• 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 34540 times and has 169 replies Next Thread
Piplicus_BNO

Member's Avatar


Joined: Aug 17, 2004
Posts: 7311
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote




----------------------------------------
 
What Pip tried to do was something totally unprecedent. He, in some strange way, used Midnight as a test server, without him even knowing. [Crap, it sucks I was not online when he did that :'(... pfft, anyways]


Like a bad penny.
[Feb 1, 2007 12:00:09 PM] Show Printable Version of Post        Send Private Message    http://www.puzzlepirates.com    Intermission #3 - click here! [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

Oooo purty! Now, why did I miss seeing that post when it was made???? /e sighs

I'm not going to change the artwork today. I have the scoring changes made (so that solo full carts don't give you more points than a team full cart - this will encourage more team cooperation) and am about to upload the new JAR file.

Here's the saved Top Scores list (using the old version of the scoring):
244	isharsh	4
226 Dcurd 5
202 isharsh 5
162 isharsh 5
127 Kewene 5
110 meekrat 3
108 Kewene 3
107 isharsh 2
105 isharsh 3
100 tcarr 4
97 Kewene 2
96 isharsh 5
95 meekrat 0
95 isharsh 4
94 meekrat 2
92 Malach 1
92 meekrat 1
91 tcarr 5
91 Dcurd 4
90 Malach 3
90 isharsh 3
87 Dcurd 3
86 tcarr 5
79 mcsolo 1
78 isharsh 4
76 isharsh 5
76 isharsh 5
76 isharsh 5
76 isharsh 5
74 amylo 2
74 Kewene 2
74 amylo 2
74 Kewene 4
74 Malach 2
74 isharsh 2
74 meekrat 2
74 isharsh 3
73 isharsh 4
72 tcarr 5
72 Kewene 4

----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 4, 2007 6:06:11 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

Now that the computer room isn't freezing, I can play with the graphics program a bit - here is a 32 x 32 transparent background cracker that I just made:

and because I am having fun... what size is the cart? Is that also 32 x 32?

Ooh - I just realized that Pip already made a cracker - I thought it was a gold nugget when I saw it before (since the discussion at the time was how Midnight is the only ocean with gems) - I just now saw the name, sorry Pip.
----------------------------------------
Pletoo of Sage
[Feb 5, 2007 7:16:48 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 cart is sitting on a 48x48 tile, but has room around it. It could be 32x32 or 48x48, either one - but the current code is set for it being 48x48.

I'm currently thinking of ways to indicate what kind of fruit was foraged by the team. Each round, each team member contributes one labor hour. Each team member could be playing at a different level. I'm tracking the points from each player's divvy of the team points (there are no longer any individual points earned).

In testing with bot team members, one time I had a horrible board and no hints left, and the team didn't make any carts at all. Each of us dropped one level, and I had 15 moves to make (quite a few left from the previous round), and could solve the board. I placed a fruit in each spot that nobody else had, and we filled the board: Vegas^7 if I remember correctly.

If the team gets one foraged item for each team cart, then that Vegas^7 would net 12 items, even with only 4 team members. Two team members alternately not placing carts (but placing any fruits they can determine) with attempting to fill the board might manage to get that Vegas^7 every other time. That would be 12 items for 4 labor hours - but at a lower difficulty level than the players are capable of playing.

Perhaps there should be a limit to the number of items foraged per round. Six sounds reasonable. This would let foragers have teams of six and still get six fruits each round just about all of the time, even if they played with a lot of greenies on the team. The team of 2 very good puzzlers who used the alternating carts strategy would get an average of 3 fruits each round, which seems reasonable for two very good puzzlers. A good solo player would average between 1 and 1.5 fruits a round, with the occasional "Nary" round.

I'm going to use limes and passion fruit here, but any pairing of "cheap fruit" and "expensive fruit" would work. The team difficulty level (average of the individual difficulty levels) would determine the spawning probabilities.

Difficulty level ---- fruits
0 - just limes
1 - 75% limes, 25% passion fruit
2 - 67% limes, 33% passion fruit
3+ - 50% limes, 50% passion fruit

Each gem used in a round gives a chance of one of the fruits turning into gold. This means that poor players won't have a chance at gold unless they are in a team with a good player (who will be generating gems). Good players will have a chance at gold if they play solo, but a better one if they are in a team with other good players and perhaps one person who is on lower difficulty (so that he has hints to be used to finish rows and columns for more carts).

The current likelihood of gold is very small. We want that to be small still..... Let's say that if a gem is used in the round, the probability of one of the fruits foraged being turned into gold ore is (team difficulty level) * 1%. This formula can be easily changed or made a straight percentage, not based on difficulty level. If foragers know that the alternating no carts round trick drastically decreases their chances of getting gold, they will be much less likely to use it.

Feedback?
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 5, 2007 7:58:10 AM] 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

To clarify:

If 4 players have an average difficulty of 2, does that mean:

The team has a 67% chance of getting 4 limes and a 33% chance of getting 4 passion fruit,

or each player has the chance so it would be possible the team would end up with 3 limes and 1 passion fruit?
----------------------------------------
Pletoo of Sage
[Feb 5, 2007 9:06:35 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

If the team forages 4 items, then the "dice" (Math.random()) would be rolled 4 times - so it might be 4 limes, it might be 3 limes and a passion fruit, it might be 2 limes and 2 passion fruit, even if the team level is two in all cases.

Most likely I'll determine if gold spawns first, then go from there. Maximum of one unit of gold ore spawning per round, no matter how big the team is.

This is certainly the type of thing that Three Rings will likely modify if they decide they want Nibbler implemented as a foraging puzzle. I'm just trying to make Nibbler a bit more interesting, getting players' attention away from the points and focusing on having fun and getting fruit.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 5, 2007 9:58:16 AM] 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

Ok, I have been having a lot of fun with the graphics programs. Since Pip already made the current images transparent, I decided to experiment with the fruit and other things. My design is based on the premise of left clicking places the fruit and right clicking cycles through the options, eliminating the fruit boxes below the board.

I made the fruit in different squares (48, 36, and 32) but 16 was ridiculously small and hard to see for some of them. However, 24 was still pretty visible for single fruit.

I then flipped or otherwise edited the fruit to make them more distinct. The carambola was still getting too small, so I just used the star part for the 24 square. The lime and rambutan were both too small as well, so I made my own for the 24 square.

While making the pieces transparent, I also ended up with silouettes. The thought came to me that they could be used instead of the nary sign (although I still prefer the pieces to disappear altogether when they are eliminated.)

I also made a couple of carts, and a different picture for the reflection clue and I used the monkey instead of the fruit bowl.

Here are some examples: (all the white areas are transparent.)


While running the color check, the single lime and mango ended up the most similar. The passion fruit turns blue and may be similar to the gem, but the shapes are different. I didn't check to see if the fruits are distinguishable from their silouettes, so a different shade may need to be used there.

Here is a sample of a board:

You can see I divided the board tiles into fourths, using three for the fruit and one for the location. I used the 24 squares in place of the initials and the 48 square for a placed fruit. I had the gem replace the location so that the fruit would still be available to help solve the other squares.

I also used the 24 and 36 squares for the clue sets (32 could be used instead of 36.)

I hope the clues are pretty clear, but just in case:
Column 1: Full cart
Column 2: Both coconuts are further up than both the carambolas.
Column 3: Monkey = fruit bowl
Row A: Reflection = Mirror
Row B: Each pomegranate is to the right of a coconut (CP:CP)
Row C: Both the pomegranates are between the carambolas (C:P:P:C)

Here is a link to all the different size pieces I made:
http://img174.imageshack.us/img174/3961/tilesetdd3.png
I have everything saved as individual tiles, both with backgrounds and transparent, so if there is anything you would like to use, let me know and I will get them to you (so you don't have to cut them apart yourself.)

Depending on the game background, the clue pieces can either be transparent (as I have them) with a solid color connecting two adjacent pieces, or if the background is textured then the clues would be on a solid color. The background effectively replaces the need to have a separate divider piece (if a piece is needed for placement purposes, just use a small strip colored the same as the game background.)

I don't know if any of this is useful or not but it sure was fun to do.
----------------------------------------
Pletoo of Sage
[Feb 6, 2007 8:52:57 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've been testing the "show forage items" tweak. Playing as two humans with no bots, starting on level 3 but ending up on level 5 and finally having to stop the marathon session due to running out of hints, what we got was:

11 rounds total, two players, so 22 labor hours used.

I booched twice, so this could have been two or three fruits more.

Total cheap fruit: 15
Total nicer fruit: 15

So, total 30 fruit (50% nice stuff) from 22 labor hours. That was two very good human players (even with the two booches). When I play with a team of bots I don't do nearly as well.

passion fruits: 1
coconuts: 2

carambolas: 1
coconuts: 1

rambutans: 1
pineapples: 1

rambutans: 2
mangos: 1

passion fruits: 2
pineapples: 2

pomegranates: 1
mangos: 1

pomegranates: 2
pineapples: 1

pomegranates: 1
bananas: 4

pomegranates: 1
Booched!

bananas: 2
Booched!

pomegranates: 3

----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 6, 2007 8:54:38 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

Thankee Pleetoo, I'll think about which version I like best (or mebbe we can have a lobby parameter to choose).

The "show what got foraged" version of Nibbler is not yet uploaded. I'm still testing.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 6, 2007 8:59:55 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

Whoops - I forgot to add that if it is too confusing using the larger fruit pictures in the clues, (because they are different) then the 24 squares could be used there just like the intials are currently.
----------------------------------------
Pletoo of Sage
[Feb 6, 2007 9:07:02 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

 
Whoops - I forgot to add that if it is too confusing using the larger fruit pictures in the clues, (because they are different) then the 24 squares could be used there just like the intials are currently.


I imagine that players could get used to either way of doing the clues, but anybody who is already used to Feasting or Nibbler will probably prefer to have the AABB type clues use two small A fruits and two small B fruits.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 7, 2007 4:10:38 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

Implementing Pleetoo's version of the user interface is going to have to wait at least a few weeks. That's the kind of thing that will consume most of an uninterrupted weekend, between the coding and the testing and debugging, and currently my GCPP coding time is in small chunks.

I'm hoping to get Piplicus' new images added this weekend, but can't make promises. I've been chasing bugs in the parts dealing with humans leaving, bots taking over, and other humans are watching, and might have the "show what items are foraged" version ready to be posted soon.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 8, 2007 6:10:24 AM] 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

Well, unfortunately, I haven't been able to play a full round in some time - I keep getting disconnected. I am pretty sure the problem is on this end though.

Let me know if you would like any of the individual tiles, if that will save some time for you. I didn't mean to make so much extra work, (but I am assuming that you wouldn't do it if you didn't like it at least a little :) )

I changed the cart and reflection tiles just a little.

----------------------------------------
Pletoo of Sage
[Feb 8, 2007 7:56:44 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 easiest setup is to have one image that is a composite of all the 48x48 transparent images, and another that is a composite of all the smaller transparent images, and a third that is the opaque images. Any image that replaces one of the exact size in one of my current image files should be in the same position (to save me from rewriting the code).

The largest problem with Pleetoo's interface is that the images aren't the same size, and won't go in the same positions as the current ones, plus changing the "what happens when you left/right click" sections of the code. That's a lot more complicated than just swapping image files.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 8, 2007 8:59:17 AM] 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

Ok, I believe I have everything compiled correctly now. I'll just give the links instead of putting all the pictures here.
The 24 x 24 images:
Opaque
Transparent

The 48 x 48 images:
Opaque
Transparent

The initials graphic with the reflection and monkey instead of the mirror and bowl.
----------------------------------------
Pletoo of Sage
[Feb 8, 2007 12:31:09 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

Thankee Pleetoo.

New version (4.0) is now live on GameGardens! ***Please*** let me know if you spot any bugs. This weekend I'll work on the new images; this version uses the old images.

I think that I have all the bugs out of players leaving and joining a table, but haven't tested the waiting list part. The idea with official version 4.0 is that a player can start a table with bots but setting a check to allow watchers to replace the bots. When a watcher joins the table, the watcher takes over from one of the bots. When a player leaves a table, a bot takes over. The table will never have more players than the original table (including bots).

And what gets foraged is announced. I haven't managed to forage gold yet, but it's supposed to be very rare.

There is no way to limit the number of rounds - the game could theoretically last until the GameGardens server is rebooted, with players being replaced. As long as one human is at the table, more rounds happen.

saved top scores list, for anybody who cares:
136	Dcurd	3
112 Respected_2 1
112 isharsh 3
105 isharsh 3
103 Able_1 0
95 isharsh 3
90 isharsh 3
86 isharsh 3
85 isharsh 3
80 isharsh 4
80 Able_1 1
79 Master_0 2
78 Master_0 3
77 Master_3 3
74 Respected_2 2
69 Kewene 2
68 tcarr 2
66 tcarr 3
66 Distinguished_1 1
64 Respected_0 0
63 isharsh 3
63 isharsh 4
62 tcarr 4
61 Master_0 3
59 tcarr 5
59 Respected_2 3
59 isharsh 4
59 Dcurd 3
59 amylo 2
58 isharsh 3
58 isharsh 3
57 Master_0 3
57 Distinguished_1 0
57 Distinguished_2 0
56 Master_0 3
55 Dcurd 2
53 isharsh 4
53 Respected_2 2
53 Kewene 1
52 Master_0 3

----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 9, 2007 5:18:50 AM] 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

I will definitely try to play again, maybe I can stay connected long enough to complete a round. The temperature has gotten above zero (not above freezing but at least above zerot) so the computer room, which faces north, is a little more bearable.
----------------------------------------
Pletoo of Sage
[Feb 9, 2007 7:11:05 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

 
I will definitely try to play again, maybe I can stay connected long enough to complete a round. The temperature has gotten above zero (not above freezing but at least above zerot) so the computer room, which faces north, is a little more bearable.


If there is another human at the table, then it doesn't matter very much if you lose your connection. Make what plays you can before you d/c. A bot takes your place. You log back in, and take over from the bot again. Ok, the bot has used up all your hints in the meantime, but if you have moves left and can determine placements, you can use them.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 9, 2007 7:15:34 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

I still haven't used the new graphics (Pletoo's or Pip's either one). I just uploaded version 4.2 of Nibbler, which has:

1. yet another bugfix (Watcher who sits down, then just closes the client was not being replaced so the game could get stalled with not enough humans locking boards for the next round to start. This should now be fixed with the person who closed the client being replaced when somebody locks a board)

2. Gold ore now has a much greater chance of spawning. This lets us test that it is indeed spawning, and also it's just plain more fun. How many units of gold can *you* spawn in one multi-round session?

3. Text animations now show the type of items foraged. I tried having separate text for each unit foraged, but that made the screen much too busy. If three units of pineapples are foraged, all you will see is one "pineapples".

4. The chat messages about foraged items are removed. I decided that the chat window was so busy with them that they were useless. The items show in the score details tab.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 13, 2007 4:34:47 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

It occurs to me that the top scores list isn't very useful at this point. Perhaps I should change it to just keep track of how many units of gold ore each player has earned.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 13, 2007 6:09:40 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 occurs to me that the top scores list isn't very useful at this point. Perhaps I should change it to just keep track of how many units of gold ore each player has earned.


I just uploaded another version of Nibbler. The "top scores" list should now keep track of how much gold ore each player has foraged (divided by number of players on the team, so if your team of three gets one gold ore it will show as 0.3) and also each player's average difficulty level.

It's still far too easy to get gold - but seeing that large orange "Gold ore!" is fun. Gold is only found when there is a gem in a filled row or column; the more gems used in rows/columns in the same round, the greater the chance of getting gold. We can easily make getting old more difficult by varying one of two constants (the chance of gems showing up, and the base probability of gold spawning).

Please share any feedback (and bug reports) you have. Hopefully this weekend I'll get a chance to fix the images to Piplicus' nice artwork.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 14, 2007 2:37:48 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

One more tweak: the top scores table now displays the number of rounds each player has finished. This will make it a lot easier to tell how hard it is to get gold.

I'm going to make two new accounts on game gardens, one that will only play nibbler solo, and one that will only play nibbler with a team of 2 bots. I'm sure that playing with two or more good human players will make it easier to get gold, but double clienting is awkward (one of the players tends to log off if I don't switch windows very often).
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 15, 2007 5:49:28 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

SoloNibbler took 11 rounds to get gold. Starting with level 3, was on level 5 for several rounds. No luck getting gold with just one gem in a row/col (several times, including level 5). The gold was found on level 5 with two gems in the same row.

There's a completely different feel to solo play of Nibbler. A lot more strategy is involved with playing as a team, even when the others are just bots.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 15, 2007 7:04:46 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

ThreeBots only took 6 rounds to get gold (but only got 1/4 of a share of it of course). Started with level 3, using three bots: Respected, Master, Master. I'm sure it would be harder to get gold with Able bots.

I had been on level 5 for two rounds I think. There were two gems on my board, but I don't know how many on the bots' boards. We managed to fill both the row and column on both gems, on the round we got gold.

I'll keep playing with both ThreeBots and SoloNibbler, to see what the long term averages are. ThreeBots will have to strike gold four times to every once that SoloNibbler does, to come out even.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 15, 2007 10:48:15 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

ThreeBots hit a fourth gold ore on round #11. Strange, but the average difficulty level for both is 4.1 as well.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 15, 2007 11:17:57 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
CraftingProj

Member's Avatar


Joined: Sep 21, 2006
Posts: 221
Status: Offline
Re: Puzzle Codename: Nibbler Reply to this Post
Reply with Quote

How much trouble would it be to record the poe from each round?

Pulling numbers out of a hat:

Cheap fruit 15 poe each.
Expensive fruit 60 poe each.
Gold ore 20K each.
Less a restocking share of 30% to cover shipping to a market in another arch.

Teams would divvy the poe evenly each round. The top score table could show each human player's average poe per round using the current formulas for probability of cheap vs expensive fruit spawning, and gold.

For example, a team of 4 gets one unit of limes, one unit of passion fruit, and one gold ore. That's 20,075 which would split as 6083 for restocking (paying jobbers for the cargo run) and 3498 each for the foragers.

In another round, the same team might get one unit of limes and four of passion fruit, which would make 255 poe; 75 poe for restocking, 45 to each of the foragers (rounding). 45 poe is nice for one hour of labor. A bad round might net them just three limes, for 8 poe each forager and 13 poe for restocking.
[Feb 15, 2007 11:56:19 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

I'll see what I can do. Dcurd and Mcsolo and I just managed to turn up yet another bug in the code that replaces bots with human players and vice versa. Really strange behavior. I'll be trying to reproduce it offline, and then trying to prevent it.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 15, 2007 12:40:36 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've managed to reproduce the bug consistently. Fixing it is going to be a headache, but I have some ideas. In the meantime, if a watcher replaces a bot at a table, have the watcher *wait* to lock until the player(s) already at the table lock.

Statistics from my play testing:

SoloNibbler played alone, always starting at level 3 and continuing for quite a long time (so that most of the time was on level 4 or 5). 39 rounds, average level was 4.4, total of 5 gold ores, so an average of 7.8 rounds per gold. Brain sweat was solving the grid as much as possible, then deciding how to best use any clues and moves, and whether to finish yet another row or leave some fruits on the board to make the next round easier.

ThreeBots played with three bots, always two masters and one respected, always starting at level 3 and continuing for a long time (so most of the time was on level 4). 23 rounds, average level 4.1, total gold was 3 - meaning that the team of four got 12 golds. Average of 7.7 rounds per gold. In addition to the brain sweat that SoloNibbler did, ThreeBots had to also consider helping team mates finish rows/cols so that they stayed at the higher levels (where they spawned gems to increase chances of gold).

Tcarr played with one bot, and sometimes a human would drop in for a bit. Always started at level 3, spent most of the time on levels 4 and 5. Sometimes the bot was Respected, sometimes Master. 39 rounds, average level 3.9 (I booched once, and the above mentioned bug finished a round before I could place my fruits once), 1 gold. A pitiful average for gold compared to the others.

Personally, I think that playing with large teams is more fun. We might need to tweak the scoring of course. Playing solo..... it might be possible for somebody to write a program that would greatly help them stay on Level 5 as long as possible. This concerns me. Cooperative play uses a lot of strategy, and to get really good scores at least some of the players need to be talking to each other and planning who will move where.
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 16, 2007 8:51:26 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
TheCosmicChz

Member's Avatar


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

 
. . . (I booched once, and the above mentioned bug finished a round before I could place my fruits once) ...

Sorry about that one, while its great we found a bug, I'm sorry i'messed up yer game.
 
Personally, I think that playing with large teams is more fun. We might need to tweak the scoring of course. Playing solo..... it might be possible for somebody to write a program that would greatly help them stay on Level 5 as long as possible. This concerns me. Cooperative play uses a lot of strategy, and to get really good scores at least some of the players need to be talking to each other and planning who will move where.

I still don't completely understand the cooperative scoring. From solo play I'm used to "make at least one full cart to stay on this level", but co-op my full carts are sometimes worth a lot less and I get demoted =(

Your new high-score method has me tempted to create a new account as well to test stats with a fresh sheet. My attempts yesterday were more booch-filled then usual and TheCosmicChz and Dcurd's stats show that well =P

I'd be curious to see fruits foraged to those stats as well, perhaps even seperated into common and rare. I love that I can see my average level tho =) If I have a lot of free time I probably will create a fresh alt to see just how high I can get that average. Without booches I can usually keep myself on level 5 for a long time on solo play.

I've been having a lot of fun with Nibbler so far, Tcarr, and just wanted to thank you. I've probably dotted in to play just about every incarnation its gone through.

Cheers.
----------------------------------------
--Dcurd :
Not the cap'n of Dcurd's crew
Stick-wielder of questionable sanity-or something
[Feb 16, 2007 9:59:01 AM] 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

I'd like to see the counts on cheap vs rare fruits myself, but am concerned about fitting all of the info in the table. It would be easier to include the total poe or average poe instead, but then the fact that gold spawns way too often will make it a lot harder to tell the cheap vs rare fruits. I'll see what fits in the table.

About the scoring (more details than anybody but a programmer wants to see)

The current scoring system determines the total points for the team in the following manner:

1. For each player, count number of carts he contributed at least one fruit to filling, number of gems he contributed, and number of fruits he contributed. A gem that is covered by both a row and column counts as two gems.

2. For each row and column, determine whether somebody on the team filled a cart there. Two players that both fill row C would only count as one row filled for the team.

3. Difficulty levels are now averaged to give the team difficulty level. teamFruits is sum of all the player fruits. teamGems is sum of all the player gems.

4. baseCartFill = 16 + (int)(8 * teamLevel)
base team score from carts = numTeamCarts * (baseCartFill + 5) - 5
extraPieces = teamFruits + teamGems - 6 * teamCarts;
if (extraPieces > 0) teamScore = teamScore + extraPieces
teamScore = teamScore + teamGems * 10

We now have the team's score determined.

5. Team score is divvied among the players in proportion to the number of pieces (fruits and gems) each player contributed to the filled carts. If one player contributes 5 pieces, and a second just 1 to make one full cart, then the first player gets 5/6 of the points.

Forage items are then calculated as a function of teamCarts and teamGems (and Math.random() of course). One item is foraged for each of the teamCarts, to a max of 6 items per round.

(current formula) Probability of getting Gold = teamGems * teamLevel * 0.0625. This means that having a lot of gems really helps, and so does getting the average difficulty level up. We probably eventually want to make 0.0625 lower, and also change the simple "teamLevel" to "teamLevel - 1" (so that lower levels don't get any gold at all), and possibly change "teamGems" to "teamGems - 1" (so that you have to have at least 2 gems to get the gold).

Any remaining items to be foraged use a probability of getting cheap fruit that is based on the teamLevel. Any fruit that isn't cheap is expensive. Current probabilities are:

Level 0: 100% cheap
Level 1: 75% cheap
Level 2: 67% cheap
Level 3+: 50% cheap
----------------------------------------
LordKalvan of Otherwhen, all oceans but mostly Midnight
[Feb 16, 2007 11:08:04 AM] 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