Jack lost to Seth in 1 turn. He did 80 at range 2. GL to Seth and everyone else in the tournament.

By Peter Bakija (Bakija) on Sunday, October 05, 2025 - 04:21 pm: Edit

>>Jack lost to Seth in 1 turn. He did 80 at range 2.>>

Yeah, GBS vs LDR seems like a particularly terrible matchup. It's one of those "The GBS just need to move forward at moderate speeds with weapons armed" situations. The LDR is gonna have to be super cagey and squirrelly, but eventually, the GBS is just going to paste it.

By Carl-Magnus Carlsson (Hardcore) on Monday, October 06, 2025 - 07:26 am: Edit

I would fight it like I was in the Orion. A range four volley on the shark's #1 and then out.

By Jack Taylor (Jtaylor) on Monday, October 06, 2025 - 09:24 am: Edit

At range 4 he would have done 60 with his rolls. I would have done 40 with my rolls. He hit with all his disruptors and has an extra phaser-1. That is 20 more. If I fire at range 8 and turn off he gets a shot on my #5 and with his rolls I take 20 internals. I am small and have no way to diminish his shot. ESGS are useless when one is destroyed and he has 2 drone racks and 2 shuttle bays. Playing against somebody like Seth I don't think there is any formula that wins that game other than Seth missing and me not missing.

By Peter Bakija (Bakija) on Monday, October 06, 2025 - 11:36 am: Edit

Yeah, that seems to be a lot of the dynamic in that fight. Like, if the GBS does something dumb, and somehow lets the faster LDR get an unanswered shot on a back shield (followed by a follow up HET return shot), the LDR could conceivably end up ahead? But yeah. Seems real rough.

By Carl-Magnus Carlsson (Hardcore) on Tuesday, October 07, 2025 - 02:42 am: Edit

Sure, if he is LUCKY he will hit with everything. The shark got no UIM however so he could just as well have missed with one of two Disruptors at range four.

If you maneuvered for the r:8/4 shot, your #2 facing hits #4 Seth would face a dilemma. If he fires everything, but you don't, you will turn in and overrun him. (He had only two drone racks iirc)
If he doesn't he risk lose much of his #1 shield for you to exploit later.

By Carl-Magnus Carlsson (Hardcore) on Wednesday, October 08, 2025 - 04:44 pm: Edit

CB over GREGG

By Carl-Magnus Carlsson (Hardcore) on Friday, October 10, 2025 - 03:47 am: Edit

Spartan over Malis1701. The top of the tree is ready for round two.

By Geoffrey Clark (Spartan) on Friday, October 10, 2025 - 04:51 am: Edit

Yes, I hope to post a summary soon, but the BBS does not like my laptop, only mobile ... ?

By Justin Royter (Metaldog) on Saturday, October 11, 2025 - 02:45 pm: Edit

I conceded to stimpy after 7 turns of •••• dice, complete •••• dice. I was literally winning.. seriously advantged the entire 7 turns but barely scratched his shields with 7 strait turns of old disr fire.

My experience with the randomizer is consistently not random, instead always HEAVILY skewed towards the extreme end of the bell, somehow an insane amount of 5s and 6s are generated.

The dice were so bad for 7 strait turns that I could not score any significant dmg over 7 turns. this happens ALL the time. This UI needs a new randomizer, something AI that is actually random.

I literally conceded while heavily advantaged without a single point of internal dmg and only light shield damage b/c the dice were so bad the entire game, and this is nothing new, it is a consistent situation in my 25 years on this platform, it needs a new randomizer badly.

There is nothing random about the sheer number of 5s and 6s I rolled in 7 turns of disr, phaser fire, and cloak effect die rolls, I rolled a crazy amount of dice in 7 turns like 20 or more in one volley with the weapon rolls and the cloak effect rolls for 7 turns the results were insanely skewed... again this is my general experience with the randomizer for 25 years now, it is complete •••• and needs a serious update to a modern AI driven randomizer.

By Daniel Bitseff (Cadet_Stimpy) on Saturday, October 11, 2025 - 05:47 pm: Edit

GG Justin, sorry about your dice. You were executing a great sub hunt, and I was having fun & wanted to continue, but wasn't going to try too hard to talk you out of conceding when you were still in great position.

I knew his dice were bad, but was curious as to how bad. So I went through the game history & did a little analysis. Out of 75 rolls for Justin, he got 5 or 6 43% of the time; conversely he saw 1 or 2 only 21% of the time. The average roll was around 3.95 (pretty much averaged "4", which is terrible). It also seemed like the worst of the rolls came at the most inopportune times for him. And while Justin was still in a position to win it (he was still scraping away at my shields every turn), I can understand the frustration at continually sanding shields when you should be getting ints.

By Justin Royter (Metaldog) on Saturday, October 11, 2025 - 06:57 pm: Edit

Thanks Dan, its not just this game and I am not tripping. Its every game I play on this interface. I am amazed I win a single game, and don't know how I won the sapphire 18 tournament, the dice were consistently just about as bad every game. The die roller is 100% cooked. Praise to the Zin and drns.

I can think of instance after instance where my dice where just terrible, not just one shot but every shot and any roll seems to be arrive at the end of the bell and not the good one.

Its obvious to anyone with half a brain and with a massive sample size like me, that the die roller is broken. Its crazy frustrating to simply not trust the die roller, there has to be a more modern and accurate solution.

This has been an issue for literally years but today was just on another level.

By Paul Franz (Andromedan) on Sunday, October 12, 2025 - 03:54 pm: Edit

Round 1 is complete!!

Lets get those Round 2 games started, ASAP.

By Justin Royter (Metaldog) on Sunday, October 12, 2025 - 07:49 pm: Edit

from AI:

Services like RANDOM.ORG provide random numbers generated from atmospheric noise. While these are not suitable for real-time, low-latency applications like a video game dice roll, they are excellent for secure, high-entropy number generation.

By Paul Franz (Andromedan) on Sunday, October 12, 2025 - 09:30 pm: Edit

Which means Random.org is perfect for what we need. SFB Online is not a low-latency application like a video game.

By Paul Franz (Andromedan) on Sunday, October 12, 2025 - 09:39 pm: Edit

Did a quick test of Random.org in the client and the numbers look fine to me.

Andromedan Rolls 1d6: 4, 6, 6, 4, 1, 5, 1, 5, 5, 4
Andromedan Rolls 1d6: 1, 4, 4, 1, 2, 2, 1, 1, 5, 3
Andromedan Rolls 1d6: 1, 4, 3, 3, 1, 1, 3, 3, 1, 1
Andromedan Rolls 1d6: 2, 3, 2, 6, 1, 6, 5, 2, 5, 3
Andromedan Rolls 1d6: 6, 3, 1, 5, 1, 3, 2, 5, 2, 4

By Peter Bakija (Bakija) on Sunday, October 12, 2025 - 10:27 pm: Edit

From my online dice monitor:

Total (Avg Roll) 1472 (3.405570652173913)

By Justin Royter (Metaldog) on Monday, October 13, 2025 - 04:51 pm: Edit

You guys can talk about it, explain it away, all that good stuff. Can we still get a new, modern, AI driven die roller?? Just for fun??

If you are saying random.org or the internal roller (the sfbol rollers do not work, crazy latency) are my only 2 options that is disappointing.

Is there some reason we cannot at least try a new AI driven die roller?

By Justin Royter (Metaldog) on Monday, October 13, 2025 - 04:53 pm: Edit

also just a quick look at your number sets and I see a ton of extreme results, your tiny sample size is skewed towards the bell yet again, far from avg.

IF all those 1s and 6s look good to you.. then I will just stop here as I am wasting my time.

By Carl-Magnus Carlsson (Hardcore) on Monday, October 13, 2025 - 05:36 pm: Edit

AI can't do random for •••• (pardon the language).

By Eddie E Crutchfield (Librarian101) on Monday, October 13, 2025 - 06:12 pm: Edit

1+28%
2=14%
3=18%
4=14%
5=16%
6=10%

By Dana Madsen (Madman) on Monday, October 13, 2025 - 06:51 pm: Edit

Justin, you asked Paul to look into Random.org. He did so and he used it to roll 50 numbers. You don't like those numbers, and asked again to get a "new AI driven die roller".

What do you mean by AI driven, do you want an AI to skew the numbers back to average if it thinks they aren't right? Do you think AI can make a better algorithm? Or do you just want some examination to see if there is a better algorithm that can be used and AI is a current buzzword. Looking online there have been improvements in random number generation in the last 10 to 20 years, and if Paul's code that he calls was written 20 years ago, maybe there a better ways?

Btw, 50 numbers isn't a very good test, run a million numbers or so. Then use a test suite (looking online there is a program called Ent or another called diehard). Like, any of our opinions looking at a sequence of 50 numbers and going thumbs up or down is kind of useless.

--- ent performs a variety of tests on the stream of bytes in infile - The program is useful for evaluating pseudorandom number generators

--- The diehard tests are a battery of statistical tests for measuring the quality of a random number generator (RNG).

Paul, what code / library do you call to generate random numbers. Looking online years ago there was back and forth comments about the Vassal engine, and they settled on "java.security.SecureRandom". Unless this is what you already use. My suggestion is that if a larger user base then us had a long discussion about how to generate random numbers for their games, then there is a high likelihood some people their spent significant time looking into it, and maybe we could just piggyback off their effort.

Paul if you have time, (or anyone else current in java programming), maybe write a short java program that uses your current code and generate a million rolls and writes them to a text file. Then use random.org (I wonder how long a million rolls will take going externally 1 after the other). Then maybe use the same system as the Vassal library.

I'd volunteer but it's been 15 years since I've tried to write anything in Java (or programmed at all). I'd have to go start from scratch and read basic tutorials on how to even setup a current programming environment on my laptop.

If you want to send me files though, I'll run them through statistical analysis programs.

By Gregory S Flusche (Vandar) on Monday, October 13, 2025 - 07:41 pm: Edit

All I know is that at range 2 I will roll 4-6s for my overloaded photons every time.

Playing Hydran the DAC will kill of all my weapons in two sets of internals.

The first thing to die on a lyran is the ESG.

Of course, at range 8 all 4 overloaded photons will Hit.

Should I go on. The dice are what they are.

By Justin Royter (Metaldog) on Monday, October 13, 2025 - 07:42 pm: Edit

Dana that would be great, I have lost faith in the randomness of the random.org die roller, I was going to try the internal roller, the sfbol rollers do not work, never have, crazy lag.

I am asking for something new, I fix pcs for a living so I dont know much about coding. I would just like a new option that has its code based in something new, like AI, if possible, maybe I dont understand, I am just a cave man, but I was thinking that trying a new option after 25 years, something based in new code, might be nice.

By Eddie E Crutchfield (Librarian101) on Monday, October 13, 2025 - 07:43 pm: Edit

I think there is a random number generator in Excel, could it be used?