• Play
  • About
  • News
  • Forums
  • Yppedia
  • Help
Welcome Guest   | Login
  Index  | Recent Threads  | Register  | Search  | Help  | RSS feeds  | View Unanswered Threads  
  Search  


Quick Go »
Thread Status: Normal
Total posts in this thread: 170
Posts: 170   Pages: 6   [ First Page | Previous Page | 1 2 3 4 5 6 | Next Page | Last Page]
[Add To My Favorites] [Watch this Thread] [Post new Thread]
Author
Previous Thread This topic has been viewed 34528 times and has 169 replies Next Thread
TheCosmicChz

Member's Avatar


Joined: Aug 24, 2003
Posts: 486
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

I . . . I think I love you =P

That explains so much. I should be able to get some better multi-play now. Now I just need to find some time to get some more testing in when I'm more rested and actually mentally agile.

I'm gonna have to think on the gold spawning probability later, my statistics skills are a tad rusty, but isn't the number of gems not truly a multiplier? If you make 3 separate clears of one gem each, you have the same probability of getting a gold as if you made one clear with three gems in it? In fact, since making individual gem clears is possibly increasing your team level, the probability could even be higher when making multiple clears (I guess lower too if one member isn't pulling their weight and dropping in level)?

In this manner multiple human players optimize gold spawn rates by coordinating only clearing one gem per round, until the team level is max, then doing whatever it takes to maintain level 5? But other then getting enough points to stay level 5, theres no incentive to clear as many gems as possible, and possibly disincentive as clearing too many rows on level 5 could leave you with a board too empty to work with.
----------------------------------------
--Dcurd :
Not the cap'n of Dcurd's crew
Stick-wielder of questionable sanity-or something
[Feb 16, 2007 5:17:20 PM] Show Printable Version of Post        Send Private Message    NotSoDreadDcurd [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

My gut feelings on this is that a good team needs one (or two) player who spends most of his time on levels 4 and 5 where gems spawn more often than they do at lower levels. Most of the players would spend most of their time at levels 3 and 4. They need one player who has a nice stash of crackers, which means level 2 or 3 or else somebody who is *very* good who doesn't mind dropping down to level 3 periodically to get more crackers.

Why the one player with lots of hints? To fill that team row or col where there's one spot nobody can determine.

I haven't tried to do any analysis on Nibbler. It's more fun to just play it and see what happens.

If you want the formulas for spawning gems, and for going up or down a level, just ask.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 16, 2007 6:15:21 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Looking more specifically at the chances of getting gems and foraging gold (with current formulas, which are subject to change of course).

gemPercent = 10 * (difficulty - 1)
gemMultiplier = number of carts the player used in the round (not just solo carts)
gemProbability = gemMultiplier * gemPercent * 0.01

This means that if you are playing on level 0 or 1, you don't get any gems, ever. If you play with a team, the odds are much higher that you will have more than one cart on your board, so your chances of getting a gem are higher. Two players who each fill part of two rows, so that both rows get carts will each have twice the chance of getting a gem (each) than if one had filled one row and the other filled the second row, without mixing.

Note that "difficulty" here is the player's difficulty level, not the teamLevel.

Now, the probability of getting Gold = teamGems * teamLevel * 0.0625. teamGems is the total number of gems used in carts, and if a gem is crossed (both row and col) it counts as two gems.

Chance of getting gold specifically is:
teamLevel  teamGems  prob of Gold
0 any 0
1 1 0.0625
1 2 0.125
1 3 0.1875
1 4 0.25

2 1 0.125
2 2 0.25
2 3 0.375
2 4 0.5

3 1 0.1875
3 2 0.375
3 3 0.5625
3 4 0.75

4 1 0.25
4 2 0.5
4 3 0.75
4 4 1.0 <----

5 1 0.3125
5 2 0.625
5 3 0.9375 <----


I don't like the idea that a team could save up enough gems to get a probability of 1, but if it is implemented in YPP, the 0.0625 would be extremely small, so this wouldn't be a problem. Keep in mind that if you booch on the higher levels you lose all your pieces and gems.

If instead of straight multiplying by the number of gems I made a loop and ran the loop teamGems times, each time having a probability of teamLevel * 0.0625, that would exactly the same as if the team always used any gem as soon as it was found. I'd rather encourage them to save up their gems.

It's too early in the morning for me to trust my grasp of the probabilities, but I do think that the above formula needs work other than just making the 0.0625 smaller. OTOH, making the gold ore next to impossible would make Nibbler (on GameGardens) less interesting.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 17, 2007 5:17:00 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Ok, in my unoffical version (desktop, not on GameGardens) I have the stats table showing the number of cheap fruits and rare fruits (divvied among team members). I *think* I have the bugs from watchers joining tables and leaving them fixed. We will still have a problem in a two human table if one player locks before the second one disconnects or just closes his client. (That is, when the first player locks, there *is* a valid active second player, but then that second player quits in a way that is now only checked at the beginning of a new round.)

The next step is to set up some kind of timer on the server, started by the first player locking the board. Once a minute it would check to see if more than 50% of the *online* human players have locked. This would fix that last scenario. The human who is left would just have to wait 60 seconds (or whatever I set the timer for) after the other player disconnects or is automatically logged off.

Now, while I have the code on the operating table, what else needs changing (other than the images - I'm not touching those today).

BTW, in my testing today offline, I've noticed that playing at level 2 is a lot of fun without the brain sweat. Ok, not going for gold, getting hardly any gems at all. Have lots of hints and moves.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 17, 2007 12:22:41 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

I'm uploading the new JAR file. If PlayerA locks, then PlayerB just closes the client instead of going back to the lobby, in less than a minute the server will notice and not count PlayerB in the voting for the next round. (whether PlayerB was locked or not)

I can't test disconnection offline, but I'm hoping that will work as well now. We still will have a problem if a watcher wants to take a seat at a table that is full, and then *all* the players just close clients instead of going back to the lobby. In that case, the board will just sit there with the watcher not getting a seat at all. IMO this isn't something we need to fix, but if you feel otherwise I can add a check every 30 seconds for this situation too.

Nibbler Stats, before they get wiped:
player         gold    level  rounds
SoloNibbler 5.0 4.4 39
ThreeBots 3.5 4.0 36
tcarr 1.0 3.9 39
TheCosmicChz 1.0 3.8 39
isharsh 1.0 3.7 11
Dcurd 0.0 2.5 65
PogoBeta 0.0 1.0 2

----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 17, 2007 2:40:47 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

ThreeBots played 10 rounds with three bots (2 master, 1 respected):
ThreeBots 1.0 4.8 2.5 3.9 10

that's 1 gold, 4.75 cheap, 2.5 rare, average level 3.9, 10 rounds. If we assume that the 1 gold would have just been another fruit, my 10 labor hours would have gotten either:
5 cheap and 2.5 rare
or
4.75 cheap and 2.75 rare.

At 15 poe for cheap and 50 poe for rare, that would be about 200-209 poe for my 10 labor hours.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 17, 2007 3:21:41 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

SoloNibbler played 20 rounds:
SoloNibbler 1.0 11.0 10.0 4.1 20

One gold (about round 15 methinks), 11 cheap fruit, 10 rare fruit, average level 4.1 (I booched once early in the game).

I noticed that two other players tried Nibbler, and never got any fruit at all. One played 7 rounds, the other 3. This disturbs me. Is the game difficult to learn? What could I do to make it easier for beginners?
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 17, 2007 5:30:19 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Another thought.... solo play on level 5 is actually rather boring most of the time. You only get enough pieces for one row, and so as soon as you have a row determined, you take it. Eventually you either drop down to level 4, or you get enough gems that you could think about trying to cross a row and a column.

We don't want to prevent people from playing Nibbler solo. However, if gold could only be foraged with at least two gems in the team carts, that would give Legendary and Ult players a large incentive to work with at least one partner. Teams would have to talk more too; if I have the only gem on the team, we have to make sure nobody fills the row or column with the gem unless we can fill both the row and the column, so that we can save it until we have two gems.

I like that idea. /e makes a note for the next version.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 17, 2007 5:39:30 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

It appears that a lot of players are finding Nibbler extremely difficult. Other than my own alts, I only noticed one player who has foraged gold who wasn't playing with me at the time.

Instead of needing multiple gems, perhaps we should just make the probability of gold based on the total team score. This would make it difficult for a solo ult to get gold (but still possible) and encourage a group of 4-6 middling type players to make a team.

Feedback?

I'm also going to add another tab to the panel on the right, that will display the number of team members, number of rounds finished in this session, the current team level, total gold in this session, total cheap fruit in this session, and total rare fruit in this session.

If I can squeeze in time to update the graphics this weekend, I'll try to redo the "quick start how to play" image then use it in the game, showing it at the beginning of the session to anybody who has either not played before or has an average difficulty level less than 2.

This weekend will be my last chance for GCPP coding for about three weeks. /e kicks RL: midterms to grade, poster to present at a conference, etc.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 23, 2007 6:03:17 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Here's another idea I'll try to add to the next version of Nibbler, to help those who find learning Nibbler too difficult.

Add extra levels, even easier than the current easiest level. Each of the new levels would have the same number of hints and moves as the current easiest level, but the board would not lock if you booch. Let's call the hardest of the new levels A, next hardest B, etc.

A: when next round starts, the booches are still there.
B: when next round starts, the booches have been healed (at no cost to hint tokens)
C: each round starts with one random fruit already placed on the board, and any booches healed.
D: each round starts with two random fruits already placed on the board, and any booches healed.
etc.

Scoring formulas are based on the difficulty level, so these new very easy levels would get very few points. The lobby would not list most of them at all, but if you are on the lowest level and don't get a single full cart, then the next round you would drop down a level. Eventually you should reach a level where you can fill carts. I'm currently thinking "B" should be the lowest setting in the lobby.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 23, 2007 3:45:07 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

apparently there is still a problem sometimes with watchers stumbling into tables, leaving, then the table being stuck. grrrrrr
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 24, 2007 4:21:44 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Official version 4.8 is now available on GameGardens.

Changes:
1. new difficulty levels added, both easier and insanely difficult. If you can't help make carts even on level 1, you drop down to level 0, then level -1, then level -2, etc. Each level is even easier. Eventually you reach a point where just sitting there will fill carts for you.

2. new scoring formulas. I'm hoping that these are balanced well enough. The points are used to help determine whether you spawn gems, and when to change levels.

3. new gold spawning formulas. It's still too easy to get gold if you have a team of good players on a high level, but even then you have to get a gem first. Even a single gem could get you gold, on any level. The more gems and the more team carts, the greater a chance you have of gold. No gem at all = no gold at all.

4. Names of fruit foraged will now appear in a color that is determined by the type of fruit. Bananas and mangos are yellow, rambutan is red. Eventually I'll take the time to find a brown color for the coconuts and a tan for the durians, etc.

5. Another tab is added to the side panel, labeled "Session". This gives totals of gold, cheap fruit, rare fruit and rounds for the session (so far), and the current team difficulty level (which is the average of the individual levels).

6. Statistics (not exactly top scores) are now sorted by gold and then average level.

Thanks to those who have helped test!
isharsh 23.2 198.5 132.7 3.6 391
SoloNibbler 13.0 63.0 66.0 3.7 122
ThreeBots 1.0 6.3 3.8 3.8 13
bsmash 1.0 55.0 18.0 2.2 69
Lanatir 0.5 2.5 4.0 2.4 7
akwellins 0.0 0.0 0.0 0.9 7
pokerace131 0.0 0.0 0.0 0.7 3
kalgarthrax 0.0 0.0 0.0 0.4 7
Kewene 0.0 2.0 0.0 2.0 2
joanna00 0.0 0.0 0.0 0.1 21
Kenku178 0.0 15.0 6.0 1.4 28
Dcurd 0.0 8.5 6.0 2.9 18
LadyQickDraw 0.0 0.0 0.0 1.2 5
mcsolo 0.0 19.3 2.5 0.6 48
mcparrot 0.0 19.3 2.5 0.9 48
hazarath 0.0 0.0 0.0 0.4 8
jermie8 0.0 1.0 0.0 0.0 1
1271351 0.0 0.3 0.0 1.5 2
jujuman 0.0 0.0 0.0 0.6 5
Aeffi 0.0 0.0 0.0 0.2 24
Vorky 0.0 2.0 0.0 0.4 8
shiberdoo 0.0 0.0 0.0 0.4 15
RaoulA 0.0 0.0 0.0 1.0 1
rundunder 0.0 0.0 0.0 1.0 1
Macaroonny 0.0 0.0 0.0 2.0 1
cptjulie 0.0 0.0 0.3 3.0 1
jefadrian 0.0 0.0 0.0 1.0 1
kevin5592 0.0 0.0 0.0 0.8 4
dmfles 0.0 0.0 0.0 1.5 2
captblack3 0.0 0.0 0.0 0.6 5
amylo 0.0 20.3 6.3 2.2 31
Kimi_ss 0.0 4.5 0.0 0.5 8
Arnstkj 0.0 0.0 0.0 0.5 6
lassie_laugh 0.0 3.0 0.0 0.8 5
knarkie 0.0 0.0 0.0 0.6 5
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Mar 1, 2007 3:49:02 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

IMO the auto level adjusting is working rather well. Gold spawns way way too easily if you are playing with other reasonable skilled opponents (even Respected/Master bots) and know anything about strategy. After you reach the point where you spawn gold, it appears easier to spawn a second one.

I'm considering having spawning of gold wipe everybody's boards. If you get gold, you get no other fruits for that round, and you lose any gems, fruits, saved hints, and saved moves - basically putting you back to where you were when you first start a game, although each player is on his/her own level. Or perhaps just doing a "game over" at that point.

It's nice to see players who aren't my alts getting gold though, and so far at least we don't have any players who aren't getting any fruits at all.
player gold cheap rare level rounds
chthon 3.0 53.8 22.3 4.7 57
ThreeBots 2.0 3.3 4.0 9.5 14
isharsh 1.7 5.7 3.0 8.2 13
tcarr 1.0 2.0 2.0 8.2 6
SoloNibbler 1.0 3.0 4.0 7.3 6
Calsan 0.0 16.3 3.5 2.5 33

Feedback would be appreciated.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Mar 2, 2007 6:39:36 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
Kewene

Member's Avatar


Joined: Jul 7, 2005
Posts: 483
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

I just needed to post this. >.<
No hints or photoshop! I finally could solve a whole board. :)



(Full window)

Too bad I only had 11 moves. <.<
There should be a bonus for this. :P

Well, I just wanted to show off *grins*

~Kew
----------------------------------------
Guiido on Jade, Keweney on Sage and Viridian!
[Mar 17, 2007 11:49:45 AM] Show Printable Version of Post        Send Private Message    http://forums.puzzlepirates.com/community/mvnforum/index [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

congrats!
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Mar 17, 2007 4:10:58 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tanonev

Member's Avatar


Joined: Nov 18, 2006
Posts: 2361
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Sorry if you've already addressed this issue, but I threw the following together:

http://www.stanford.edu/~tanonev/Piranha.java



Feed the following into Piranha (put it into a text file and include the file name on the command line):


DPR

order PPRR D
pair RPD
mix RPD
pair RPD
adjacent RD P
order RPPR D
order DDRR P
pair RPD
order PDDP R
adjacent PD R
mix PDR
order DPPD R

------
------
------
------
------
------


and get the following analysis out:


Possible boards: 4

PPRRDD
DDPRRP
DPRPDR
RRDDPP
PRDPRD
RDPDPR

PDPRRD
DDRRPP
DPRPDR
RRDDPP
PRDPRD
RPPDDR

DPPRRD
DDRRPP
PDRPDR
RRDDPP
PRDPRD
RPPDDR

DDPPRR
PRRDPD
PRDRDP
DPDPRR
RDRDPP
RPPRDD

------
------
----D-
--D---
------
R-P---

Best hint location (0-indexed): row 0, column 0
D: 2
P: 2
R: 0


The fifth board is what the 4 possible boards have in common.

Rerun Piranha with, say, a P at (0, 0) and we get


Possible boards: 2

PPRRDD
DDPRRP
DPRPDR
RRDDPP
PRDPRD
RDPDPR

PDPRRD
DDRRPP
DPRPDR
RRDDPP
PRDPRD
RPPDDR

P--R-D
DD-R-P
DPRPDR
RRDDPP
PRDPRD
R-PD-R

Best hint location (0-indexed): row 0, column 1
D: 1
P: 1
R: 0


So this board is completely solveable using 2 hints.

I've run Piranha on 6 other empty boards (not much, but I'm lazy), and out of those 6, 4 were completely solveable, 1 had 30 solveable squares, and 1 had few solveable squares but had a single hint location that would solve the entire board.

Piranha runs on the (perhaps erroneous, but so far accurate) assumption that a fruit basket row/column can only be 1 of 30 (instead of all 90) possible permutations. If you make it so that a fruit basket row could be anything at all, then Piranha's power would be greatly reduced.
----------------------------------------
Tanonev on all oceans; currently exploring Meridian.
Puppetar by Tilinka
[Mar 22, 2007 11:45:42 PM] Show Printable Version of Post        Send Private Message    http://www.alpha-slash.com [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

The strategy on Nibbler isn't about solving the boards from the clues. The higher levels get fewer (or even zero) clues. The strategy comes from deciding which rows/cols to fill, which of your moves to take even if they don't fill a row/col, and how many moves to save for next round.

I usually start on level 8. That gives me a single hint and only 7 moves. Sometimes the clues fully determine the board, sometimes they determine one or two rows/cols, sometimes I can't determine anything without using my single hint. Do I use the hint, or save it? Strategy. Do I use the gem on the board now, or wait until I can cross it (both row and col so the gem counts twice) or until there are two or three gems on the board?

There's a lot more strategy in multiplayer Nibbler games as well. Another player tells me he can solve row A and column 4, but doesn't have enough moves to finish them. If I can solve something in row A or col 4, then perhaps we will get a better score if I use one of my moves to help the other player finish.

If players say that Nibbler is too easy even at the higher levels (possibly because they are using a program like Piranha) we can always use even fewer clues. The current version of Nibbler added a lot of levels even easier than my original easiest level (which I considered ridiculously easy) because I was seeing too many players on the high scores lists playing multiple rounds but never finishing a single row or column.

Does this make sense? And please, anybody who likes logic puzzles, let me know if we need to make even harder levels than we have now. Making harder levels is easy; we can either start taking away clues (not every row/col has a clue) or else not use the easiest clues (palindrome, ABAB). I already have the "solve the board" code written (from Parfait and Feasting) so it would not be much work to make sure at least one hint is needed to solve a board on the higher levels (which get no hints, and so people have to guess, but guess wrong and the entire board clears on those high levels; conservative players would just say "i'll pass on this board, and have more moves next board").
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Mar 23, 2007 4:59:14 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

 
The strategy on Nibbler isn't about solving the boards from the clues and hints . The higher levels get fewer (or even zero) clues hints . The strategy comes from deciding which rows/cols to fill, which of your moves to take even if they don't fill a row/col, and how many moves to save for next round.


grrrr.... one of these days I'm going to learn to reread the post three times, not just once, particularly before breakfast. I got clues/hints confused here. Currently, all difficultly levels of Nibbler get one clue for each row/col, but the higher levels get fewer and fewer hints (crackers).
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Mar 23, 2007 5:02:13 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tanonev

Member's Avatar


Joined: Nov 18, 2006
Posts: 2361
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

This makes sense, and I agree that Piranha doesn't actually break Nibbler; however, it seems to me that this exposes a sort of disconnect in the puzzle. There are two parts to the puzzle: the logic part (the clues and solving the board) and the strategy (risk/reward and cooperation) part. In other games, the mechanics and the strategy of the game are intertwined (YPP example: bilge's mechanic is to make clears, its strategy is to make simultaneous clears in limited numbers of switches; GCPP example: Knightfish's mechanic is to move from one square to the next, its strategy is to make moves in such an order as to maximize the length of the game). Here, though, the entire basis of strategy has (IMHO) nothing to do with the basic premise of the puzzle. You handle all of the logic stuff all at once in the first step, and then all of the strategy and planning in the second step.

Or, to look at it from a different angle, there is (at least right now) no logic element that a player can use that Piranha can't. A person using Piranha differs from a person not using Piranha solely in that the former saves a lot of time analyzing the board and will never make a mistake by making a move that can't possibly be correct. As a result, Piranha places other players at a distinct disadvantage, not in that the other players can't do what Piranha can, but in that Piranha never hurts the user (unless the user makes a typo and is misinformed about the board as a result).

The other problem that may come up is that Piranha makes it that much easier to multiclient the puzzle. As a logic puzzle, Nibbler's defense against multiclienting was the sheer amount of information that needed to be kept straight among the boards. Piranha takes care of all of that, and could even be modified to conveniently combine the boards from each player to tell you which square is solveable by which player(s).

Don't get me wrong--I love Nibbler and the idea of multiplayer cooperative puzzling in general. I just wanted to point out (and I know that this is a tall order) that while Nibbler is a logic puzzle and a team strategy game, it would be nice if it could be both at the same time.
----------------------------------------
Tanonev on all oceans; currently exploring Meridian.
Puppetar by Tilinka
[Mar 23, 2007 10:32:58 AM] Show Printable Version of Post        Send Private Message    http://www.alpha-slash.com [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

I understand the point you are trying to make. Most people know that I lubs me logic puzzles. The very nature of logic puzzles means that a machine could help solve them. This makes them tricksy for building a workable YPP puzzle around them.

What I decided was to not care about machines solving the puzzles at the lower levels. I assume that at the higher difficulty levels, everybody can solve the board as much as it *can* be solved, without making mistakes. So the key to high scores on the higher levels is cooperation and strategy.

Removing the restriction of exactly two of each item in each row and each column would not fix the machine solvability issue. Any logic puzzle would have a similar problem. This might mean we will be forced to give up on logic puzzles. I hoped that the strategy part needed after the board is solved as much as possible would make the game interesting enough. Is bilge made less interesting if you write a program that can spot the largest one move combo on the board for you? Are SF tournaments made less interesting if you have access to a program that displays your opponent's drop pattern for you at the beginning, instead of needing to see the first few drops?

A person who multiclients Nibbler would still need to use a good strategy on the higher levels, but aye twould be a lot less of a headache to multiclient if you had access to a cheat program.

Do you have suggestions that would make the strategy clearer from the beginning? I don't want it to devolve into a Sudoku clone (or even Inspector Parker).
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Mar 23, 2007 11:04:56 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tanonev

Member's Avatar


Joined: Nov 18, 2006
Posts: 2361
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

I believe removing the restriction of exactly two of each item in each row and each column increases the search space about 300000 times, which would at least make the machine solver a lot slower. Unfortunately, it might also affect human logic for solving the puzzle adversely.

What if instead of making hints reveal tiles, have hints reveal clues? And/or perhaps if you clear a row or column, the clue for that row or column is revealed for the other players? This way the logic step is spread throughout the duration of the puzzle, not all bunched up at the beginning. Also, strategy would be able to affect the logic puzzle itself.
----------------------------------------
Tanonev on all oceans; currently exploring Meridian.
Puppetar by Tilinka
[Mar 23, 2007 11:32:27 AM] Show Printable Version of Post        Send Private Message    http://www.alpha-slash.com [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

I'll think about these suggestions. I was trying to avoid the situation that would encourage people to multiclient; having your alt clear a row or column to make it easier for you to solve the board sounds like it would beg people to cheat.

The way Nibbler works currently, each player has a different board. This was so we wouldn't have one great player solving the board then just telling everybody else the solution.

Searching a solution space even for Sudoku (nine different items, exactly one in each row, each column, each box) won't take a significant amount of time on a client. Doing it on a server that is juggling many different players would be a problem. Also, you don't have to generate the full solution space then test each solution against the clues. I wrote the AI for Parfait so that I could generate sets of clues that would fully determine a board without having extra clues. Then I modified that AI for Feasting, where the bot opponent has very few clues, and the two opponents are working on the same board so placing a fruit gives the opponent another clue (A3 = banana). The AI I wrote for Parfait and Feasting solves the board as much as possible with the given clueset. This doesn't give a set of possible boards but rather for each location what are the possible fruits that could go there. It would be straightforward to write something similar for Sudoku or other logic puzzle.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Mar 23, 2007 12:23:18 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tanonev

Member's Avatar


Joined: Nov 18, 2006
Posts: 2361
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

When I suggested the row clear hint, I intended to work in the direction of better players helping out the lesser ones so that the scores "even out." I imagined a Leg/Ult player doing lots of single clears in order to provide Able players more clues so that they could clear stuff on their own.

To prevent alt abuse, perhaps the row clear hints only occur if your current score is lower than that of the person who cleared the row?
----------------------------------------
Tanonev on all oceans; currently exploring Meridian.
Puppetar by Tilinka
[Mar 23, 2007 1:13:41 PM] Show Printable Version of Post        Send Private Message    http://www.alpha-slash.com [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

There's already a way for Leg/Ult players to help those who aren't as good. The lower level player can use two hints on a palindrome or ABAB clue row/col to solve any parts of that row/col that aren't already covered, then any remaining hints to try to finish a row/col that somebody else started. The GM+ player tries to solve his/her board completely before placing any fruits, then makes moves that complete other player's rows/cols.

If you help complete three rows/cols (even if you just place three fruits) then you either stay on your own difficulty level or rise to the next hardest one.

Another strategy is for mid-level players to each try for one full row, and an Able player use his/her hints as directed by the other players finish the best possible set of row/cols.

Hmm.... how about any time you score well enough to rise to a higher level, it gives you a special token. That token can be played at any point later in the game, to give a different player (at a lower difficulty level) something that would help him solve his board (perhaps a clue of the form A3 is not a banana, when A3 is currently empty on everybody's boards). Perhaps a differently shaped cracker, so the player can choose which position gets it? Or perhaps the player who generated the token places it on the board and all players except himself get a clue of the form "A3 is not a banana"? I think I like that last possibility.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Mar 23, 2007 1:32:44 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

I'm still considering adding the generation of tokens that will give a "nary" clue to every player for the position the token is played on. Not sure about that.

I'm starting to think that it might be better to have any gems on the board vanish, so that you can't hog them. This would make gems rarer, and since that's the only way to get gold, gold will be rarer. A team would get very excited when one of the players has a gem, and work very hard to try to cover the row and column for that gem, to have a chance at the elusive gold. Does this sound more interesting to you?
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Apr 5, 2007 1:25:18 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
Patetch

Member's Avatar


Joined: Apr 22, 2006
Posts: 3340
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Yes, added risk does make the game more interesting, and choices have to be made.
I did want to note, you may have seen this already, but apparently there are plans to release gems on all the oceans, but the types and variety will probably be different for each ocean (since most do not have archs named after gems.)
----------------------------------------
Pletoo of Sage
[Apr 5, 2007 1:41:15 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

 
Yes, added risk does make the game more interesting, and choices have to be made.
I did want to note, you may have seen this already, but apparently there are plans to release gems on all the oceans, but the types and variety will probably be different for each ocean (since most do not have archs named after gems.)


I had noticed that, yes. It would be pretty easy if Nibbler were chosen as a foraging puzzle to have an input parameter for which arch/ocean you are in, that would decide which cheap fruit is foraged, which expensive fruit is foraged, and the set of possible gems that could be foraged (anything for THIS ocean, except this arch's gem). I can add a chance of a gem being spawned if gold isn't as well, for the GameGardens version. Since I won't have much (if any) coding time until about May 8th or so, I'll have lots of time to think and discuss changes.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Apr 5, 2007 2:19:41 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
tcarr

Member's Avatar


Joined: Jun 7, 2004
Posts: 4294
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Videos are now available of me playing Nibbler, with bots, on level 8. If I can arrange to have a human or two play with me, that would be an even better video methinks. YouTube quality leaves something to be desired, but it should give players an idea of some of the features/strategies.

Level 8 with 3 bots, first two rounds

Round 3 (gold ore!)
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[May 2, 2007 5:04:22 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
Stefania_33

Member's Avatar


Joined: Sep 4, 2006
Posts: 604
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Ok, I have decided to put on here some of what we already talked and some new questions.


1. I still think that it would be more challanging if the black holes could not be repaired, and maintained till the board locks ( i.e not being able to complete any line due to the holes). Because at this current stage where you can repair them, it makes it very easy to get fast to a high level, and from there gold is very easy to be obtain. At least in my team games played with the bots(once I got the hold of the game), I had one gold after another. And I started from level 8, 9 and 10. So I think pirates should not be able to get to those levels very fast and also the bingo and dokeys wouldnt be so easy this way. And the permanent holes will stop this easy access to the gold forageable . I know that starting all over is not motivating, but also patching the hole is not challanging either.

2. If first proposal is not accepted maybe it should be limited as the patching crackers are. One per 5 levels or any other condition you can think of.

3 There is a question that keeps coming when I play the game. Without any bad temptation sometimes when I play and I see that the other bots have some fruits placed already on the board, I can not resist to say I wish I knew what those fruits are. The question is: if I play in a team with other 3 pirates ( not bots) do they have the same table as I have in terms of same fruit placement?
Because in this situation, when they have some fruits placed on board ( the plus sign or the already carted sign) I could just easily ask them what those fruits are, and help myself with the information without using any cracker, creating myself a not fair advantage.

4. And the last and not so important, if possible, when I press set the initials, those to appear even on the blocks with the gems, since that info still plays a role in the judgement made on the board.

That is all for now. I really think this is a very nice game, especially from the cooperative point of view. Initially I started to play alone, but after a while I started to play with bots and I now I only play it in team.
----------------------------------------
Shooshy
(Sage) Emerald Ocean

"Cei ce stapanesc pe altii sunt puternici. Cei ce se stapanesc pe ei insisi au adevarata putere" Lao Tzu
[May 10, 2007 8:47:00 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
Stefania_33

Member's Avatar


Joined: Sep 4, 2006
Posts: 604
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

Also another suggestion.

The letters are very difficult to distinguish when they all have the same color ( ie. brown). I know that letter have the color of the fruits they represent, but I think a very distinctive color would make it a bit easier for the eyes. My eyes gets teary when I play a game with all 3 or 2 brown, and most of the time I quit the game looking for distinctive colors.
----------------------------------------
Shooshy
(Sage) Emerald Ocean

"Cei ce stapanesc pe altii sunt puternici. Cei ce se stapanesc pe ei insisi au adevarata putere" Lao Tzu
[May 10, 2007 8:54:25 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
Posts: 170   Pages: 6   [ First Page | Previous Page | 1 2 3 4 5 6 | Next Page | Last Page]
[Show Printable Version of Thread] [Post new Thread]

Puzzle Pirates™ © 2001-2020 Grey Havens, LLC All Rights Reserved.   Terms · Privacy · Affiliates