Sloane teaches math

By Blail Blerg, in Star Wars: Armada

From another topic: (by me)

Each Tie has a .25 chance of actually getting an accuracy. To strip a token at average chances (though this is still random chance, results will vary incredibly) (spend twice) you'd need this equation, lovely provided by my dear elementary teacher: 2 = .25x + .25(.25)x, solve for x. (Via Sloane rerolls off crits)

Although in essence, if you attack something with 6-8 ties, youre probably doing about 4-5 damage too. =)

Damage of 8 ties: 8(.5) + 8(.25)(.5) = 4 + 4(.25) = 4 + 1 = 5.

5 damage and 2 accuracies and 1 crit miss. (Though, I tend to double roll into crits a lot it seems)

[the answer is about 6.4 ties btw, on average. You're gonna want 7-8, and in your build maybe around 10 lol]

8 ties with Sloane will deal avg 5 dmg 2 acc for64 points

7 tie bombers no sloane = 7(.5)1 + 7(.25)2 = 7 dmg for 63 points

Sloane really helps with AA squadron builds. But she's extremely build-around: Otherwise I find her very underwhelming and takes up too much opportunity cost for good admirals. Even Motti or Moffy J. Moffy J is awesome.

*Maths

I think there's an underlying x factor with Sloane that is hard to quantify beyond the math. Spending these defense tokens is brutal, and ups the offensive output of every ship in your fleet. And Bombers are bad at fighting other squadrons, while TIEs and Interceptors are amazing at it, and Flight Controllers ups that value, while she makes their anti ship *almost* as good as bombers. I agree you have to build for her, but that is also the case for Mothma, Ackbar, Screed, Ozzel, and Jerry and all of those are really strong Admirals. Dodonna, Rieekan, and Motti IMO are the only true-blue generalists, but of course can be optimized through builds.

So if it takes 6.4 ties to discard a token, should we be running sloane with Intel officers to spread the damage token stressing ever further?

7 minutes ago, Ginkapo said:

So if it takes 6.4 ties to discard a token, should we be running sloane with Intel officers to spread the damage token stressing ever further?

you can, but i worry that you (general you, not you specifically) will end up building too heavily into token removal and not as much into actual killing of ships. Like Konstantine; you CAN put tractor beams and everything on all his ships, but all that does is mess with my speed. It doesnt make me dead, just slower.

Your game shouldn't be about stripping tokens, the odds are too low to count on. With that crit reroll you get you're using those ties as reasonably consistent ship damage dealers, with the bonus of spending a token. As long as you're commanding the squads so IF you get lucky and spend a token, you can then take advantage of it with a ship then you're in good shape.

5 minutes ago, geek19 said:

you can, but i worry that you (general you, not you specifically) will end up building too heavily into token removal and not as much into actual killing of ships. Like Konstantine; you CAN put tractor beams and everything on all his ships, but all that does is mess with my speed. It doesnt make me dead, just slower.

If you are going to do something then DO IT.

I run Konstantine with Tractors, Slicers, Grav Shift and Grav Well. Fully commited.

Run Sloane with Avenger or OP on a Raider with DC.

I don't think you need to force a discard with Sloane if you follow up with an attack. Spending 2 or 3 tokens and make your opponent decide if they want to discard them allowing further attacks to be more effective.

Here we go with the "chance" and "probability" confusion again....

The chance of each result is fixed, 4 hits, 2 crits, 2 accuracys, 50%, 25% & 25%, rerolling the dice does not change the chance of results, it remains fixed at 50%, 25% & 25%, what does increase or decrease is the probability of a specific result happening or not happening.

You can roll 4 dice, get zero accuracys, the 5th dice still only has a 25% chance of coming up an accuracy result. The probability of not rolling it decreases per reroll.

Other games the target number changes per reroll, for example first roll on a d8, 1-2 is a fail, 2nd roll 1-3 is a fail, 3rd roll 1-4 is a fail, here we see the chance of failure increasing with every reroll. (Xcom boardgame)

I have used Sloane and there is also a psychological effect as much as a real threat.

When you fire a round into a ship and they spend tokens they know those Ties can come screeming in and remove them.

ThIs can have an effect on when the spend their defensive tokens or not.

You also still need Bomber to some degrees to get consisent damage in and the hullpoints to do so over several turns.

Sloane effect on ships are good if you can make your opponent avoid spending tokens and you actually don't need to do anything. If they only save one or two damage on that brace they might actually not do it.

The effect on Aces are probably more severe if you run Howlrunner and Flight Controllers.

Edited by jorgen_cab

I'm only stating average maths. Not actual results. I did note that there is extremely high variance to those numbers above.

Actually, it might be a good time to get schooled: Who can actually explain probability/statistics well?

*ignores all this, because math was invented by Satan, as far as I'm concerned*

*takes Sloane and crams as many TIE and Interceptors in there as possible*

Never tell me the odds. Lol

If you rearrange the letters you get Santa.

Lupine, want to bat rep when youre done?

Oh and yes, I only tested Sloane vs no-sqds, just to see if it made full AA builds better. Haven't tried vs aces. I'm pretty sure sloane vs aces will be much more potent.

20120208160239-1_0.jpg

Right, so above is the Normal Distribution Curve of Bell Curve. The arithmetic mean is shown by the u symbol. This is the point where on average the data set will point to, in this case, that it takes 6.4 ties to remove a token. Now as you can probably summise, its also true that it mores likely for 5 ties to remove a token than 2 ties. So the graph flattens as it moves away from the mean.

The higher the variance, the flatter the curve, the lower the variance, the steeper and narrower the curve. Thats basically all there is to it.

If you want to go further, ask @Baltanok about the two tail test, or the other types of distribution. For most situations the Bell Curve is applicable.

2 hours ago, Blail Blerg said:

If you rearrange the letters you get Santa.

Lupine, want to bat rep when youre done?

Once I actually get a chance to try her, I'll post a report. Current fighter group I'm trying with is four defenders and three interceptors, iirc.

3 hours ago, OccasionallyCorrect said:

Your game shouldn't be about stripping tokens

Oh but the deffense tokens game is essential in this game.

25 minutes ago, Ginkapo said:

Right, so above is the Normal Distribution Curve of Bell Curve. The arithmetic mean is shown by the u symbol. This is the point where on average the data set will point to, in this case, that it takes 6.4 ties to remove a token. Now as you can probably summise, its also true that it mores likely for 5 ties to remove a token than 2 ties. So the graph flattens as it moves away from the mean.

The higher the variance, the flatter the curve, the lower the variance, the steeper and narrower the curve. Thats basically all there is to it.

If you want to go further, ask @Baltanok about the two tail test, or the other types of distribution. For most situations the Bell Curve is applicable.

Well, don't blow my skills all out of proportion. I have a couple of stats classes, and a whole lot of time using Excel.

The question I'm asking is: "how many TIEs do I need to strip one token from my target ship most of the time?" So we have to make some decisions about how much "most of the time" means. If you want to be 75% likely to take out Home One's Brace, you need more TIEs than if you are willing to accept only 60% likely. And if you want 98% likelihood? You need all the TIEs.

6 TIEs are 60% likely to strip a defense token. 8 are about 77% likely. 12 are 93% likely, 16 are 98% likely.

But those numbers don't take into account the enemy fighter screen. It's pretty unlikely that 16 TIE's are going to be able to attack Home One with no interference.

For those of you who'd like to do the calculations yourself, look up the BINOMDIST function in google sheets. The syntax of the function is:

BINOMDIST( # of successes, # of trials, Probability of Success, FALSE)

# of successes is how many successes you want

# of trials is how many TIEs & TIE/INTs you have

Prob success is the chance to get an accuracy, or 0.3125, including the reroll on crits.

False is to turn off the cumulative part of this function, because it's cumulative the wrong way for what we want.

This only gives you the probability of that exact number of successes, when you really want the probability of at least that number of successes. So, calculate the binomial distribution for 3, 4, 5 successes, etc, until you get to the total number of TIEs you are throwing at the enemy ship. This is where having a cumulative inverse binomial distribution function would be nice.

All the math, when the answer is simply 'all the TIES you can cram in there'. ????

I think the 5 damage you're doing from the Ties will be just as meaningful if not moreso than the 2 discards. I think the biggest loser for Sloan is the no-squadron build, because even if you could have given up some attacks on your squads, the fact that any defense tokens that you do spend might become easily discarded affects your decision points.

My sense is that Sloan is going to be at her best winning the squadron game with squads, while winning the ship game with ships, and then getting some support from your squads. If you do win the squad game early, then you get support from your squads against the opponent's ships.

10 minutes ago, Baltanok said:

Well, don't blow my skills all out of proportion. I have a couple of stats classes, and a whole lot of time using Excel.

The question I'm asking is: "how many TIEs do I need to strip one token from my target ship most of the time?" So we have to make some decisions about how much "most of the time" means. If you want to be 75% likely to take out Home One's Brace, you need more TIEs than if you are willing to accept only 60% likely. And if you want 98% likelihood? You need all the TIEs.

6 TIEs are 60% likely to strip a defense token. 8 are about 77% likely. 12 are 93% likely, 16 are 98% likely.

But those numbers don't take into account the enemy fighter screen. It's pretty unlikely that 16 TIE's are going to be able to attack Home One with no interference.

For those of you who'd like to do the calculations yourself, look up the BINOMDIST function in google sheets. The syntax of the function is:

BINOMDIST( # of successes, # of trials, Probability of Success, FALSE)

# of successes is how many successes you want

# of trials is how many TIEs & TIE/INTs you have

Prob success is the chance to get an accuracy, or 0.3125, including the reroll on crits.

False is to turn off the cumulative part of this function, because it's cumulative the wrong way for what we want.

This only gives you the probability of that exact number of successes, when you really want the probability of at least that number of successes. So, calculate the binomial distribution for 3, 4, 5 successes, etc, until you get to the total number of TIEs you are throwing at the enemy ship. This is where having a cumulative inverse binomial distribution function would be nice.

If you can force a ship to use one or two defense tokens by shooting at it first and THEN attack it with Tie-fighters, you will generally roll at least one accuracy by throwing about 1-4 squadrons at it (roughly 75% chance to get at least one accuracy). If all you want to do is force them to discard one of those tokens it might happen on the first roll or later (or not at all). But four squadrons will have a very high chance to force a discard at this time. Needing to throw six to eight squadrons on a ship is not very feasible as a basic strategy in my opinion to reach that effect. You need a ship to spend a token on its own first and then attack it with Tie-fighters.

You probably will not have more that 6-8 regular Tie-fighter/interceptors in a list so using it as a base strategy will not be advisable.

Edited by jorgen_cab

I still find that the best squadron for Sloane is Jendon/Stele. Each attack has around 60% chance for a accuracy. And if you don't have an accuracy, you will deal two damage with a crit. Even the so annoying double accuracy is no problem with Stele.

These two are not this good against squadrons as the 5 Tie Fighter would be (from the damage). But they are better against ships. And they have more HP compared to 5 Tie Fighters.

From my experience a mix of aces and Tie fighters is way better. A pure swarm of generic Tie Fighters is to weak or to hard to activate.

Edited by Tokra
1 minute ago, jorgen_cab said:

If you can force a ship to use one or two defense tokens by shooting at it and THEN attack it with Tie-fighters you will generally roll at least one accuracy by throwing about 1-4 squadrons at it (roughly 75% chance to get at least one accuracy). If all you want to do is force them to discard one of those tokens it might happen on the first roll or later (or not at all). But four squadrons will have a very high chance to force a discard at this time. Needing to throw six to eight squadrons on a ship is not very feasible as a basic strategy in my opinion to reach that effect. You need a ship to spend a token on its own first and then attack it with Tie-fighters.

You probably will not have more that 6-8 regular Tie-fighter/interceptors in a list so using it as a base strategy will not be advisable.

Yes, Flight Commander on an ISD/VSD might be nice. Or lead with the hitter & follow with the carrier.

Boarding Troopers, then squad dial would also be a way to discard defense tokens, if you want to do it in one activation.

51 minutes ago, Baltanok said:

Yes, Flight Commander on an ISD/VSD might be nice. Or lead with the hitter & follow with the carrier.

Tried both and it works... ;)

Last game I had a VSD with; Flight Commander, Quad Battery Turrets, Disposable Capacitor, NK-7 Ion Cannon and Veteran Gunners.

I actually managed to discard three defense tokens in one activation AT LONG RANGE!!!... but that was luck... ;)

Edited by jorgen_cab
55 minutes ago, Baltanok said:

Yes, Flight Commander on an ISD/VSD might be nice. Or lead with the hitter & follow with the carrier.

Boarding Troopers, then squad dial would also be a way to discard defense tokens, if you want to do it in one activation.

Are we sure about that? You still reveal the dial outside of close range. And then you arent resolving a squadron command post manuever, you are discarding it for boarding troopers. Im not convinced that works the way you think it does. Dont see it on the rules forum either.