1st Automated Beta Tournament!

[IainMcNeil]

To be clear it is the % of total BG losses for a side, not actual BG losses.

This is a first pass scoring system for beta testing, not necessarily set in stone for the future. It does not reward routing the enemy army enough I think.

[maximvs]

Keeping score? -

Apparently the ancient Egyptians used to ask their soldiers to cut off the hands of the enemy slain so they could be brought back and counted. Not that pure numbers necessarily gives a true picture, but it certainly avoided the possibility of the same body being counted twice!

[Blathergut]

The time system seems to be a bit of a concern. If you happen to send your opponent his turn once he's tumbled into sleep, or trudged off to slavery, he loses (or you do, in reverse) 8-10 hours.

Could the time count down system instead be tied to when you actually load the game?

If I only had 1 hour maximum to complete all my moves (or whatever time amount works), that might be a better system.


This is very noticable for those in Australia. My opponent is there, 13 hours difference, and we have a very small window each morning/evening to get turns in. The current system I think is also open to abuse. I could deliberately wait until my opponent is sleeping/working to send his turn.

[IainMcNeil]

No, this is by design this way. It is the only logical way for it to work so it can't change I'm afraid. If you can't connect and play the turns you'll have to miss out and not play. It is the only way to ensure games have completed by the deadline to kick off the next round. Someone who is losing could just stop playing.

Also if you have a time difference working against you so does your opponent. You'll have an 8-10 hour gap after you move before they get the turn to you so there is none who has an advantage. You may have to play a turn first thing in the morning to keep it moving.

Fundamentally if you don't have the time to play you can't enter.

[hidde]

The forumla is :
Player 1 % = (Player2Losses / (Player1Losses + Player2Losses)) * 100%
Player 1 Score = 25 * Player 1 %
Player 2 Score = 25 - Player 1 score.

I don't think your formula is calculating correctly. I have done it manually below. Also your scoring system seems odd as it has potential pitfalls if used to calculate anything other than a mirrored match. Wouldn't it be better to have a single formula that will resolve all results? You already calculate % losses of total army in multi-player matches so why not tie in your scoring system to it.

pantherboy vs cheerfullyinsane (slitherine puts it at 21 to 4 points)
6/22 vs 48/37

Player 1% = (48/(6+48))*100 = 88.88%
Player 1 Score = 25*88.88% = 22 (rounded down)
Player 2 Score = 25-22 = 3

pantherboy vs cheerfullyinsane (slitherine puts it at 17 to 8 points)
20/37 vs 23/22

Player 1% = (23/(20+23))*100 = 53.48%
Player 1 Score = 25*53.48% = 13 (rounded down)
Player 2 Score = 25-13 = 12

I'm still curious about this.
When I do the math I get my result to be: 13+15
The scoreboard show 16+11
The end result is only one pts difference but it would be nice to have an explanation.

[Blathergut]

How does the scoring work if your opponent's time runs out and you were no where near finishing the game?

[deeter]

Fundamentally if you don't have the time to play you can't enter. -- Ian

I don't mean to be a problem, but this policy seems rather arbitrary and not very inclusive. It's limiting to people in different parts of the world, with different daily schedules and with different lifestyles. I may work both ways with the time zones, but what about those who work on weekends or have other constraints?

For example, I'll be out of town for the first three days of round four. What can I do aobut that?

Deeter

[IainMcNeil]

Nothing - sorry.

[Jonah75]

Keeping score? -

Apparently the ancient Egyptians used to ask their soldiers to cut off the hands of the enemy slain so they could be brought back and counted. Not that pure numbers necessarily gives a true picture, but it certainly avoided the possibility of the same body being counted twice!

Or at least only twice. Two hands, see.

[batesmotel]

Any chance of getting the evade bug fixed in time for the remaining rounds of the tournament? For anyone who hasn't run into it yet, the bug leads to issues like Swiss unarmoured, crossbow, average cavalry refusing to evade from pikes or knights.

Chris

[deeter]

Ditto.

Deeter

[deadtorius]

Ouch.....

Will be getting another e-mail before each round begins or should we just keep logging in to see if any unscheduled games have shown up?

[Blathergut]

One would expect them up on Friday. I think there was a notification email last time.

[maximvs]

Just so you all know, I'm down in London for the next couple of days, but I will be home again by Sunday lunchtime.

So whoever gets drawn against me tomorrow, have your first turn as normal, but obviously don't expect me to respond (or take my first turn of the other game) until Sunday PM.

If I have calculated correctly, this will be within the 2.5 day time allowance (just!). Phew ...

Max

[IainMcNeil]

Time allowance is 3 days so you have plenty of time

When the draw is made you'll all be notified by e-mail.

[maximvs]

Thanks Ian.

Hey, I just noticed, you scored 25 points times 2!!! You jammy devil ...

[IainMcNeil]

Yes... only because he ran out of time. Oh good - I get to play Pantherboy

[IainMcNeil]

By the way 1.3.5 is out and fixes the evade bug... hopefully.

[batesmotel]

Have you ever played in a chess tournament using Swiss system pairings? I just looked at the 2nd round beta tournament pairings and the beta match ups using the scoring system aren't going to work well. The Swiss system assumes that you are using scores per round that will cause players to break up into a few large groups. Normal Chess scoring is 1/.5/0 or was back when I played back in high school. Based on having a rating system so you you have some idea who the best players are, the first round takes the top half of the players and matches them against the bottom. After the first round, for example, you would take all the winners and split them in half, and then match the top half versus the bottom half the idea of the system is that the best players (assuming no upsets) will quickly sort out with the best scores and in theory the top players will only face off in the last few rounds and then place accordingly.

The way the scoring system for the beta match is working is that nearly every player has a different score so the second round ends up being each player against the next lower one which is going to tend to force all the scores to the middle I suspect rather than sorting out the best player. The final ranking will probably have the top half of the players in the top half but details of the ordering after that will probably be pretty random. With the scoring system used for the beta, it would probably be better to take the top half of the standings at the end of a round and then split them with the first quarter versus the second quarter. Similarly for the 3rd quarter versus the fourth. This will be much more likely to produce a clear winner for the tournament. Alternatively, changing to a scoring system with points awarded for win/loss/draw and then just using the relative AP loss levels for tie break might be a better way to go for the long term.

Hopefully the beta tournament will be good for shaking out a lot of these issues at any rate.

Chris

[IainMcNeil]

This is the system we've used in tabletop tournaments for 20 years or so and works fine so I think we'll be ok