Dumping Archon, creating a new tournament guiding app?

Ankha wrote:

Sambomb wrote:

In a new app we can evaluate each round when the round start, following as many seating rules as possible.

The computing power for calculating the seatings exceeds any server's capabilities when you have more than 20 players (see www.vekn.net/forum/generic-v-tes-discussion/69811-optimal-seatings-changes-3r-f?limitstart=0)

The best you can do is to take the optimal seating, remove the dropping players, then try adding/permutating players so that the rule 1 and 2 are at least enforced.

Some people may not be aware that the current Optimal Seatings are looked-up from a previously designed table.

I've been playing around with the current Archon and have made a way to facilitate what you propose here - basically a table which indicates which player positions are currently breaking rules 1 and 2. I 've combined this with a couple of macros to randomly assign seatings as a way to quickly find an alternative seating. I will need a bit more time to clean up and fine-tune the file, but I won't have any availability to do so in the coming days - alternatively I can send it to anyone interested in developing the idea.

Beacuse we have late arrivals also, 2 players, who joined in round 2.

Ankha wrote:

ScoundrelAtHeart wrote:

So a server could not handle 20 players but a normal pc or laptop can? How could this be?
In that case maybe do the calculation on client side of the App and the server side is just for data storage.

I never said that a normal PC could. On my computer (i7-4770K overclocked at 3.7GHz), it took 688 seconds to compute the optimal seating for 19 players. And it grows exponentially with each additional player. (That's why I did some grid computing for 26+ players). I don't thing people will want to wait 10 minutes, 1 hour, or more to get the next round seating.

ScoundrelAtHeart wrote:

My goal here is to contribute to 1/ a more intuitive, easier to understand, GUI based version of Archon 2/ that does not need Excel ir other spreadsheet programm.

I agree on the first part, but disagree on the second part (it's not a goal, who cares if it's using Excel underneath or not?)

This is indeed very long. Are these numbers from the xls Archon or did you create an app and put it to the test?

Maybe someone could optimize the algorithm to make it faster.

Also not everyone can handle spreadsheets

Ankha wrote:

Sambomb wrote:

Help me to understand it better:
1st round 15 players - Optimal Seating for 15 players
2nd round 17 players - Optimal Seating for 17 players
3rd round 15 players - Optimal Seating for 17 players

Lets say 1 and 2 drop early, and turn the tables on: (round 3 for 17 players edited to drop 2 players)

Instead of: ( normal r3 for 15 players )

I don't understand your example. Why is there only 15 players on round 1? If players 1 and 2 drop early, it means they play round 1 and 2, right?

Round 1 - 15 starting players
Round 2 - 2 Late arrivals (16 - 17 )
Total players 17
Player 1 .. 17
Round 3 - 2 Early drops ( 1 - 2 )
Total players 15
Player 3..17

About the app, and if the app import optimal seating from the current archon and try to enforce it each round.
If late arrivals or early drops happens and tables need to be fixed, the skimflux algorithm can be used to fix them in a smother way.

We can turn the XLS optimal seating in a SQL table.
Players, Round, Table, Position 1, Position 2, Position 3, Position 4, Position 5

The harder thing is how to handle 11 players, maybe if 3 rounds are disabled we can work around it.