lines, but 100 BPM players are not able to do this midgame. As a result 100 BPM players have no chance to beat the 150 BPM players.

So I think Cultris 2 should be more randomized. I'd like to see an option (a check mark) that let the room owner randomize sequence and garbage. You know, in Cultris 2 all players get the same piece sequence in a match. The positions of the garbage holes are also the same. It's good for a 1vs1 tournament format but it is bad for large rooms.

12/19/2013 1:15:55 PM
Gravatar
Total Posts 59

Re: [Suggestion]: Option to randomize piece sequence and garbage holes

I'd say, that the speed itself is not the main reason.

Yes, speed matters a lot. Even with changes according to your suggestion, there will be no way (except really bad luck) for a 100bpm player to beat a 150r. I'm pretty certain that any player can win vs a faster player, as long as the speed difference is not too much. 130 vs 140 - that works. 90 vs 100 might work as well.

But.
As I said, I think speed is not all. There are some players which can play fast, but do many mistakes and lose vs slower (~15%) players. I think alongside of getting faster in speed, one is also improving block placement. Starting off with block combinations, artificial gaps and finally the spins. I guess players do not simply play 150bpm without these skills. At least the ones I see are able to place very well.

I agree that it's the combo which determines the winner in the end. And speed is a factor to that. But having a solid downstack with 1 line/piece allows you to compensate a certain lack of speed to get the same combo. Players who have both are simply much better.

That aside, I don't think your suggestion is a good one (sorry). I see your point, but I don't think we can simply solve it with that. The way the combo system works - 150bpm players will still do 7s/8s without even trying. Because in the first few combo counts you'll get time added and later on even removed (I think there was a thread about that some time ago).

All in all I like the fact that everyone has the same premise every game. If 2 players match in speed, the block placements will determine the winner. I don't want to involve more luck in the game. A 100bpm player can't beat a 150bpm player - is that wrong? I don't think so.
Isn't the formula for cultris somewhat like: skill = speed*1.5 + placement + spins/2
I have the feeling that you want to aim towards: placement*1.5 + speed + spins

12/19/2013 7:13:21 PM
Gravatar
Total Posts 514

Re: [Suggestion]: Option to randomize piece sequence and garbage holes

The formula for Cultris is roughly:

skill = speed*speed*placement

I'd rather like it to be:

skill = speed*placement

(speed is measured in average BPM, placement is a factor between 0 and 1, 1 means perfect stacking, 0 means random piece dropping, placement is > 0.5 for most players)

I didn't want to reinvent Cultris, I just wanted to reduce the negative side effect of a time based combo system. I am sure it's effect would be noticeable in a 2 player room. But it's true that the effect would be rather small in a large room (game takes longer, more opponents must have unlucky garbage/sequence.

Another way to make FFA more random is by making the garbage distribution more random.  Similar to Cultris 1, but but not as biased/unfair. For example 20 % of your combo (the whole combo) is sent to 1 particular opponent and the rest is randomly distributed.

------

I'd like to see Simon making statistics involving bots. At first, how big are the chances of a slower bot to win against a faster bot? Bots' speed could be: 100 BPM vs 120 BPM, 120 BPM vs 150 BPM and 100 BPM vs 150. Then, what's the average position of a slower bot in a room full of fast bots? For example a room with 6 bots: 100 BPM vs 110 BPM vs 120 BPM vs 130 BPM vs 140 BPM vs 150 BPM.

12/20/2013 12:39:39 AM
Gravatar
Total Posts 348
cuz my hips don't lie

Re: [Suggestion]: Option to randomize piece sequence and garbage holes

I kind of agree with making a bigger skill gap by making the randomizer even worse than it already is.

But if you change the fact that every player gets the same garbage and pieces does not make it more difficult, this adds in the factor of luck. The great thing about c2 right now is there is no luck, and if you get "lucky" you created it yourself. 

12/20/2013 9:20:33 AM
Gravatar
Total Posts 626

Re: [Suggestion]: Option to randomize piece sequence and garbage holes

(Following the discussion)

but in the opposite direction: People asked for less randomness. Very interesting.

Letting bot's play against eachother is not meaningful. All bots play with the same strategy and skill, and even the exact same piece placements - so the faster bots will always win. That holds, unless I add randomness (which the bots in the game have). But then the outcome would only depend on the amount of randomness that I artificially added.

12/20/2013 7:24:12 PM
Gravatar
Total Posts 514

Re: [Suggestion]: Option to randomize piece sequence and garbage holes

With less randomness you mean the randomizer, don't you? Many players are used to 7-bag randomizer (which is also called "Random Generator", although it's pretty much the least random randomizer). It's no wonder that those people dislike Cultris or don't even give Cultris a chance. On the other hand, there are also people like me who are used to memoryless randomizer (which is the most random randomizer). Honestly, I dislike that Cultris 2's repetition rate is so low (chances to get a chain of equally shaped pieces is pretty low).

I don't understand your argument regarding bots. Statistics involving real players would be meaningless, because the average 150 BPM player will have better stacking than the average 100 BPM player. It's desired that the bots play with the same strategy and skill. This way you can make research how much the garbage output is affected by speed (the garbage output pretty much decides who wins a 1 vs 1). Let 2 bots play 100 games and log how many lines they send, block and downstack (plus their speed). Of course, humans play differently than bots (humans will stack for higher combos in particular), but it's still a pointer.

The faster bot will not always defeat the slower bot, provided the speed gap is not too big. The play of 2 differently fast bots will desynchronize after a while, because the faster bots sends a line which the slower bot won't cancel. If this doesn't happen fast enough, you could also make the bots drop the first piece on a different spot.

I'd really like to see how random garbage/piece sequence will affect the outcome in a 1 vs 1. Can you let a 100 BPM bot play 100 games against a 120 BPM and log how often the 100 BPM bot wins (once with random garbage/sequence and once without)?

12/20/2013 7:46:58 PM
Gravatar
Total Posts 52

Re: [Suggestion]: Option to randomize piece sequence and garbage holes

The rankings in FFA really depend on the player. If a player is a consistent downstacker, then I would expect their rank to not change much between games. On the other hand, if they are not consistent, their rank should be expected to vary between games. If the game was made more random for consistent players, I imagine it would be much worse for the rest.

I think changing the penalties for no line clears in the late stages of a combo may alter the dynamics and place more emphasis on stacking. Increasing consecutive clears in the late stages of a combo may be better.

What ticks me is when a combo is limited by time, even when it consists entirely of consecutive line clears!

12/20/2013 8:27:54 PM
Gravatar
Total Posts 344
The only way to get smarter is by playing a smarter opponent

Re: [Suggestion]: Option to randomize piece sequence and garbage holes

I don't think that more randomness will improve the gaming experience. I think that it would lead many players to conclude that winning is just possible with enough luck.

Morphling is on the right track: Reward good downstacking. Maybe the combo time could be influenced by some more parameters like:

  • Stack height
  • Garbage height
  • Type of cleared line (Tetrimino or garbage)
  • Line clear consecution
12/21/2013 10:27:51 AM
Gravatar
Total Posts 514

Re: [Suggestion]: Option to randomize piece sequence and garbage holes

I think that it would lead many players to conclude that winning is just possible with enough luck.

Nope, players would be proud when they win, because it would still happen very seldom. Everybody would still play with Cultris 2 randomizer; it's just that everybody would start with a different seed.

I think changing the penalties for no line clears in the late stages of a combo may alter the dynamics and place more emphasis on stacking.

Can you really prevent making no line clears at the end of a combo? I don't think you should be penalized for what you cannot influence.

What ticks me is when a combo is limited by time, even when it consists entirely of consecutive line clears!

That sounds interesting. Are you suggesting the following: Your last piece placement was a line clear and you run out of time. If your next placement is a line clear too, then the combo counter is still increased. I guess with such a system people would try more to go for 3 wide combos which puts emphasis more on stacking, because 3 wide is harder to stack.

12/21/2013 1:20:12 PM
Gravatar
Total Posts 52

Re: [Suggestion]: Option to randomize piece sequence and garbage holes

That sounds interesting. Are you suggesting the following: Your last piece placement was a line clear and you run out of time. If your next placement is a line clear too, then the combo counter is still increased. I guess with such a system people would try more to go for 3 wide combos which puts emphasis more on stacking, because 3 wide is harder to stack.

Yes, something along the lines of that, but without giving unlimited time. So maybe if time could be added for combos where there is a minimal amount of pieces that do not clear lines. For example, if you reached a 10 combo and only missed 2 line clears, say 0.2s could be given. Missing 1 line clear could give 0.3s, and missing none could give 0.4s.

I think in effect, it would be the same as the current system, with the only difference being it would penalize or reward combos with minimal line clears a little more.

 

 

Share This Using Popular Bookmarking Services