• 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: 4
[Add To My Favorites] [Watch this Thread] [Post new Thread]
Author
Previous Thread This topic has been viewed 335 times and has 3 replies
Robdonald



Joined: Mar 11, 2016
Posts: 2
Status: Offline

Puzzle Competitions, Secret Strats, and Replays Reply to this Post
Reply with Quote

Execute summary: use replays or livestreaming to fix puzzle competition secret strategy issues.


After the Grey Havens takeover, I logged on a bit, won some familiars, tried this duty puzzle competition nonsense, then ran out of free time and stopped logging on long ago. Feel free to tell me my concerns are no longer valid and this has all changed!

As I see it, winning 1v1 puzzle tournaments still (mostly) works, as a reasonable amount of how-to-play is published, having a some top level matches and discussions with top players can bring one up to the current metagame, and you just have to do a bit better than the people logged on to win the familiar.

Duty puzzle competitions, however, have been broken by the age of the game. The issue is one of secret knowledge.

In order to win a duty puzzle competition, you must:
1) Have the secret knowledge: what the strategy is to score very high.
2) Execute it better than anyone else there.

While back when puzzle pirates was new, the competition of figuring out the secret knowledge before everyone else lent excitement to the game, in this current bygone era, secret knowledge is primarily a matter of archaeology. Someone, many years ago, figured out how to maximize their puzzle score for a given puzzle in a competition, and if you don't know what that is, you have zero chance against someone doing the optimal strategy even passably well. You might discover it by talking to oldsters, or mining old forum posts, but it's really hard to do this for a sufficient number of puzzles to even compete. And even if you do, you don't know if it's actually currently the optimal known technique. There might have been discoveries or game patches since it was pioneered.

Here's what my current ideas/guesses of the secret knowledge are, without doing much archaeology:
1) Alchemistry:
* First, get your rating high enough to get quicksilvers. Then, fill all but one bottle to the three-star level, using your usual score maximization via maximizing colors and preferring to do an opposite color fill. Leave a quicksilver tube somewhere.
* Then, fill every bottle to within one label of the top, converting your quicksilver the fill before this is done.
* Then, use quicksilver to fill every single bottle at once. Not clear if maximizing non-quicksilver color fills is worthwhile here or not, but the difference is probably very small, only as much as one label score in one bottle.

The idea here is to maximum score per filled bottle, and additionally maximize bonus score for extra bottles filled on the last move. Presuming the extra bottle bonus is significant, this would probably be unbeatable without doing a max-final-fill strategy.

2) Rigging:
* Be aware of the secret or not-so-secret rules as to what duty reports get scored, whether pre-contest starts are counted and in what circumstances, in what circumstances pre-contest sparkliness can contribute to contest score. I don't know ANY of that.
* Loop the entire board with a bunch of tars contained.
* Use the awesome pieces you get to loop again before first loop sparkly fades.
* Somehow convert this to a maximum sort-of-duty-report(s) that count for the contest????

3) Foraging.
* Be aware of the forage scoring bug, assuming that's still a thing?
* Get a very very very specific rating to get the specific set of forage boxes you need to win the contest.
* Convert those boxes to a really fast forage session, without losing score due to the forage bug. Presumably, no one who isn't in that exact rating/experience range can beat you.

4) Blacksmithing.
* Get pretty high combo counts as you approach a mostly-silver board.
* Calculate a brute force solution that removes all remaining pieces precisely.

...That's all I've got off the top of my head. But, the example suffices. You see the problem here? We can't even truly COMPETE unless we know what the heck the optimal strategy we're competing at, and the hidden knowledge of puzzle pirates and the mists of time mean basically one or two people at most in the entire contest have any clue.

Given source code access to the game, I'd solve it by creating a replay functionality by recording people's moves and initial board state, only doing familiar contests for things I had a replay function for, and publishing the replays of the winning player and runner up every time.

If I were to fix it with merely the power of a player, I would fix it by winning a ton of familiar contests, then running two-familiar-prize contests in which you were required to livestream your play in order to win the contest. This would enable the publication of optimal strategies over time.

Anyway, I don't have time to fix it the player way, but some thoughts that randomly ran through my head today about competitions in this game I don't currently play! Do with them as you wish.

(Edit 1: added executive summary)
----------------------------------------
[Edit 1 times, last edit by Robdonald at Mar 15, 2019 2:45:09 PM]
[Mar 11, 2019 7:33:59 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
hidemyhoney

Member's Avatar


Joined: Dec 26, 2017
Posts: 375
Status: Offline
Re: Puzzle Competitions, Secret Strats, and Replays Reply to this Post
Reply with Quote

 
After the Grey Havens takeover, I logged on a bit, won some familiars, tried this duty puzzle competition nonsense, then ran out of free time and stopped logging on long ago. Feel free to tell me my concerns are no longer valid and this has all changed!

As I see it, winning 1v1 puzzle tournaments still (mostly) works, as a reasonable amount of how-to-play is published, having a some top level matches and discussions with top players can bring one up to the current metagame, and you just have to do a bit better than the people logged on to win the familiar.

Duty puzzle competitions, however, have been broken by the age of the game. The issue is one of secret knowledge.

In order to win a duty puzzle competition, you must:
1) Have the secret knowledge: what the strategy is to score very high.
2) Execute it better than anyone else there.

While back when puzzle pirates was new, the competition of figuring out the secret knowledge before everyone else lent excitement to the game, in this current bygone era, secret knowledge is primarily a matter of archaeology. Someone, many years ago, figured out how to maximize their puzzle score for a given puzzle in a competition, and if you don't know what that is, you have zero chance against someone doing the optimal strategy even passably well. You might discover it by talking to oldsters, or mining old forum posts, but it's really hard to do this for a sufficient number of puzzles to even compete. And even if you do, you don't know if it's actually currently the optimal known technique. There might have been discoveries or game patches since it was pioneered.

Here's what my current ideas/guesses of the secret knowledge are, without doing much archaeology:
1) Alchemistry:
* First, get your rating high enough to get quicksilvers. Then, fill all but one bottle to the three-star level, using your usual score maximization via maximizing colors and preferring to do an opposite color fill. Leave a quicksilver tube somewhere.
* Then, fill every bottle to within one label of the top, converting your quicksilver the fill before this is done.
* Then, use quicksilver to fill every single bottle at once. Not clear if maximizing non-quicksilver color fills is worthwhile here or not, but the difference is probably very small, only as much as one label score in one bottle.

The idea here is to maximum score per filled bottle, and additionally maximize bonus score for extra bottles filled on the last move. Presuming the extra bottle bonus is significant, this would probably be unbeatable without doing a max-final-fill strategy.

2) Rigging:
* Be aware of the secret or not-so-secret rules as to what duty reports get scored, whether pre-contest starts are counted and in what circumstances, in what circumstances pre-contest sparkliness can contribute to contest score. I don't know ANY of that.
* Loop the entire board with a bunch of tars contained.
* Use the awesome pieces you get to loop again before first loop sparkly fades.
* Somehow convert this to a maximum sort-of-duty-report(s) that count for the contest????

3) Foraging.
* Be aware of the forage scoring bug, assuming that's still a thing?
* Get a very very very specific rating to get the specific set of forage boxes you need to win the contest.
* Convert those boxes to a really fast forage session, without losing score due to the forage bug. Presumably, no one who isn't in that exact rating/experience range can beat you.

4) Blacksmithing.
* Get pretty high combo counts as you approach a mostly-silver board.
* Calculate a brute force solution that removes all remaining pieces precisely.

...That's all I've got off the top of my head. But, the example suffices. You see the problem here? We can't even truly COMPETE unless we know what the heck the optimal strategy we're competing at, and the hidden knowledge of puzzle pirates and the mists of time mean basically one or two people at most in the entire contest have any clue.

Given source code access to the game, I'd solve it by creating a replay functionality by recording people's moves and initial board state, only doing familiar contests for things I had a replay function for, and publishing the replays of the winning player and runner up every time.

If I were to fix it with merely the power of a player, I would fix it by winning a ton of familiar contests, then running two-familiar-prize contests in which you were required to livestream your play in order to win the contest. This would enable the publication of optimal strategies over time.

Anyway, I don't have time to fix it the player way, but some thoughts that randomly ran through my head today about competitions in this game I don't currently play! Do with them as you wish.


Old news. Same people win them anyway. Theres a previous post somewhere showing Secret Strats in all the puzzles.
----------------------------------------
Support Sov! and I miss Jake :(

Remove dead oceans (Jade/Opal/Meridian), all those wasted new players :/

Idk why people use alts to be relevant bashing everything.......
----------------------------------------
[Edit 1 times, last edit by hidemyhoney at Mar 11, 2019 7:38:08 PM]
[Mar 11, 2019 7:36:49 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
Kyura94

Member's Avatar


Joined: Oct 31, 2010
Posts: 284
Status: Offline
Re: Puzzle Competitions, Secret Strats, and Replays Reply to this Post
Reply with Quote

Robdonald wrote: 
The issue is one of secret knowledge.
...
In order to win a duty puzzle competition, you must:
1) Have the secret knowledge: what the strategy is to score very high.
2) Execute it better than anyone else there.
...
if you don't know what that is, you have zero chance against someone doing the optimal strategy even passably well
...
You see the problem here? We can't even truly COMPETE unless we know what the heck the optimal strategy we're competing at
It's not that prevalent across puzzles actually, and not much of a secret anymore given the experience of the playerbase. Anyway, this is what I find still being relevant, for the curious folks:

Top-Down according to pirate page listing:

Sails
DR manipulation isn't exclusive to sails, but it's about the best you can do to in terms of 'secret knowledge'. Set up strong combo, hit Esc and idle for something like 3 minutes, then release combo close to a LP. Your score would have reset, so you jump from 0 to sparkles and just idle out the Incred.
You still need the brains to set up a strong combo. Won't beat the best sailors even if they don't use this trick.

Rigs
Massive Tar loop is correct. Set up, 3min DR manipulation, double massive Tar, profit. But Rigs has so many botters so there's no pride in winning Rigs anyway.

Carp
No 'secret knowledge' AFAIK.

Patch
Theoretically you could use the 3min DR manipulation but the benefit is so diminished due to the high-speed nature of Patch.
You still need the brains to patch well for the whole LP. Won't beat the best patchers even if we don't use this trick.

Bilge
Multiple techniques, don't think there's been thorough study on which is the best, but there's some element of SetupTime vs HighScore involved.
1) Build up water to accumulate crabs, do the 3min DR manipulation whilst letting water clear, then click to clear multiple crabs at once, profit.
2) Set up multiple patterns, do the 3min DR manipulation, profit.
3) Learn disjointed combos, do the 3min DR manipulation, pull off a supervegas or superSD, profit.
Easily beats the best bilgers who don't use tricks. But bilge has so many botters so there's no pride in winning Bilge anyway.

Guns
Fill all the cannons 2/3 of the way, pause your board such that the cannonballs are ready to enter 2 cannons. You should have also counted your barrel so you know where the next 2 cannonballs will come from. When the next DR starts, unpause for easy 2/4 cannons, do the other 2, then gun the rest of the LP as normal.
You still need the brains to gun well for the whole LP. Won't beat the best gunners even if we don't use this trick.

Dnav
No 'secret knowledge' AFAIK.

Distil
Pick your board religiously because every spice counts. But meh.
You still need the brains to distil well for the session. Won't beat the best distillers even if we don't use this trick.

Alch
Pick your board religiously. Also, the way you're using Quik isn't optimal. Use it as an additional colour as opposed to a one-size-fits-all autofill for more score.

Shipwright
Pick your board religiously because the opening few moves can be gambled with and abandoned at no cost. But meh, it's all luck anyway.

Blacksmith
Nothing. Just regular blacksmith. But BS has so many botters so there's no pride in winning BS anyway.

Forage
No 'secret knowledge' AFAIK. Just be aware of the bugs.

Weave
No 'secret knowledge' AFAIK.
----------------------------------------
Ryuken on Obsidian (active) and Emerald (not really)
I made an in-depth Distilling guide here, and a guns one somewhere.
I gave Obsidian their Owls c:
[Mar 12, 2019 6:14:05 AM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
Robdonald



Joined: Mar 11, 2016
Posts: 2
Status: Offline

Re: Puzzle Competitions, Secret Strats, and Replays Reply to this Post
Reply with Quote

The post was actually about general fixes to secret knowledge issues, to bust out if Grey Havens ever gives up and open sources it, or passionate players try to fix the problem via well-structured competitions. But replying to your comments on existing secret knowledge for posterity:

 

[... other duty puzzles]
Dnav
No 'secret knowledge' AFAIK.


Every single duty puzzle has a bit of secret knowledge you haven't mentioned: the optimal rating/experience/star level to maximize score.

For many of these, it may be optimal to have the highest rating possible. Back in the day before puzzle competitions were a thing and restart penalties existed, I optimized dnav by having the precise rating needed for three-star starts, and restarting the puzzle until three-stars-on-the-inner-ring were available. I don't recall if I used this technique to win my first familiar in the best pirate race final or not.

During the first duty puzzling competition on Midnight, there was some secret knowledge regarding carpentry: it was not clear whether it was better to be ultimate, or to have a fresh alt who had never played the puzzle and got instant sparkles via the tutorial levels. But that competition was more about first league incredible than highest league. I'd guess that max rating is always optimal for carp now, and many other puzzles.

 

Alch
Pick your board religiously. Also, the way you're using Quik isn't optimal. Use it as an additional colour as opposed to a one-size-fits-all autofill for more score.


Not that I currently play, but I think you are misinterpreting the comment, and could be willing to log on to head-to-head you if you disagree with the presented strategy to prove which was better.

It is optimal to use quicksilver to get an extra color to maximize your score when you fill your initiial all-but-one bottles, essentially always.

What I am saying is that on the FINAL FILL, it is optimal to fully fill every single bottle in order to maximize the extra bottles bonus, and this is most easily done by quicksilvering every bottle. I express doubt that we need to do anything other than Voodoo the final fill. Yeah, a 5-color-fill that fills every bottle WILL score slightly higher, but this probably only effects the score of one label on one filled bottle on the final fill, so it's probably not worth worrying about in terms of value per time, and difficulty of achievement. It is difficult to fill every bottle AND multi-color at the same time, and the vast majority of the value on the final fill would be to maximize extra bottles filled, assuming extra bottle bonus exists.

Non-competition alchemistry typically does not bother doing this, as filling every bottle precisely one label from the top is time consuming and not worthwhile outside of competition.

 

Forage
No 'secret knowledge' AFAIK. Just be aware of the bugs.


Again, the secret knowledge consists of what level to play the puzzle at in order to maximize one's score. Of course, our default position is that maximum level is best, so there's no necessity of rating tanking to win a competition, but whether or not that is true is still secret knowledge.

This post over here implies there's a very specific rating range that is NOT the top level (yay archaeology): https://www.reddit.com/r/Games/comments/1zz1jj/what_happened_to_puzzle_pirates_how_can_it_be/cfyi4bw
[Mar 12, 2019 11:21:26 PM] Show Printable Version of Post        Send Private Message [Link]  Go to top 
[Show Printable Version of Thread] [Post new Thread]

Puzzle Pirates™ © 2001-2016 Grey Havens, LLC All Rights Reserved.   Terms · Privacy · Affiliates