Simulations of One-On-One Battles: Original Starfire Book 1

Home of the 1st and 2nd editions of STARFIRE

Moderators: SDS Members, SDS Owner

Forum rules
Forum rules
1. Nothing obscene.
2. No advertising or spamming.
3. No personal information. Mostly aimed at the posting of OTHER people's information.
4. No flame wars. We encourage debate, but it becomes a flame when insults fly and tempers flare.

Try to stick with the forum's topic. Threads that belong to another forum will be moved to that forum.

Re: Simulations of One-On-One Battles: Original Starfire Book 1

Postby Whitecold on Sun 17 Jul 2016 05:13

For simulating range I would just ignore BS and turning, just let each ship move to its desired range at top speed. So if both want to get closer, they move towards each other at top speed, if the faster ship wants to get closer, they close by the difference, if the slower ship wants to get closer, no luck, the range stays open.
Whitecold
Rear Admiral
Rear Admiral
 
Posts: 446
Joined: Fri 19 Sep 2014 15:03

Re: Simulations of One-On-One Battles: Original Starfire Book 1

Postby Graywulffe on Wed 20 Jul 2016 10:12

thebard wrote:One way to simulate this could be varying the starting distance, so that in different simulations , firing will happen at different distances.


This is a good idea, and I have not tried it. I like Whitecold's suggestion, too. I will first look into randomizing the starting distances, as this could be implemented most quickly. Thanks much for the suggestions.

-best
Graywulffe
Lieutenant JG
Lieutenant JG
 
Posts: 23
Joined: Sat 28 May 2016 08:35
Location: Vancouver, BC

Re: Simulations of One-On-One Battles: Original Starfire Book 1

Postby Graywulffe on Sat 06 Aug 2016 09:18

Part VII: Return to Primary Beams Vs. Force Beams

I have now implemented random starting positions in the Original Starfire Book I battle simulator that I have been programming. At the start of each game, the ships are randomly placed from 0 to 15 hexes away from zero. Both ships have the same distance, but for one it is a negative number. Thus, if the random starting location is 8, then Ship1 is placed at 8, and Ship2 is placed at -8 on the number line. They are 16 hexes apart. The rest of the simulation remains the same, with the ships using their full movement allowance each turn, and doing a full-frontal charge.

Using the same destroyers from the previous two experiments to test primary beams against force beams:

DD1 – Stinger – (2) – SSSAAAIIPIPIPDI (5)
DD2 – Ram – (2) – SSSAAAIIFIFIFDI (5)

I ran 100,000 games. The Ram, with a compliment of three force beams, had a 9:1 advantage. Thus, the Stinger, zapping away with a trio of primaries, won one game in nine on average, or 11.1% of the time. A successful outcome for the Stinger, of course, centered on having good luck with the weapon's random system destruction. The early removal of the Ram's force beams is important. This is exemplified during those games where the starting location was zero: When the Ram won initiative and fired first, the Stinger with 15 total systems was destroyed instantly (5 damage per force beam). If the Stinger fired first and hit one or more of the Ram's force beams, then the primary beams provided a fighting chance.

I have the program save the data from each game in the test. Looking at the average number of turns per game for the given starting location, there is an upward trend with the turn average increasing with increasing distance. Here is the data:

Starting Position, Average Number of Turns Per Game
0, 2.9
1, 4.8
2, 6.1
3, 7.5
4, 5.4
5, 6.0
6, 4.0
7, 7.7
8, 10.6
9, 7.6
10, 8.1
11, 5.4
12, 10.9
13, 14.3
14, 9.8
15, 9.8

A linear fit returns an R-squared of 0.54 with a positive slope—not the best statistical approach for count data (maximum likelihood would be better), but a quick-and-dirty glimpse that hints about a relationship. Part of the relationship has to do with the fact that the further away the starting location, the more turns are required to reach close range. However, of interest is that starting with the first peak in average games per turn, 7.5 at a distance of 3, further average-game-turn maxima occur every five steps: starting location eight and again at 13. The movement allowance of each ship is also 5 at the beginning of each game. The apparent relationship is likely not a coincidence. Note that, with 100,000 games, each starting location has roughly 6,250 observations, so these game-turn averages are quite robust.

Thinking about a starting location of 3: After firing at a distance of six on turn 1, the ships move five hexes toward each other (assuming no engines were destroyed) on turn 2. Thus, they zip past end up four hexes apart for the next round of fire. This keeps the force beam damage down to one or two systems per hit, while the primaries destroy one system per hit. The Stinger will do better under these circumstances than at closer ranges where the force beams can deliver serious blows.

Longer games (more turns) translate into a higher likelihood of the Stinger surviving, in part because it is likely getting off more shots. This is captured in the data. The Stinger had no wins during any game lasting just 1-4 turns (n = 38,384 games), and more than 50% wins for games with more than 16 turns (n = 7,786 games). Long games likely also reflect the Stinger successfully destroying the Ram's force beams early, and many of its engines. The Stinger then ends up chasing the Ram down while trying to hit the last few internal systems. Or, the Ram loses all of its engines and the Stinger just hovers nearby taking shots. The longest game was 98 turns—this likely was one of these situations, probably also with the Stinger quite crippled and moving slowly. Yes, during that game the Stinger suffered the loss of 11 out of 15 systems, leaving just two intact engines and one primary beam. During the game, the Stinger got off 51 shots, almost the highest of all the games in the test. This game, by the way, had a starting position of 13.

Thus, my interpretation of the data is that when faced with force beams, a ship loaded with primaries is probably better off attempting to surgically remove the enemy's weapons while keeping some distance (>5 hexes)—if possible. Note that the captain of the Ram may elect for a distance approach, trying to stay either out of range of the primaries while delivering ½ damage points to the Stinger, or hovering at a distance where the force beams have an effect of 1 while the primary beam attacks are only capable of ½. Or, the Ram's captain may attempt a close-range strike, hopefully destroying the Stinger with one pass.

-best
Graywulffe
Lieutenant JG
Lieutenant JG
 
Posts: 23
Joined: Sat 28 May 2016 08:35
Location: Vancouver, BC

Re: Simulations of One-On-One Battles: Original Starfire Book 1

Postby igycrctl on Sat 06 Aug 2016 12:05

Would you consider making your source available? I would be interested to see it.
igycrctl
Lieutenant JG
Lieutenant JG
 
Posts: 20
Joined: Tue 15 Apr 2014 12:45

Re: Simulations of One-On-One Battles: Original Starfire Book 1

Postby Graywulffe on Sun 07 Aug 2016 16:35

igycrctl wrote:Would you consider making your source available? I would be interested to see it.


Hi, igycrctl. Thanks for the query. Currently I am not releasing the code. This is in part because it is not finished and also because it almost certainly can be refined. I have not ruled out making the source available sometime in the future, however.

-best
Graywulffe
Lieutenant JG
Lieutenant JG
 
Posts: 23
Joined: Sat 28 May 2016 08:35
Location: Vancouver, BC

Previous

Return to Original Starfire

Who is online

Users browsing this forum: No registered users and 2 guests

cron