Technical question on shuttle percentage calculation

Capt. Chaos

I am trying to calculate the percentage numbers that are displayed for faction event shuttles.
According to the Wiki (https://stt.wiki/wiki/Faction_Missions), the pre-boost success is based on the average skill of the crew in each shuttle slot. That's a straight-forward calculation, just add up each base skill that matches the mission traits to get a score for each assigned crew member, taking into account the AND/OR slots. Then take the average of those scores to get the AverageSkill for the shuttle.

The Wiki says mission difficulties range from 50 to 2500. I am assuming that a 4000 VP mission has a difficulty of 2500. So the success rate should be

100. / (1. + exp(3.5  * (0.5 - AverageSkill/2500.)))

This gives a result that is consistently lower than the displayed percentage. Am I interpreting the Wiki info incorrectly? I'm not sure how that equation was derived, or if it is indeed accurate.

[PAaA] Peanut3141

Capt. Chaos wrote: »

I am trying to calculate the percentage numbers that are displayed for faction event shuttles.
According to the Wiki (https://stt.wiki/wiki/Faction_Missions), the pre-boost success is based on the average skill of the crew in each shuttle slot. That's a straight-forward calculation, just add up each base skill that matches the mission traits to get a score for each assigned crew member, taking into account the AND/OR slots. Then take the average of those scores to get the AverageSkill for the shuttle.

The Wiki says mission difficulties range from 50 to 2500. I am assuming that a 4000 VP mission has a difficulty of 2500. So the success rate should be

100. / (1. + exp(3.5  * (0.5 - AverageSkill/2500.)))

This gives a result that is consistently lower than the displayed percentage. Am I interpreting the Wiki info incorrectly? I'm not sure how that equation was derived, or if it is indeed accurate.

Could you perhaps show your work somehow? I find it hard to provide any sort of useful comment without seeing how you got here.

AviTrek

It's unclear if that 2500 max difficulty applies to events or general shuttles. The page you're on is talking about general shuttle difficulty. Also, these are numbers derived by players, not provided by DB, the formula may not be perfect.

Capt. Chaos

[PAaA] Peanut3141 wrote: »

Capt. Chaos wrote: »

I am trying to calculate the percentage numbers that are displayed for faction event shuttles.
According to the Wiki (https://stt.wiki/wiki/Faction_Missions), the pre-boost success is based on the average skill of the crew in each shuttle slot. That's a straight-forward calculation, just add up each base skill that matches the mission traits to get a score for each assigned crew member, taking into account the AND/OR slots. Then take the average of those scores to get the AverageSkill for the shuttle.

The Wiki says mission difficulties range from 50 to 2500. I am assuming that a 4000 VP mission has a difficulty of 2500. So the success rate should be

100. / (1. + exp(3.5  * (0.5 - AverageSkill/2500.)))

This gives a result that is consistently lower than the displayed percentage. Am I interpreting the Wiki info incorrectly? I'm not sure how that equation was derived, or if it is indeed accurate.

Could you perhaps show your work somehow? I find it hard to provide any sort of useful comment without seeing how you got here.

It's basically just what is on the referenced Wiki page. However, that page could be a bit clearer. I'm hoping the Wiki contributors who came up with that formula can comment on how it was derived.

Some more details:

    for the first mission trait
        score1 = trait base for the assigned character
    if there is a second trait
        score2 = trait base for the second trait
   else
        score2 = 0
    if the slot is an AND
        if (score1 > score2)
            totalscore = score1 + 0.25 * score2
        else
            totalscore = score2 + 0.25 * score1
    else if (score1 > score2)
        totalscore = score1
    else
        totalscore = score2
    multiply totalscore by the character's event bonus (1, 2, or 3)

Then calculate an AverageSkill of each slot's totalscore.
(eg. for a three slot shuttle, AverageSkill = (totalscore1 + totalscore2 + totalscore3)/3

This gets plugged into the formula.

The formula looks like this when graphed.


I suppose I can spend some time curve fitting some shuttle data to see if the formula needs tweaking.

Capt. Chaos

Update.

I find that setting the mission difficulty for the 4000 VP missions to 2025 (rather than 2500) gives me results that are within one percent of the displayed success figures.

{SST}WeAreBorg

I would love to see if this data holds true of event shuttle missions as well. Thank you for the post and please keep us posted.

[BL] C6pilot

I've been watching ST for decades and in all those years I've never seen so many missions fail. In fact I've seen them pull successful missions outta their *** at the 59 minute mark on a regular basis. We need THOSE odds when playing this game.

Paund Skumm

C6pilot wrote: »

I've been watching ST for decades and in all those years I've never seen so many missions fail. In fact I've seen them pull successful missions outta their *** at the 59 minute mark on a regular basis. We need THOSE odds when playing this game.

A Klingon, a Romulan, a Ferengi and a Gorn walk into a shuttle... OK bad bar jokes aside, what did you think the outcome would be with this circus act of a crew?

Astrometrics

Capt. Chaos wrote: »

Update.

I find that setting the mission difficulty for the 4000 VP missions to 2025 (rather than 2500) gives me results that are within one percent of the displayed success figures.

Yeah well that was easy.

Shuttle difficulty, if I remember correctly, varies by 25 each couple of successful missions. The VP don't: I think that all missions for difficulty between 2000 and 2500 always give you 4000 VP.

Capt. Chaos

To be clear, the 2025 number was the result for the 4000 VP event missions I was running on Saturday. My success rate plateaued at 85% and agreed well with the displayed percentages.

WaldoMag

@Capt. Chaos
I treat difficulty as unknown. Then I reverse calculate it using known crew. Once I have the difficulty I use it to calculate other crew success rate. Works fairly well.

WaldoMag

Non event shuttles are said to have no limit in difficulty. Therefore not sure where that upper limit came from. As long as you have a string successes difficulty will go up.

Capt. Chaos

WaldoMag wrote: »

@Capt. Chaos
I treat difficulty as unknown. Then I reverse calculate it using known crew. Once I have the difficulty I use it to calculate other crew success rate. Works fairly well.

That's basically how I approached it. Wrote a program that let me enter my event crew, mission traits, and difficulty setting. Played with it to get the difficulty that matched.

Capt. Reynolds

The difficulty level for faction event is equal to the vp you're running the shuttles for.
The AvgSkill the wiki talks about is not an average, but the sum of each crew stats.

Success rate probability = (1/(1+(EXP(3,5 * (0,5-(AvgSkill/Difficulty)))))) * 100

AvgSkill = sum(max_base_stat1 + 0,25 min_base_stat1 + skill_boost) * Event_bonus + .. + max_base_statn + 0,25 min_base_statn + skill_boost) * Event_bonus)

skill_boost is applied whether the crew has the skill or not, it is applied to the slot for the AND condition

For Faction events,

Difficulty = 2/(vp_level * n_crew), where
n_crew is the number of crew required for the mission
vp_level is exactly the victory point you're playing for

Well, the difficulty level is not always equal to the vp point you're playing for because DB likes to play with the formula for whatever reason... Missions for 2750vp have a difficulty level of 3000, I don't know the current difficulty rate for lower level missions, while for 3500 and 4000 vp these are exact difficulty levels.

The fun thing is that with the formula at hand you can see right away when bonus crew has wrong bonus or to plan ahead who to put in shuttles...

WaldoMag

Thanks Reynolds,
Will have something to check next faction event. And will get to see how it matches up for that event. ( will see how DB plays with the numbers)

Capt. Chaos

[DB:DB] ~SE~ Capt. Reynolds wrote: »

The difficulty level for faction event is equal to the vp you're running the shuttles for.
The AvgSkill the wiki talks about is not an average, but the sum of each crew stats.

Success rate probability = (1/(1+(EXP(3,5 * (0,5-(AvgSkill/Difficulty)))))) * 100

AvgSkill = sum(max_base_stat1 + 0,25 min_base_stat1 + skill_boost) * Event_bonus + .. + max_base_statn + 0,25 min_base_statn + skill_boost) * Event_bonus)

skill_boost is applied whether the crew has the skill or not, it is applied to the slot for the AND condition

For Faction events,

Difficulty = 2/(vp_level * n_crew), where
n_crew is the number of crew required for the mission
vp_level is exactly the victory point you're playing for

Well, the difficulty level is not always equal to the vp point you're playing for because DB likes to play with the formula for whatever reason... Missions for 2750vp have a difficulty level of 3000, I don't know the current difficulty rate for lower level missions, while for 3500 and 4000 vp these are exact difficulty levels.

The fun thing is that with the formula at hand you can see right away when bonus crew has wrong bonus or to plan ahead who to put in shuttles...

Thanks for that exposition, Capt. Reynolds! The wiki really isn't clear about that AvgSkill bit.
Apparently Difficulty = VP * crew /2 is what you meant. That gives the right answer for the 4000 VP missions I was running. Cheers!

Capt. Reynolds

Last time I had a 2250vp mission the difficulty rate was about 2600, that means that for beginner players overcoming certain threshold levels may be too difficult for what it's worth. It just increases the gap between long time players and new ones...
If anyone has data on lower level vp it would be very much appreciated.

I asked it already, but here it is again, DB, can you please lower the difficulty level to the actual vp?

For regular missions the formula is the same but the difficulty level increases by a step of 100 or similar. I never reached the 2500 difficulty level, I fail much earlier and then I actually fed up in trying to track it properly.

Capt. Reynolds

sorry in advance if any of you is already up to date with the other thread I'm linking:
https://forum.disruptorbeam.com/stt/discussion/3467/do-they-understand-what-95-chance-to-succeed-actually-means
I certainly don't read all threads...
Did any of you keep track of the last event shuttles? Got any data to share?