Variable length endings

[Colonel Talvela]

I am courious if they are completely variable, or if they are dependent on flag contesting/point considerations.

I was wondering if anybody wanted to test variable length endings for me. I am too lazy.

[Der Kuenstler]

LOL - I'm off tomorrow CT - I've always wondered that myself - I'll see what I can find out...

[Colonel Talvela]

Thank you good man. I would give you another clickie ... nevermind.

[Colonel Talvela]

Well I checked the manual (imagine that)...it says the ending is dependent on flag control. Bummer for the game I am trying to end ASAP

[The Coil]

I was wondering if anybody wanted to test variable length endings for me. I am too lazy.

Actually, I ran a test on that once a long time ago, when I apparently could find nothing better to do for 4 hours or so. If I remember right, I set up a test wherein there was a squad of infantry sitting around the flag. One set of trials, an attacking squad engaged them from close range, the next set they stood way far away and didn't shoot at all.

My main conclusions were:
1) Even me, with my considerable threshold for boredom (go look up that treeburst test from a while back as exhibit A), got bored before I ran a statistically significant number of trials.

2) There seemed to be some marginal average increase in extra turns with the firefight going on, but the manual seems to overstate the difference a bit (IIRC it says something like "effectively a game won't end if there is a firefight going on around a flag"). And even if there was a difference in the average game length, the range of #'s of extra turns wasn't all that different - ie there were sometimes games with no shooting that went 5+ extra turns, and sometimes games with shooting that ended with 0 or 1 extra turns. So expect on average a longer game, but in any given game, it could end right away or go on for a while.

[Der Kuenstler]

Variable Endings

First let’s start with a direct quote that the manual gives us:

“When the turn limit has been set to ‘variable’ in the parameters, and one or more victory flags changes hands near the end of the battle, the battle may be extended several more turns, out to a maximum of ten extra turns or a 25% increase in the original battle’s length, whichever is less. Essentially this means a battle won’t stop short while active fighting is going on, within limits.” (P.42)

The person I found who appears to know the most about this is Treeburst155 at Battlefront. I will include some of his quotes along with the manual quotes:

1) The manual says extra turns kick in when “one or more victory flags changes hands near the end of the battle” - This is as clear as mud - what is the designers’ definition of “near the end”?

According to Treeburst:
“The variable ending kicks in with flag status changes AT LEAST SIX turns before the scenario end turn. That's why it is so common to go into overtime. Any flag changes in virtually the last quarter of a typical game will trigger it!!”
“I KNOW the variable ending is triggered REAL early. I would have guessed flag changes in the last two or three turns, but it's at least six turns.”

2) Apparently you can extend a game by moving away from a flag located behind your lines - this changes the flag to neutral. “What I know right now is that the game can be extended at will, and with no risk, at least once, by any player who has a flag behind his lines....By [deserting] one of your VLs near the end you could have gone into overtime if you wanted.” - Treeburst155

3) This is the quote I’m inclined to believe (a “die roll” explanation) because it is so consistent with my experience in all of my variable ending games:

“A 40+ game will go at least 40 complete turns, but no more than 50 complete turns (10 extra turns max, or 25% of designer specified game length, whichever is less). This "die roll" is thrown at the end of turn 40, and ONLY at the end of turn 40. The die has 11 sides (in the case of a scenario designed with 40 or more turns), numbered 0-10. The number of extra turns is set by this die. What this means is that the chances of the game ending on any specific turn are equal (9%). So, there is about a 55% chance the 40 turn game will go 5 or more extra turns. There's only a 9% chance of the game ending on the designer specified turn. It's best to think of a 40+ game as 45 +/-5. Odds are over 70% that you will get at least two extra turns.”

If Treeburst155 is correct then there is no use spending hours testing variable endings in the editor - once a flag changes it’s status near the end of the game a random “die roll” decides how much longer it will be. I have sent Treeburst an email to find out where he got his info in order to confirm its validity, and will add updates and corrections here as needed.

[jawsconan]

Hi,

I just finished a game VS Nashorn and the game ended at 50 turns for a 40+ (scen "well of the souls")!
10 additionnal turns. I never have seen that before.
It is a problem to conduct strategic plans to keep the flag. Because the opponent has more time to manage the same way and usually, a great number of additionnal turn means more random end.

cheers.
Hubert

[Der Kuenstler]

10 additionnal turns. I never have seen that before.

That would be the maximum according to the manual - it happened to me once, too - also in a 40+ turn game - it will only go that long if the battle is 40 or more turns. The 25% rule would mean that the very longest one of POS' 20+ turn battles could go is 25 turns.

Here's a follow-up email from Treeburst155, whose real name is Michael Meinecke - thanks to him for the quick response:

"I discovered how the variable ending works by MUCH empirical testing....time and time again. Finally I figured out how the program behaves. Can I prove it? Not really....but I'm right. :) I used to have FAR too much time on my hands. I tested this variable ending thing for days. This is the only conclusion as to program behavior that I could come up with that fit my data."

[Mike Abberton]

I wonder hwo the results in a game with less than 40+ turns is calculated. The game could:

a) recalc the "die roll" using whatever the max extra turn number is (so 0-5 for a 20+ turn game); or

b) use the standard 0-10 roll and just make any result over 5 equal to 5.

I sort of suspect option B, since I seem to have a lot of short variable scenarios go 5 or 6 extra turns, but I could be imagining it.

[Der Kuenstler]

In Treebursts theory if it was 25 turns with 25% rule the most it could go would be 25 X 25% = 6 turns (rounded down) so a 7 sided die would be used (0-6) to see how many extra turns it would be.